Milan (Italy) - It is fair to say that Barcelona has not turned out the CPU core many hoped it would be and lots of people inside AMD would rather forget about this evolutionary step and quickly move on to the next architecture. Custom PC was able to squeeze some more information out of AMD about this next-gen architecture. It’s not much, but we now know for sure that it will be "completely different" from what AMD is selling today.
This confirmation comes from Giuseppe Amato, AMD’s technical director of sales & marketing for the EMEA region, one of the very few executives in the company who occasionally provide straight-forward answers.
"If I look at the next generation architecture of our CPU, then it will definitely not be, how can I say, comparable with the Phenom. It will look completely different," Custom PC quoted Amato. Of course, this does not really come as a surprise and was previously indicated. However, Amato also noted that the new architecture, which is expected to end up in the Bulldozer core, will "solve problems that today we think can never be addressed by hardware."
This comment is likely to fuel a new rumor mill and if it was up to us, we’d be betting the farm on addressing bandwidth challenges and simplify virtualization. We also believe that we could be seeing a GPU influence and AMD may be applying its Fusion LEGO concept to this new architecture - in order to apply massively parallel designs.
AMD would do well to introduce a brand new architecture, and not soon enough. The company has been losing market share to Intel over the last year, and much of what it gained during the Athlon days have been lost. AMD has also posted many consecutive losses in revenue in the order of billions of dollars. AMD recently announced a brand new initiative to produce and sell its own brand of computers. Company executives say that they have thought about it long and hard, and feel that it makes sense for AMD to introduce a business class group of computers with a guaranteed 24-months life cycle, something AMD says is a concern for most businesses.
Sunday, November 29, 2009
AMD officially announced the new family name for the 45nm Phenoms and Phenom II, at the firm’s Financial Analyst Day last week. More interestingly, re
ZoomBack in September there were industry reports that AMD was going to change their Phenom X4 and X3 model number scheme from a 4-digit model number to a 5-digit model number. This no longer appears to be the case, but there has been no official word from AMD about these model numbers yet.
Reported 45nm Phenom II 700/800/900 Processor Roadmap
Model Cores Frequency Codename Socket Total Cache TDP Release Date
X4 945 4 3.0GHz Deneb AM3 8 MB *125W Q2 2009
X4 940 4 3.0GHz Deneb AM2+ 8 MB 125W Jan. 8 2009
X4 925 4 2.8GHz Deneb AM3 8 MB 95W Feb. 2009
X4 920 4 2.8GHz Deneb AM2+ 8 MB 125W Jan. 8 2009
X4 910 4 2.6GHz Deneb AM3 8 MB 95W Feb. 2009
X4 810 4 2.6GHz Deneb AM3 6 MB 95W Feb. 2009
X4 805 4 2.5GHz Deneb AM3 6 MB 95W Feb. 2009
X3 720 3 2.8GHz Propus AM3 7.5 MB 95W Feb. 2009
X3 710 3 2.6GHz Propus AM3 7.5 MB 95W Feb. 2009
*Expreview reports the Phenom II X4 945 as a 125W part. While the source did not specify, the 945 is assumed to have an unlocked multiplier as it is the flagship model. The Phenom II X4 940 and 920 will also have unlocked multipliers.
An interesting thing to note is the similarity between this naming scheme and that used by Intel’s Core i7 family of processors, specifically the Phenom II X4 920 and 940. Hopefully this won’t be a source of confusion for the less knowledgeable crowd should they decide to build their own system. At least the 3-digit model numbers are easier to say and remember than the 5-digit numbers.
April will reportedly see the launch of the mainstream triple- and quad-core Phenom II X3 400 and X4 600 series, and June the budget dual-core X2 200 series chips.
We contacted AMD and a representative told us that while the name changes are true, the processor model numbers cannot be confirmed at this time. We were also told that official numbers will be unveiled very shortly
Reported 45nm Phenom II 700/800/900 Processor Roadmap
Model Cores Frequency Codename Socket Total Cache TDP Release Date
X4 945 4 3.0GHz Deneb AM3 8 MB *125W Q2 2009
X4 940 4 3.0GHz Deneb AM2+ 8 MB 125W Jan. 8 2009
X4 925 4 2.8GHz Deneb AM3 8 MB 95W Feb. 2009
X4 920 4 2.8GHz Deneb AM2+ 8 MB 125W Jan. 8 2009
X4 910 4 2.6GHz Deneb AM3 8 MB 95W Feb. 2009
X4 810 4 2.6GHz Deneb AM3 6 MB 95W Feb. 2009
X4 805 4 2.5GHz Deneb AM3 6 MB 95W Feb. 2009
X3 720 3 2.8GHz Propus AM3 7.5 MB 95W Feb. 2009
X3 710 3 2.6GHz Propus AM3 7.5 MB 95W Feb. 2009
*Expreview reports the Phenom II X4 945 as a 125W part. While the source did not specify, the 945 is assumed to have an unlocked multiplier as it is the flagship model. The Phenom II X4 940 and 920 will also have unlocked multipliers.
An interesting thing to note is the similarity between this naming scheme and that used by Intel’s Core i7 family of processors, specifically the Phenom II X4 920 and 940. Hopefully this won’t be a source of confusion for the less knowledgeable crowd should they decide to build their own system. At least the 3-digit model numbers are easier to say and remember than the 5-digit numbers.
April will reportedly see the launch of the mainstream triple- and quad-core Phenom II X3 400 and X4 600 series, and June the budget dual-core X2 200 series chips.
We contacted AMD and a representative told us that while the name changes are true, the processor model numbers cannot be confirmed at this time. We were also told that official numbers will be unveiled very shortly
Phenom 9950 Cracks The 4GHz Barrier On LN2
At the Games Convention in Leipzig, Germany, members of AwardFabrik.de managed to break 4GHz using liquid nitrogen cooling on an AMD Phenom 9950 processor. CPU-Z did not correctly validate the clockspeed however, and the AwardFabrik team are still unsure why this was the case. Zoom
Nevertheless, at a clock speed of 3952MHz, the team broke a SuperPi 1M record for AMD CPUs - 19.954 seconds - which is quite impressive for an AMD system considering SuperPi heavily favors the Core 2 architecture. Other hardware used included a Foxconn A79-S motherboard and an OCZ 1000W power supply.
Intel will soon change the overclocking scene with Core i7. Intel’s platform will finally lose its ancient Front Side Bus and adopt the QuickPath Interconnect for a high-bandwidth, point-to-point connection with the chipset which is dedicated to I/O traffic. The new interconnect will mean enthusiasts currently using Intel systems will have to break old habits and learn new methods to successfully overclock their CPU. This interconnect is reportedly quite temperamental when the reference clock is run much over its default speed of 133MHz, so this may pose problems for overclockers when using CPUs with upwards locked multipliers. Newer BIOS and motherboard revisions should improve stability of the QPI when overclocked.
SuperPi 1M may soon become obsolete with the current world record sitting at 7 seconds and quickly running out of headroom for faster times. Nehalem’s improved single-threaded performance when in Turbo Mode may push that world record further.
Nevertheless, at a clock speed of 3952MHz, the team broke a SuperPi 1M record for AMD CPUs - 19.954 seconds - which is quite impressive for an AMD system considering SuperPi heavily favors the Core 2 architecture. Other hardware used included a Foxconn A79-S motherboard and an OCZ 1000W power supply.
Intel will soon change the overclocking scene with Core i7. Intel’s platform will finally lose its ancient Front Side Bus and adopt the QuickPath Interconnect for a high-bandwidth, point-to-point connection with the chipset which is dedicated to I/O traffic. The new interconnect will mean enthusiasts currently using Intel systems will have to break old habits and learn new methods to successfully overclock their CPU. This interconnect is reportedly quite temperamental when the reference clock is run much over its default speed of 133MHz, so this may pose problems for overclockers when using CPUs with upwards locked multipliers. Newer BIOS and motherboard revisions should improve stability of the QPI when overclocked.
SuperPi 1M may soon become obsolete with the current world record sitting at 7 seconds and quickly running out of headroom for faster times. Nehalem’s improved single-threaded performance when in Turbo Mode may push that world record further.
AMD Plans to Release Eight-Core Processors Until 2009
Advanced Micro devices hardly managed to pose a threat for the chip manufacturer Intel lately, but recent information shows that the company is far from waving the surrender flag. Intel's
pre-briefing conference for the upcoming Spring Developers Forum in China revealed Intel's next goals on the company's roadmap. However, AMD might have quite some hidden aces up its sleeve.
According to Randy Allen, Corporate Vice President of the Server and Workstation Division at AMD, Intel's Nehalem chip is just the company's attempt at catching up with AMD's existing technology. He claims that the Barcelona server processors are currently shipping to a large amount of OEM and system integrators worldwide, and it is expected to sell extremely well.
The upcoming Nehalem chips will come with Intel's Quick Path Interconnect, which is a clone of AMD's HyperTransport 3.0 technology, and this is just one of the aspects that Intel is trying to copy. The Nehalem architecture also was inspired from AMD's cache structure, with small per-core Level 1 and Level 2 caches, that are both connected to a large shared Level 3 cache pool.
"We have level three cache, we've had integrated memory since 2003 and high speed serial links since 2006. I don't think there is anything new here," claimed Allen.
As far as the upcoming six-cored monster in the Dunnington family is concerned, AMD claims that they have the right weapons to defeat it. AMD's upcoming Shanghai chip will be manufactured at the 45-nanometer process node, and Allen claims that it will come with extra features, such as an enhanced HyperTransport interconnect and native support for DDR3 memory.
"We'll have four cores by the end of the year and you can expect to see eight cores in the 2009 timeframe," Allen concluded.
pre-briefing conference for the upcoming Spring Developers Forum in China revealed Intel's next goals on the company's roadmap. However, AMD might have quite some hidden aces up its sleeve.
According to Randy Allen, Corporate Vice President of the Server and Workstation Division at AMD, Intel's Nehalem chip is just the company's attempt at catching up with AMD's existing technology. He claims that the Barcelona server processors are currently shipping to a large amount of OEM and system integrators worldwide, and it is expected to sell extremely well.
The upcoming Nehalem chips will come with Intel's Quick Path Interconnect, which is a clone of AMD's HyperTransport 3.0 technology, and this is just one of the aspects that Intel is trying to copy. The Nehalem architecture also was inspired from AMD's cache structure, with small per-core Level 1 and Level 2 caches, that are both connected to a large shared Level 3 cache pool.
"We have level three cache, we've had integrated memory since 2003 and high speed serial links since 2006. I don't think there is anything new here," claimed Allen.
As far as the upcoming six-cored monster in the Dunnington family is concerned, AMD claims that they have the right weapons to defeat it. AMD's upcoming Shanghai chip will be manufactured at the 45-nanometer process node, and Allen claims that it will come with extra features, such as an enhanced HyperTransport interconnect and native support for DDR3 memory.
"We'll have four cores by the end of the year and you can expect to see eight cores in the 2009 timeframe," Allen concluded.
Making Sense of AMD's Processor Road Map
Advanced Micro Devices Inc. (AMD) Thursday gave journalists and analysts a peek at its road map of upcoming processors, including more details on the company's quad-core processors and plans for a family of server chips with 16 cores.
However, each of the processors and chip packages on AMD's road map are tagged with code names that do little to explain what each chip actually does. To help you make sense of what each code name represents, here's a quick overview of what AMD has in store, and when each chip is expected to hit the market:
Barcelona: The code name given to AMD's upcoming Quad-Core Opteron processors, made using a 65-nanometer (nm) process. AMD plans to ship the first Barcelona chips next month, with the first servers based on the processors appearing on the market in September.
Bobcat: Code name for a future low-power CPU architecture for mobile devices such as ultramobile PCs and consumer electronics products. Will consume from 1 watt to 10 watts of power. Due in 2009.
Bulldozer: Code name for a CPU core designed for servers and clients that consumes from 10 watts to 100 watts of power. Set for release in 2009.
Eagle: Code name for an upcoming notebook chip package based on the Falcon processor. To ship in 2009.
Falcon: Code name for the first Fusion chip that will combine a CPU and graphics processor. Designed for laptops, Falcon will offer up to four Bulldozer cores. Due to ship in 2009.
Fusion: The code name for AMD processors that combine multiple components with the aim of lowering power consumption and improving performance. The first Fusion chips, called Falcon, will ship in 2009.
Griffin: Code name for an upcoming dual-core mobile processor. To ship in 2008.
Hardcastle: The code name given to upcoming chip packages designed for business users, including Perseus and Puma.
Leo: Code name for a desktop chip package based on the 45nm Phenom processor, which will offer 6MB of cache. Set for release in 2008.
Perseus: Upcoming desktop chip package designed for business users. To ship in 2008.
Phenom: The brand name for AMD's quad-core desktop processors, which are slated to start shipping during the fourth quarter of 2007.
Puma: Chip package for laptops based on the Griffin processor. To debut in 2008.
Ridgeback: Code name for AMD's 45nm desktop processors. Will include 6MB of cache. To be released in mid-2008.
Sandtiger: The code name for a family of server chips with eight to 16 cores that will be released in 2009. Will be produced using a 45nm process and include supp
(Double Data Rate 3) memory. Due in 2009.
Shanghai: Code name for the 45-nanometer successor to Barcelona. The quad-core Shanghai will offer some architectural improvements over Barcelona, as well as 6MB of cache. To be released in mid-2008.
Spider: Desktop chip package based on 65nm Phenom processors with 2MB of cache. To be released in 2007.
AMD’s CPU Roadmap, 2008-2011
While speaking at their annual financial analyst day today, AMD provided members of the press a glimpse of their latest CPU and GPU roadmaps. AMD’s GPU update was rather slim, with the company reiterating previously disclosed guidance from CEATEC for next year: DirectX 11 and 40-nm are all on tap for release with their next-generation RV870 GPU in 2009.
Because these details were already disclosed last month in Japan, the real news wasn’t made in graphics, rather AMD has made sweeping changes to their CPU roadmap. The following chart quickly summarizes the changes:
As you can see, AMD has totally reworked their mobile CPU roadmap, with four new cores slated for release between now and the end of 2010, “Caspian”, “Conesus”, “Champlain”, and “Geneva”. The desktop roadmap is dominated by Deneb, which we now know has been officially branded as Phenom II. Then in 2011, AMD plans to introduce Orochi for high-end desktops as well as their first CPU-GPU hybrid processors (AMD refers to them as APUs or accelerated processing units) that combine traditional CPU processing with graphics duties. These processors are codenamed “Llano” and “Ontario”.
Previously AMD’s roadmap indicated we’d see the first CPU-GPU APUs in 2009 with the introduction of their Swift processors. At the time of AMD’s ATI
acquisition in 2006, the company made the case that their 45-nm manufacturing process was the right time for the CPU to gain graphics functionality, now the company feels that the tech isn’t feasible until they shrink to 32-nm with Llano and Ontario, with the Swift core disappearing entirely from AMD’s new roadmap.
Until we see the introduction of the first APUs however, AMD plans to service the mobile segment with more conventional 45-nm processors.
Up first will be Caspian, Conesus, and Huron, all slated to be introduced in the second half of 2009.
Caspian is AMD’s solution for mainstream notebooks
. Like today’s Griffin-based Turion X2 CPUs, the Caspian core will feature two processing cores with 2MB of cache and DDR2 memory. Unfortunately AMD wouldn’t provide further details on Caspian’s architecture (i.e. is the core simply a die-shrink of Griffin or are its origins based on Phenom II), when asked the company declined to answer the question. However, the “S1G3” (socket 1 generation 3) designation in the second chart implies that it’s definitely more advanced than just a simple die-shrink of Griffin, although it may not be quite as complex as Phenom II to keep power consumption and die size down.
Caspian is AMD’s solution for mainstream notebooks
. Like today’s Griffin-based Turion X2 CPUs, the Caspian core will feature two processing cores with 2MB of cache and DDR2 memory. Unfortunately AMD wouldn’t provide further details on Caspian’s architecture (i.e. is the core simply a die-shrink of Griffin or are its origins based on Phenom II), when asked the company declined to answer the question. However, the “S1G3” (socket 1 generation 3) designation in the second chart implies that it’s definitely more advanced than just a simple die-shrink of Griffin, although it may not be quite as complex as Phenom II to keep power consumption and die size down.
Caspian will ship with AMD’s next-generation Tigris platform, which replaces today’s Puma. Tigris will feature a new RS880M North Bridge + SB710 South Bridge. Once again AMD wasn’t specific on details when it comes to RS880M, so we don’t know if it’s derived from RV770 or RV870 technology. AMD also expects manufacturers to offer Tigris-based notebooks with discrete graphics, presumably using upcoming Mobility Radeon 4000 parts.
Moving further down the rung, for the ultraportable/mini-notebook (netbook) segment, AMD plans to offer Conesus in 2H’09. Yesterday, rumors were swirling that AMD would offer a CPU to rival Intel’s Atom CPU. Today AMD reiterated their position on the topic: NO! The company still feels consumers want a more fully-fledged notebook experience with no compromises; as anyone who has tried to seriously game or encode a video on a netbook can tell you, Intel’s Atom CPU really isn’t up to the task for these duties. Rather than offer a dedicated CPU for the netbook segment, AMD plans to counter with lower priced ultraportable systems based around Conesus and Huron.
Conesus is a dual-core part with 1MB of cache and DDR2, while Huron is a single-core CPU. Conesus will rely on AMD’s upcoming Congo platform, which consists of RS780M + SB710 South Bridge, while Huron will ship with Yukon, which will utilize RS690E + SB600. To minimize power consumption we’d guess that these chipsets will be built using TSMC’s 40-nm manufacturing process, although AMD wouldn’t provide specifics on that topic either.
Graphics cards News
AMD plots 16-core super-CPU for 2009
"Bulldozer" to be "highest performing processor core" ever
is preparing an all-new PC processor with up to 16 execution cores. Due out in the first half of 2009, the new architecture is codenamed Bulldozer. In an official announcement, AMD said Bulldozer will be its first substantially new CPU core since the original Athlon 64 processor of 2003. And it'll be the first ground up architectural redesign since the K7 Athlon of 1999.
In that context, it's likely the Bulldozer architecture will provide the foundations of AMD CPUs for many years to come. In other words, the architecture is rather important. The news is the latest in a flurry of announcements from AMD's PR machine in the past week. AMD has already revealed plans for a second-gen quad-core processor known as Shanghai and demoed its upcoming Phenom quad-core desktop chip running at 3GHz.
Meanwhile, the world is still waiting for Barcelona, AMD's first generation quad-core CPU , to appear. AMD has promised that Barcelona will launch later this August.
Bulldozer is the name, crushing Intel is the game
But what about Bulldozer? The big news is that it will form the basis of AMD's first massively multi-core PC processor with up to 16 execution cores. Bulldozer will also be fully compatible with AMD's so-called M-SPACE modular CPU design.
Along with traditional PC processors, therefore, expect to see AMD Fusion CPUs powered by Bulldozer cores but also offering a range of specialised processing units. Think graphics processing cores and high-definition video decoding engines and you'll get an idea of the sort of additional functionality Bulldozer-based Fusion processors will deliver.
As for the main execution cores themselves, they retain the same basic out-of-order superscalar design as AMD's existing PC processors. However, Bulldozer will utilise a deeper instruction pipeline. That's a measure traditionally introduced to allow higher clockspeeds. However, deeper pipelines typically also reduce clock-for-clock performance.
Bulldozer also receives a range of new instructions designed to accelerate media processing and performance in high performance computer clusters. With as many as 16 cores humming away, keeping Barcelona fed with data will obviously be a tough task. In response, AMD says the chip will benefit from "highly scalable memory and I/O performance".
Specifically, that means a new version of AMD's Direct Connect technology along with four HyperTransport 3.0 links per processor. Support for ultra fast DDR3 memory as well as AMD's G3MX memory extender technology will further boost the data and bandwidth available to Bulldozer. AMD's G3MX technology is designed to allow increased system memory without resorting to Intel 's power-hungry FBDIMM approach.
How fast is she, mister?
It all sounds pretty impressive on paper. But how fast will this 16-core chip be in practice? Well, according to AMD, Bulldozer is designed to be nothing less than "the highest performing single and multi-threaded compute core in history".
If AMD is to be believed, Bulldozer will improve upon every metric of CPU performance. From performance per watt to outright multi-threaded performance and old school single-threaded oomph, it's promised Bulldozer will be the new king.
Of course, by 2009 Intel will be shipping some fairly exotic kit of its own. A major revision of Intel's Core architecture, complete with an integrated memory controller and HyperTransport-bashing Common Serial Interface, is pencilled in for 2008 with a further revised variant in 2009.
What's more, with that longer instruction pipeline in mind, it will be interesting to see how Bulldozer pulls off improved single-threaded performance. Rumours are currently circulating that Bulldozer may be capable of thread-fusing or using multiple cores to compute a single thread. Thread fusing is one of the holy grails of PC processing. If Bulldozer is indeed capable of such a feat, the future could be very bright indeed for AMD.
"Bulldozer" to be "highest performing processor core" ever
is preparing an all-new PC processor with up to 16 execution cores. Due out in the first half of 2009, the new architecture is codenamed Bulldozer. In an official announcement, AMD said Bulldozer will be its first substantially new CPU core since the original Athlon 64 processor of 2003. And it'll be the first ground up architectural redesign since the K7 Athlon of 1999.
In that context, it's likely the Bulldozer architecture will provide the foundations of AMD CPUs for many years to come. In other words, the architecture is rather important. The news is the latest in a flurry of announcements from AMD's PR machine in the past week. AMD has already revealed plans for a second-gen quad-core processor known as Shanghai and demoed its upcoming Phenom quad-core desktop chip running at 3GHz.
Meanwhile, the world is still waiting for Barcelona, AMD's first generation quad-core CPU , to appear. AMD has promised that Barcelona will launch later this August.
Bulldozer is the name, crushing Intel is the game
But what about Bulldozer? The big news is that it will form the basis of AMD's first massively multi-core PC processor with up to 16 execution cores. Bulldozer will also be fully compatible with AMD's so-called M-SPACE modular CPU design.
Along with traditional PC processors, therefore, expect to see AMD Fusion CPUs powered by Bulldozer cores but also offering a range of specialised processing units. Think graphics processing cores and high-definition video decoding engines and you'll get an idea of the sort of additional functionality Bulldozer-based Fusion processors will deliver.
As for the main execution cores themselves, they retain the same basic out-of-order superscalar design as AMD's existing PC processors. However, Bulldozer will utilise a deeper instruction pipeline. That's a measure traditionally introduced to allow higher clockspeeds. However, deeper pipelines typically also reduce clock-for-clock performance.
Bulldozer also receives a range of new instructions designed to accelerate media processing and performance in high performance computer clusters. With as many as 16 cores humming away, keeping Barcelona fed with data will obviously be a tough task. In response, AMD says the chip will benefit from "highly scalable memory and I/O performance".
Specifically, that means a new version of AMD's Direct Connect technology along with four HyperTransport 3.0 links per processor. Support for ultra fast DDR3 memory as well as AMD's G3MX memory extender technology will further boost the data and bandwidth available to Bulldozer. AMD's G3MX technology is designed to allow increased system memory without resorting to Intel 's power-hungry FBDIMM approach.
How fast is she, mister?
It all sounds pretty impressive on paper. But how fast will this 16-core chip be in practice? Well, according to AMD, Bulldozer is designed to be nothing less than "the highest performing single and multi-threaded compute core in history".
If AMD is to be believed, Bulldozer will improve upon every metric of CPU performance. From performance per watt to outright multi-threaded performance and old school single-threaded oomph, it's promised Bulldozer will be the new king.
Of course, by 2009 Intel will be shipping some fairly exotic kit of its own. A major revision of Intel's Core architecture, complete with an integrated memory controller and HyperTransport-bashing Common Serial Interface, is pencilled in for 2008 with a further revised variant in 2009.
What's more, with that longer instruction pipeline in mind, it will be interesting to see how Bulldozer pulls off improved single-threaded performance. Rumours are currently circulating that Bulldozer may be capable of thread-fusing or using multiple cores to compute a single thread. Thread fusing is one of the holy grails of PC processing. If Bulldozer is indeed capable of such a feat, the future could be very bright indeed for AMD.
AMD Athlon II X2 Processor Details
• The AMD Athlon II X2 250 performs exceptionally well when combined with AMD chipsets and integrated graphics solutions to create an all-AMD platform. Platforms featuring all-AMD technology can deliver up to twice the graphics performance of those with Intel integrated graphics.¹
• Windows® 7 is optimized for multi-core processors like AMD Athlon™ II processors to give consumers an amazingly fast, simple and engaging PC experience.** For example, Windows 7 is tuned to make the most of these new processors’ power management features, such as AMD PowerNow!™ 3.0 technology. AMD power management technologies, in combination with Windows 7, can help OEMs and partners to build exceptionally green, cool and quiet PCs.
• Based on AMD’s acclaimed 45nm process technology, the AMD Athlon II dual-core processor has a TDP of 65W and can slash power consumption by up to 50 percent when doing basic tasks, up to 40 percent when running heavy workloads and up to 50 percent when at idle.²
AMD Phenom II X2 550 Black Edition Details
• AMD Black Edition processors, like the AMD Phenom™ II X2 550, help users to take control and unleash the maximum potential of Dragon platform technology’s unprecedented performance tuning capabilities.* The same massive headroom that set world records in recent months is at users’ finger tips, offering impressive performance at a price the competition can’t beat.³
• Users can also maximize their overclocking experience by utilizing the new features and capabilities of AMD OverDrive™ 3.0, designed to enable quick and effective tuning of their PC experience for optimal performance.*
• With dual-, triple- or quad-core processors, AMD provides platform level solutions at multiple price points, each of which exceeds expectations for virtually any user.
• Windows® 7 is optimized for multi-core processors like AMD Athlon™ II processors to give consumers an amazingly fast, simple and engaging PC experience.** For example, Windows 7 is tuned to make the most of these new processors’ power management features, such as AMD PowerNow!™ 3.0 technology. AMD power management technologies, in combination with Windows 7, can help OEMs and partners to build exceptionally green, cool and quiet PCs.
• Based on AMD’s acclaimed 45nm process technology, the AMD Athlon II dual-core processor has a TDP of 65W and can slash power consumption by up to 50 percent when doing basic tasks, up to 40 percent when running heavy workloads and up to 50 percent when at idle.²
AMD Phenom II X2 550 Black Edition Details
• AMD Black Edition processors, like the AMD Phenom™ II X2 550, help users to take control and unleash the maximum potential of Dragon platform technology’s unprecedented performance tuning capabilities.* The same massive headroom that set world records in recent months is at users’ finger tips, offering impressive performance at a price the competition can’t beat.³
• Users can also maximize their overclocking experience by utilizing the new features and capabilities of AMD OverDrive™ 3.0, designed to enable quick and effective tuning of their PC experience for optimal performance.*
• With dual-, triple- or quad-core processors, AMD provides platform level solutions at multiple price points, each of which exceeds expectations for virtually any user.
AMD Introduces Next Generation AMD Athlon™ II Processor, Adds Dual Core to Record-Setting AMD Phenom™ II
AMD Athlon™ II processor delivers new native dual-core architecture, efficient 45nm technology and 3 GHz performance at an affordable price
AMD Phenom™ II X2 Black Edition processor combines value and unlocked potential for gamers and tuners on a budget
COMPUTEX 2009 (TAIPEI, Taiwan) --6/2/2009 Bringing its acclaimed 45nm technology to new high-volume processor designs, AMD (NYSE: AMD) today announced two new dual-core desktop processors. Building on 10 years of AMD Athlon™ processor innovation, the new 45nm AMD Athlon™ II X2 250 processor gives mainstream consumers exceptional performance, efficiency and value. For enthusiasts and overclockers, AMD also announces the AMD Phenom™ II X2 550 Black Edition processor, the first ever dual-core AMD Phenom II CPU.* With this latest addition to the AMD Phenom II processor family, users can now experience the power of AMD platform technology, codenamed “Dragon,” with dual-, triple- and quad-core configurations.
AMD Phenom™ II X2 Black Edition processor combines value and unlocked potential for gamers and tuners on a budget
COMPUTEX 2009 (TAIPEI, Taiwan) --6/2/2009 Bringing its acclaimed 45nm technology to new high-volume processor designs, AMD (NYSE: AMD) today announced two new dual-core desktop processors. Building on 10 years of AMD Athlon™ processor innovation, the new 45nm AMD Athlon™ II X2 250 processor gives mainstream consumers exceptional performance, efficiency and value. For enthusiasts and overclockers, AMD also announces the AMD Phenom™ II X2 550 Black Edition processor, the first ever dual-core AMD Phenom II CPU.* With this latest addition to the AMD Phenom II processor family, users can now experience the power of AMD platform technology, codenamed “Dragon,” with dual-, triple- and quad-core configurations.
AMD Phenom FX CPU Expected in 2009
Looking to put pressure on Intel once again in the high-end market, sources say AMD plans on releasing the AMD Phenom FX in mid-2009 on its new AM3 socket platform. AMD previously acknowledged that AM3 processors will work in AM2 sockets but not the reverse.
AMD is attempting to bring back their FX-line of processors in mid-2009 according to documents Tom’s Hardware saw. Although little is known about the specifications of these processors, we do know they will be based on the Deneb FX core, feature four processing engines, have shared level-three cache and be based on a 45 nm manufacturing process.
The FX-branding used by AMD in the past had represented products of unmatched performance for enthusiasts who were willing to spend a hefty premium to have the extra performance. The Athlon FX line-up was the last to use processors with the FX-branding and were in production from 2003 to 2006. The Athlon FX processors featured unlocked multipliers and offered the highest clockings of the AMD processors available.
With the great success of Intel’s Core 2 Duo processors in 2006, AMD lost the performance crown it once had held and was forced to focus on mid and entry-level offerings instead, ending production of the Althon FX. AMD continued to offer processors that featured unlocked processors with its Black Edition line-up of Athlon and Phenom processors, which were offered at a fair price for those looking for an overclockable processor and still loyal to AMD. Much the same as AMD’s FX processors, Intel’s Extreme Edition processors offer high clock rates and unlocked multipliers, with similarly high prices.
If history serves us well, these upcoming FX processors will likely carry the name AMD Phenom FX, carry with them a steep price and offer clock-speeds higher than their future mainstream brethren. Future AMD processors produced using 45nm process technology are expected to reach 3.0GHz in speed, possibly putting the AMD Phenom FX processors at over 3GHz.
AMD is attempting to bring back their FX-line of processors in mid-2009 according to documents Tom’s Hardware saw. Although little is known about the specifications of these processors, we do know they will be based on the Deneb FX core, feature four processing engines, have shared level-three cache and be based on a 45 nm manufacturing process.
The FX-branding used by AMD in the past had represented products of unmatched performance for enthusiasts who were willing to spend a hefty premium to have the extra performance. The Athlon FX line-up was the last to use processors with the FX-branding and were in production from 2003 to 2006. The Athlon FX processors featured unlocked multipliers and offered the highest clockings of the AMD processors available.
With the great success of Intel’s Core 2 Duo processors in 2006, AMD lost the performance crown it once had held and was forced to focus on mid and entry-level offerings instead, ending production of the Althon FX. AMD continued to offer processors that featured unlocked processors with its Black Edition line-up of Athlon and Phenom processors, which were offered at a fair price for those looking for an overclockable processor and still loyal to AMD. Much the same as AMD’s FX processors, Intel’s Extreme Edition processors offer high clock rates and unlocked multipliers, with similarly high prices.
If history serves us well, these upcoming FX processors will likely carry the name AMD Phenom FX, carry with them a steep price and offer clock-speeds higher than their future mainstream brethren. Future AMD processors produced using 45nm process technology are expected to reach 3.0GHz in speed, possibly putting the AMD Phenom FX processors at over 3GHz.
Monday, January 19, 2009
Quad-Core AMD Opteron processors leave Barcelona
AMD has released its long-awaited native quad-core processor, previously codenamed "Barcelona".
Describing the new Quad-Core AMD Opteron as the "world's most advanced x86 processor", AMD claims it as the first native x86 quad-core microprocessor, meaning the four processing cores share a single die of silicon.
The company also highlights other aspects of the 55- and 75-watt server chips. It claims gains for energy efficiency, a 50 percent increase in integer and floating-point performance, and improved virtualisation support. Note that the server chips also share the same power and thermal envelopes as their dual-core predecessors.
"Today marks one of the great milestones in microprocessor achievement as AMD again raises expectations for industry-standard computing," said AMD chairman and chief executive officer, Hector Ruiz. "We've worked closely with our customers and partners to design a new generation of processing solutions embodied by today's Quad-Core AMD Opteron processor - a four-way winner in performance, energy efficiency, virtualisation and investment protection. Early customer response has been extremely positive."
leshing out the details on the energy efficiency claims, AMD highlights the use of proprietary "CoolCore Technology" (turning off unused parts of the processor) an independent core-specific enhancement to its PowerNow! Technology (allowing cores to vary their clock frequency depending on application requirements) and Dual Dynamic Power Management (DDPM), which allows the cores and memory controllers to operate on different voltages, again determined by usage.
Sun, HP, IBM and Dell will be among those shipping systems based on the new server chip will be shipping. Details of AMD's pricing for the Quad-Core Opteron will be found at www.amd.com/pricing (based on 1,000 unit orders).
More information on AMD's quad-core processing can be found on the company's website.
For its part, Intel released updated quad-core server microprocessors last week - the Quad-Core Xeon 7300 series, its first multi-processor (MP) chips based on its now standard Core micro-architecture.
See also: Electronics Weekly's focus on microprocessors, a roundup of content related to microprocessor technologies and developments.
Describing the new Quad-Core AMD Opteron as the "world's most advanced x86 processor", AMD claims it as the first native x86 quad-core microprocessor, meaning the four processing cores share a single die of silicon.
The company also highlights other aspects of the 55- and 75-watt server chips. It claims gains for energy efficiency, a 50 percent increase in integer and floating-point performance, and improved virtualisation support. Note that the server chips also share the same power and thermal envelopes as their dual-core predecessors.
"Today marks one of the great milestones in microprocessor achievement as AMD again raises expectations for industry-standard computing," said AMD chairman and chief executive officer, Hector Ruiz. "We've worked closely with our customers and partners to design a new generation of processing solutions embodied by today's Quad-Core AMD Opteron processor - a four-way winner in performance, energy efficiency, virtualisation and investment protection. Early customer response has been extremely positive."
leshing out the details on the energy efficiency claims, AMD highlights the use of proprietary "CoolCore Technology" (turning off unused parts of the processor) an independent core-specific enhancement to its PowerNow! Technology (allowing cores to vary their clock frequency depending on application requirements) and Dual Dynamic Power Management (DDPM), which allows the cores and memory controllers to operate on different voltages, again determined by usage.
Sun, HP, IBM and Dell will be among those shipping systems based on the new server chip will be shipping. Details of AMD's pricing for the Quad-Core Opteron will be found at www.amd.com/pricing (based on 1,000 unit orders).
More information on AMD's quad-core processing can be found on the company's website.
For its part, Intel released updated quad-core server microprocessors last week - the Quad-Core Xeon 7300 series, its first multi-processor (MP) chips based on its now standard Core micro-architecture.
See also: Electronics Weekly's focus on microprocessors, a roundup of content related to microprocessor technologies and developments.
AMD Athlon 64 X2 6000+ Review
The AMD Athlon 64 X2 6000+ processor is the top of the line for the AMD Athlon 64 X2 line of CPU's. At 125 watts of power this processor begs for some good cooling and a very steady hand when it comes to overclocking. But I won't be overclocking this beast, it would be a shame to mess with something
this good straight from the box.
The AMD Athlon 64 X2 6000+ processor is a 3 Ghz processor that rivals the Intel E6600 CPU and clearly is the best of the AMD AM2 sockets at this time. The 6000+ processor is built with the L2 cache at 1 MB per core that gives you plenty of internal memory for the processor.
I'd like to thank AMD for letting me review this great processor, now on to the review.
AMD as well as plenty of other reviews give all this kind of information that I guess I will include for the real tech guys who are reading:
Frequency: 3.0GHz
Manufactured: Fab 30 and 36 / Dresden, Germany
Process Technology: 90-nanometer DSL SOI (silicon-on-insulator) technology
Packaging: Socket AM2 (940-pin organic micro PGA)
HyperTransport technology: Supports single HT link - up to 8.0 GB/sec per link bandwidth
Memory Controller: Shared integrated 128-bit wide memory controller
CPU to Memory Controller: 3.0GHz
Supported Memory Speeds: DDR 2 memory up to and including PC2 6400 (DDR2-800) unbuffered
HyperTransport Links: 1
HyperTransport Spec: 2.0GHz (2x 1000MHz / DDR)
Total Processor bandwidth: Up to 20.8 GB/sec [8.0GB/sec HyperTransport + 12.8GB/sec dual-channel memory]
L1 Cache Sizes: 64K - L1 instruction + 64K - L1 data cache per-core (256KB total L1)
L2 Cache Sizes: 1MB L2 data cache per-core (2MB total dedicated L2 cache)
Approximate Transistor count: 227.4 million
Approximate Die Size: 218mm2
Nominal Voltage: 1.35-1.40V
Max Thermal Power: 125 W
Max Ambient Case Temp: 55o Celsius to 63o Celsius
Max Icc (processor current): 90.4 A
Min P-State (with C'n'Q): 1.0 GHz
o Nominal Voltage @ min P-state: 1.1V
o Max Thermal Power @ min P-state: 36.4 W
o Max Icc @ min P-state: 30.4 A
this good straight from the box.
The AMD Athlon 64 X2 6000+ processor is a 3 Ghz processor that rivals the Intel E6600 CPU and clearly is the best of the AMD AM2 sockets at this time. The 6000+ processor is built with the L2 cache at 1 MB per core that gives you plenty of internal memory for the processor.
I'd like to thank AMD for letting me review this great processor, now on to the review.
AMD as well as plenty of other reviews give all this kind of information that I guess I will include for the real tech guys who are reading:
Frequency: 3.0GHz
Manufactured: Fab 30 and 36 / Dresden, Germany
Process Technology: 90-nanometer DSL SOI (silicon-on-insulator) technology
Packaging: Socket AM2 (940-pin organic micro PGA)
HyperTransport technology: Supports single HT link - up to 8.0 GB/sec per link bandwidth
Memory Controller: Shared integrated 128-bit wide memory controller
CPU to Memory Controller: 3.0GHz
Supported Memory Speeds: DDR 2 memory up to and including PC2 6400 (DDR2-800) unbuffered
HyperTransport Links: 1
HyperTransport Spec: 2.0GHz (2x 1000MHz / DDR)
Total Processor bandwidth: Up to 20.8 GB/sec [8.0GB/sec HyperTransport + 12.8GB/sec dual-channel memory]
L1 Cache Sizes: 64K - L1 instruction + 64K - L1 data cache per-core (256KB total L1)
L2 Cache Sizes: 1MB L2 data cache per-core (2MB total dedicated L2 cache)
Approximate Transistor count: 227.4 million
Approximate Die Size: 218mm2
Nominal Voltage: 1.35-1.40V
Max Thermal Power: 125 W
Max Ambient Case Temp: 55o Celsius to 63o Celsius
Max Icc (processor current): 90.4 A
Min P-State (with C'n'Q): 1.0 GHz
o Nominal Voltage @ min P-state: 1.1V
o Max Thermal Power @ min P-state: 36.4 W
o Max Icc @ min P-state: 30.4 A
AMD Athlon 64 X2 6000+ Review
Now that the required specifications are done with I will get to the meat of the matter, installing and use. The processor is easy to install and setup for the heat sink. You will have to take into consideration the heat requirements of this processor if you're thinking of using one in your system.
The processor uses up to 125 watts of power so you need a decent cooling solution but it does not need to be the best out there for normal operation.
I do not and will not be overclocking the AMD Athlon 64 X2 6000+ in this review. The frame rates I received for testing and playing since I have had this processor are more than adequate and I just do not want to take the time for a few frame rates increase. I do not like to overclock components as the benefits do not equal the time spent when I have so much better things to do with my time.
I tried the stock fan and heat sink that came with my system for the AMD Athlon 64 X2 5200+ and found that running any games or CPU intensive programs would cause errors due to heat. The core would get up towards 70 degrees Celsius and would shut down. I also tried a lower model of CPU cooler from EverCool, the Shark, and it did much better than the stock heat sink. At some very intensive testing and benchmarking it does not keep the 6000+ as cool as I would like so I tried out the Buffalo from EverCool and it does very well. I have run the processor through its paces with both games and benchmark programs and the chip never gets over 60 degrees Celsius.
I would recommend a very good CPU cooler for this much power but you really don't have to buy the best on the market, just a good one. Installing the processor is easy and the coolers I used were also very easy to install. Just like any computer installation you need to learn what you're doing if you have never done this before but it is not that difficult to replace a CPU. If you don't know how to do this kind of upgrade it is not even as difficult as changing a video card as you do not have to install any drivers. You just swap out the processor and install the CPU cooler.
As long as you have the correct processor and know that your motherboard and system will handle the new processor you simply remove the CPU cooler, wipe off the old thermal compound and install the new one. You will need to put some thermal paste on that should come with your new CPU cooler. If not you can buy a tube for about ten dollars at a computer store or order some from the internet. The best brand out is Artic Silver 3 but you can use the more common thermal compound and just smear a small amount on the CPU.
The processor uses up to 125 watts of power so you need a decent cooling solution but it does not need to be the best out there for normal operation.
I do not and will not be overclocking the AMD Athlon 64 X2 6000+ in this review. The frame rates I received for testing and playing since I have had this processor are more than adequate and I just do not want to take the time for a few frame rates increase. I do not like to overclock components as the benefits do not equal the time spent when I have so much better things to do with my time.
I tried the stock fan and heat sink that came with my system for the AMD Athlon 64 X2 5200+ and found that running any games or CPU intensive programs would cause errors due to heat. The core would get up towards 70 degrees Celsius and would shut down. I also tried a lower model of CPU cooler from EverCool, the Shark, and it did much better than the stock heat sink. At some very intensive testing and benchmarking it does not keep the 6000+ as cool as I would like so I tried out the Buffalo from EverCool and it does very well. I have run the processor through its paces with both games and benchmark programs and the chip never gets over 60 degrees Celsius.
I would recommend a very good CPU cooler for this much power but you really don't have to buy the best on the market, just a good one. Installing the processor is easy and the coolers I used were also very easy to install. Just like any computer installation you need to learn what you're doing if you have never done this before but it is not that difficult to replace a CPU. If you don't know how to do this kind of upgrade it is not even as difficult as changing a video card as you do not have to install any drivers. You just swap out the processor and install the CPU cooler.
As long as you have the correct processor and know that your motherboard and system will handle the new processor you simply remove the CPU cooler, wipe off the old thermal compound and install the new one. You will need to put some thermal paste on that should come with your new CPU cooler. If not you can buy a tube for about ten dollars at a computer store or order some from the internet. The best brand out is Artic Silver 3 but you can use the more common thermal compound and just smear a small amount on the CPU.
AMD: Shanghai CPU Presentation in Berlin
Even though the launching of “Barcelona” (previous server processor from AMD) was in city Barcelona, Spain, “Shanghai” was launched in Berlin, Germany, and not on Far East (as we expected). Presentation opened Ian McNaughton and opening word continued Emilio Ghilardi, Senior Vice President and General Manager, AMD EMEA, who will succeed next year Mr. Henry Richard on same position but worldwide. Considering that he’s relatively newcomer in AMD, with only few spent months in company, Ghilardi first spoke about reasons why he accepted new position regardless on situation in AMD.
Ghilardi said that he was impressed by Dirk Meyer, new CEO and president of company, with his exceptional vision for further AMD development. The second reason was AMD’s success that regardless on pressure created from competitors, big financial crisis and other circumstances, reflected in income increase from 1.77 billion USD in Q3, which is for 32% more than in Q2. He said that they are especially proud on successes of graphics department that took 40% of the discrete graphics market in Q3. He was extremely satisfied with launching of Shanghai one quarter earlier than it was at first planned and with successful continuing of philosophy: performance per watt per dollar.
According to Ghilardi, Shanghai CPU brings certain technology novelties like additional 300 million transistors and 45nm manufacturing process that enabled for them frequency increase for 400MHz. Despite the frequency increase and large number of transistors, 45nm manufacturing process made possible for them to achieve average thermal dissipation (TPB - Thermal Power Band) at 75W. It is very interesting information that in Q1 next year will be launched, now already traditional variation – HE (High efficiency) with 55W thermal power band. We should also mention that terms TPB and TDP (Thermal Design Power) are not the same. Simply put, TPB is average thermal dissipation and TDP is max. thermal dissipation. Performance increase from 35%, comparing to solutions present at the moment, is huge success which was also recognized from most AMD partners. Since new CPU is compatible with SocketF, in start were available 25 platforms that could improve performances by simply adding this new processor into existing motherboards (SocketF). With this move, power consumption in idle mode will be reduced for about 35% while under full load for about 21%. According to Guiseppe Amato, beside more sophisticated manufacturing process, reasons for this drastic consumption increase should be looked in architecture improvements. Enlarged amount of L3 cache (8MB) and option of copying L1 and L2 cache contents into it is one of the architecture improvements. Changes in architecture enabled faster passing into “sleep” mode of CPU cores and “deeper sleep” (lower power consumption of cores in sleep mode). Even though processors still relay on OS in determining how “deep” sleep will be, it’s important to mention new driver for Shanghai processors that better cooperates with OS in this matter. It takes 1ms for waking up the core, which isn’t that fast but this “sluggishness” is compensated with significantly higher power savings.
Leslie Sobon, Vice President Worldwide Product Marketing, briefly spoke about transition from Growth to Savings economy. Regarding on global trends in sense of recession and worldwide financial crisis, reducing the pollution and increased efficiency, because of raising energy price, AMD considered that in following period market will be more oriented on savings than on race for achieving high speeds. During last year current consumption used for working and cooling servers had already attained half of system price, making this a very important item in all business calculations. Since AMD offered possibility of using new processors even on older platforms with updated BIOS, business partners can improve ratio performance per watt per dollar with minimal investments in new hardware.
After presentations have finished, Damian Schmidt, CEO of STRATO Company, briefly spoke to us. He invited us to visit their (secret) location where we participated in demonstration and plugging in of new server rack. There we had a chance, at first hand, to compare performances of new server against old one. At first glance it seems that claims reveled from AMD are truthful, but for more information you’ll have to wait until we get whole system on test. Regarding on huge number internet locations that Strato hosts, it is obvious that they needed extremely demanding hardware for this task and AMD was very pleased because they had opportunity to demonstrate performances of new processors on real example.
Ghilardi said that he was impressed by Dirk Meyer, new CEO and president of company, with his exceptional vision for further AMD development. The second reason was AMD’s success that regardless on pressure created from competitors, big financial crisis and other circumstances, reflected in income increase from 1.77 billion USD in Q3, which is for 32% more than in Q2. He said that they are especially proud on successes of graphics department that took 40% of the discrete graphics market in Q3. He was extremely satisfied with launching of Shanghai one quarter earlier than it was at first planned and with successful continuing of philosophy: performance per watt per dollar.
According to Ghilardi, Shanghai CPU brings certain technology novelties like additional 300 million transistors and 45nm manufacturing process that enabled for them frequency increase for 400MHz. Despite the frequency increase and large number of transistors, 45nm manufacturing process made possible for them to achieve average thermal dissipation (TPB - Thermal Power Band) at 75W. It is very interesting information that in Q1 next year will be launched, now already traditional variation – HE (High efficiency) with 55W thermal power band. We should also mention that terms TPB and TDP (Thermal Design Power) are not the same. Simply put, TPB is average thermal dissipation and TDP is max. thermal dissipation. Performance increase from 35%, comparing to solutions present at the moment, is huge success which was also recognized from most AMD partners. Since new CPU is compatible with SocketF, in start were available 25 platforms that could improve performances by simply adding this new processor into existing motherboards (SocketF). With this move, power consumption in idle mode will be reduced for about 35% while under full load for about 21%. According to Guiseppe Amato, beside more sophisticated manufacturing process, reasons for this drastic consumption increase should be looked in architecture improvements. Enlarged amount of L3 cache (8MB) and option of copying L1 and L2 cache contents into it is one of the architecture improvements. Changes in architecture enabled faster passing into “sleep” mode of CPU cores and “deeper sleep” (lower power consumption of cores in sleep mode). Even though processors still relay on OS in determining how “deep” sleep will be, it’s important to mention new driver for Shanghai processors that better cooperates with OS in this matter. It takes 1ms for waking up the core, which isn’t that fast but this “sluggishness” is compensated with significantly higher power savings.
Leslie Sobon, Vice President Worldwide Product Marketing, briefly spoke about transition from Growth to Savings economy. Regarding on global trends in sense of recession and worldwide financial crisis, reducing the pollution and increased efficiency, because of raising energy price, AMD considered that in following period market will be more oriented on savings than on race for achieving high speeds. During last year current consumption used for working and cooling servers had already attained half of system price, making this a very important item in all business calculations. Since AMD offered possibility of using new processors even on older platforms with updated BIOS, business partners can improve ratio performance per watt per dollar with minimal investments in new hardware.
After presentations have finished, Damian Schmidt, CEO of STRATO Company, briefly spoke to us. He invited us to visit their (secret) location where we participated in demonstration and plugging in of new server rack. There we had a chance, at first hand, to compare performances of new server against old one. At first glance it seems that claims reveled from AMD are truthful, but for more information you’ll have to wait until we get whole system on test. Regarding on huge number internet locations that Strato hosts, it is obvious that they needed extremely demanding hardware for this task and AMD was very pleased because they had opportunity to demonstrate performances of new processors on real example.
AMD Phenom II X4 940: Compared to Phenom X4 9950 BE and Intel Core 2 Q9550
In last few years it looked like AMD lost its bearings in CPU market. No one expected that company, that is in constant transition, financial problems and without complete product portfolio, can produce interesting, equivalent and concurrent product. Phenom CPUs with K10 micro architecture didn’t claim speed crown from Intel, but they were significant step forward compared to “worn out” K8 cores.
We must be aware of the fact that every processor is result of compromise between investments and goals. Probably K10 would be better product if it had larger L2 and L3 cache memory, with better and larger Branch Target Buffers, with different cache policies that are favoring intensive operations with cache memory, but in that case Phenom, that was manufactured in 65nm process, wouldn’t give adequate results. We must remind you that Intel only recently, with 45nm manufacturing process that is fully operational, introduced monolith quad-core design. Unfortunately for Intel, and on the other side, fortunately for AMD, Nehalem is totally different price group (at this moment) and thanks to this fact AMD has maneuvering space in mainstream CPU market. Thanks to new 45nm manufacturing process, developed in collaboration with “big blue” (IBM), new cores with code name “Deneb” are much cooler than older 65nm cores, can achieve much higher working frequencies and can be overclocked much better and easier. Last year AMD introduced very solid Spider platform based on AMD 790FX chipset, ATI Radeon HD 3870 graphics cards and 65nm Phenom X4 CPUs. New platform, called “Dragon”, very similar to Spider platform, is based on 790GX chipset and Radeon HD 4870 graphics cards along with new 45nm Phenom II X4 CPUs. Deneb core is compatible with older 790FX chipset, so with update of BIOS new CPUs can be used on older motherboards. This allowed us to test new Phenom II CPU on DFI LanParty 790FX motherboard with 4GB OCZ DDR2 1066 memory. All tests were done on 790FX motherboard because there were very small differences in performances compared to test results on 790GX motherboard. When Phenom II was operating at default clock we used Thermalright True cooler without fan and it worked stable. When we overclocked it, two fans (mounted in push-pull configuration) were needed.Biggest innovation of new AMD CPUs is new and very advanced 45nm lithographic manufacturing process that introduces significant advances in pFET transistor performances compared to 65nm manufacturing process. New pFET transistors have higher drive current (specific current intensity) per micrometer: 660µA/µm compared to 510µA/µm for 65nm transistors. This allows higher performances: faster “switching speed” (speed of change from logical zero to one). Higher specific current intensity is one of the Intel’s “secrets” in achieving higher CPU frequencies because their HKMG (High-K Metal Gates) are one of the fastest dielectrics on the market when it comes to “switching speeds”. On the other side, on the same frequency AMD’s 45nm CPU has lower heat dissipation compared to Intel’s 45nm CPU. AMD claims that Shanghai (45nm Opteron) has for 15% lower power consumption under full load and 30% in idle compared to Intel 45nm Xeon CPU. This can make huge difference for servers but what it brings to usual desktop user?
One of goals that AMD was hoping to achieve was smaller dimensions of chip with increase in performances, because smaller dimensions of chip allow more L3 cache to be implemented in CPU. Thanks to 45nm manufacturing process, minimal size of gate on Shanghai and Deneb cores is 38nm: 7% smaller then on 65nm CPUs. Performances are for 19% higher for nFET and 23% higher for pFET compared to 65nm process. These results are achieved not solely thanks to better manufacturing process but also thanks to advanced optimizations. One of those optimizations is SOI (Silicon On Insulator – used also on some earlier models) that resulted in lower amount of electron “leakage” and “parasite” current. Another optimization of nFET is “stress memorization” that resulted in wider n channel. Thanks to this optimization scaling to 45nm allows better mobility of electrons in “narrow” cavities that resulted in higher specific current intensity per micrometer. This increased “drive current” for 19% on nFET and for 23% on pFET transistors compared to 65nm CPUs. This is great achievement because Ion/Ioff ratio on AMD pFET transistors is about ten times lower compared to Intel’s pFET transistors. Interesting fact is that electron “leakage” is 3 times lower compared to CPUs made in older 65nm manufacturing process. Since power consumption depends on amount of “leakage”, lower electron “leakage” results in lower power consumption of entire CPU.
Next important novelty is manufacturing process itself. AMD uses so called “Immersion Lithography”. This technique improves resolution of lithographic process: it uses liquid medium instead of air that is between lithographic lens and wafer which resulted in better refractive index. Achieved resolution is proportional to refractive index of fluid. Since fluid refracts beams on wafer, resulted resolution is higher and it can achieve sizes lower then 37nm. This resulted in better CPUs, higher frequencies and great overclock possibilities.
We must be aware of the fact that every processor is result of compromise between investments and goals. Probably K10 would be better product if it had larger L2 and L3 cache memory, with better and larger Branch Target Buffers, with different cache policies that are favoring intensive operations with cache memory, but in that case Phenom, that was manufactured in 65nm process, wouldn’t give adequate results. We must remind you that Intel only recently, with 45nm manufacturing process that is fully operational, introduced monolith quad-core design. Unfortunately for Intel, and on the other side, fortunately for AMD, Nehalem is totally different price group (at this moment) and thanks to this fact AMD has maneuvering space in mainstream CPU market. Thanks to new 45nm manufacturing process, developed in collaboration with “big blue” (IBM), new cores with code name “Deneb” are much cooler than older 65nm cores, can achieve much higher working frequencies and can be overclocked much better and easier. Last year AMD introduced very solid Spider platform based on AMD 790FX chipset, ATI Radeon HD 3870 graphics cards and 65nm Phenom X4 CPUs. New platform, called “Dragon”, very similar to Spider platform, is based on 790GX chipset and Radeon HD 4870 graphics cards along with new 45nm Phenom II X4 CPUs. Deneb core is compatible with older 790FX chipset, so with update of BIOS new CPUs can be used on older motherboards. This allowed us to test new Phenom II CPU on DFI LanParty 790FX motherboard with 4GB OCZ DDR2 1066 memory. All tests were done on 790FX motherboard because there were very small differences in performances compared to test results on 790GX motherboard. When Phenom II was operating at default clock we used Thermalright True cooler without fan and it worked stable. When we overclocked it, two fans (mounted in push-pull configuration) were needed.Biggest innovation of new AMD CPUs is new and very advanced 45nm lithographic manufacturing process that introduces significant advances in pFET transistor performances compared to 65nm manufacturing process. New pFET transistors have higher drive current (specific current intensity) per micrometer: 660µA/µm compared to 510µA/µm for 65nm transistors. This allows higher performances: faster “switching speed” (speed of change from logical zero to one). Higher specific current intensity is one of the Intel’s “secrets” in achieving higher CPU frequencies because their HKMG (High-K Metal Gates) are one of the fastest dielectrics on the market when it comes to “switching speeds”. On the other side, on the same frequency AMD’s 45nm CPU has lower heat dissipation compared to Intel’s 45nm CPU. AMD claims that Shanghai (45nm Opteron) has for 15% lower power consumption under full load and 30% in idle compared to Intel 45nm Xeon CPU. This can make huge difference for servers but what it brings to usual desktop user?
One of goals that AMD was hoping to achieve was smaller dimensions of chip with increase in performances, because smaller dimensions of chip allow more L3 cache to be implemented in CPU. Thanks to 45nm manufacturing process, minimal size of gate on Shanghai and Deneb cores is 38nm: 7% smaller then on 65nm CPUs. Performances are for 19% higher for nFET and 23% higher for pFET compared to 65nm process. These results are achieved not solely thanks to better manufacturing process but also thanks to advanced optimizations. One of those optimizations is SOI (Silicon On Insulator – used also on some earlier models) that resulted in lower amount of electron “leakage” and “parasite” current. Another optimization of nFET is “stress memorization” that resulted in wider n channel. Thanks to this optimization scaling to 45nm allows better mobility of electrons in “narrow” cavities that resulted in higher specific current intensity per micrometer. This increased “drive current” for 19% on nFET and for 23% on pFET transistors compared to 65nm CPUs. This is great achievement because Ion/Ioff ratio on AMD pFET transistors is about ten times lower compared to Intel’s pFET transistors. Interesting fact is that electron “leakage” is 3 times lower compared to CPUs made in older 65nm manufacturing process. Since power consumption depends on amount of “leakage”, lower electron “leakage” results in lower power consumption of entire CPU.
Next important novelty is manufacturing process itself. AMD uses so called “Immersion Lithography”. This technique improves resolution of lithographic process: it uses liquid medium instead of air that is between lithographic lens and wafer which resulted in better refractive index. Achieved resolution is proportional to refractive index of fluid. Since fluid refracts beams on wafer, resulted resolution is higher and it can achieve sizes lower then 37nm. This resulted in better CPUs, higher frequencies and great overclock possibilities.
AMD Offers Sempron 210U and 200U Processors for Embedded Applications
AMD today announced immediate availability of the AMD Sempron 210U and 200U processors for embedded systems. These new processors are available with five year longevity that is standard for AMD embedded components and feature lidless Ball Grid Array (BGA) packaging with the low power and high performance of AMD's Direct Connect Architecture. These new lidless BGA packaged processors can help customers significantly shrink their embedded design into new smaller, flatter form factors without having to sacrifice any computing performance,” said Buddy Broeker, director, Embedded Product Marketing, AMD. “Retail touch screens, self-service kiosks and digital signage are a growing part of the consumer experience and thin client computing continues to play a strong role in helping businesses become more efficient. These new processor features will help our customers stay on the cutting edge of embedded system design.”
These AMD Sempron processors are specifically designed with unique computing features and a thin, compact form factor to help enable new and uncompromising designs. The AMD Sempron 200U and 210U processors are packaged in a smallfootprint, lidless 812-ball BGA (Ball Grid Array) package. This enables small form factor and rugged designs to be created with optimal thermal dissipation utilizing a soldered down ASB1 device. The AMD Sempron 200U and 210U processors are well suited for designs including enterprise-class thin client systems, point of sale kiosks, digital signage, ruggedized systems for military aero or other field implementations, telecommunications and networking devices, gaming machines and industrial controls with these models supporting extended longevity.
Customers can further streamline development by pairing the new AMD Sempron processors with the AMD M690E chipset.
Additionally, for those solutions that need extreme performance graphics the ATI Radeon E2400 discrete GPU can be incorporated into the platform. This discrete GPU offers support for Microsoft DirectX 10.0 and OpenGL 2.0 and like the AMD Sempron 200U and 210U processors, delivers a compact physical size to accommodate shrinking embedded system form factors.
Product features and benefits:
- Featuring AMD’s innovative Direct Connect Architecture for leading-edge performance by providing separate, dedicated high-speed links between processor and memory, processor and I/O, and I/O to memory, to enable predictability in real-time applications
- HyperTransport technology helps boost overall system performance through a dedicated high-speed, low-latency I/O interface
- AMD Digital Media Xpress technology designed to be compatible with the largest installed base of multimedia enhanced software, taking 3D and graphics to new levels for medical imaging, gaming, kiosk, and point-of-sale applications
- Simultaneous 32- and 64-bit performance, designed to be compatible with Microsoft Windows Vista
- Enhanced Virus Protection to increase the reliability of your network-connected applications
Embedded systems based on the AMD Sempron 210U and 200U processors are also available from iBASE, aValue, EVOC, Gigabyte and Inventec while additional AMD embedded customers are expected to bring systems to market in 2009.
These AMD Sempron processors are specifically designed with unique computing features and a thin, compact form factor to help enable new and uncompromising designs. The AMD Sempron 200U and 210U processors are packaged in a smallfootprint, lidless 812-ball BGA (Ball Grid Array) package. This enables small form factor and rugged designs to be created with optimal thermal dissipation utilizing a soldered down ASB1 device. The AMD Sempron 200U and 210U processors are well suited for designs including enterprise-class thin client systems, point of sale kiosks, digital signage, ruggedized systems for military aero or other field implementations, telecommunications and networking devices, gaming machines and industrial controls with these models supporting extended longevity.
Customers can further streamline development by pairing the new AMD Sempron processors with the AMD M690E chipset.
Additionally, for those solutions that need extreme performance graphics the ATI Radeon E2400 discrete GPU can be incorporated into the platform. This discrete GPU offers support for Microsoft DirectX 10.0 and OpenGL 2.0 and like the AMD Sempron 200U and 210U processors, delivers a compact physical size to accommodate shrinking embedded system form factors.
Product features and benefits:
- Featuring AMD’s innovative Direct Connect Architecture for leading-edge performance by providing separate, dedicated high-speed links between processor and memory, processor and I/O, and I/O to memory, to enable predictability in real-time applications
- HyperTransport technology helps boost overall system performance through a dedicated high-speed, low-latency I/O interface
- AMD Digital Media Xpress technology designed to be compatible with the largest installed base of multimedia enhanced software, taking 3D and graphics to new levels for medical imaging, gaming, kiosk, and point-of-sale applications
- Simultaneous 32- and 64-bit performance, designed to be compatible with Microsoft Windows Vista
- Enhanced Virus Protection to increase the reliability of your network-connected applications
Embedded systems based on the AMD Sempron 210U and 200U processors are also available from iBASE, aValue, EVOC, Gigabyte and Inventec while additional AMD embedded customers are expected to bring systems to market in 2009.
AMD Quadfather – first part of an epic trilogy?
AMD's new platform has been informally dubbed either 4x4 or Quadfather, but it is now officially called Quad FX. The 4x4 name was derived from the ability to have four processor and four graphics cores in one system. To achieve this, AMD has made a curious hybrid of desktop and workstation technologies. Like the latest Opterons, it's based on a pair of 1,207-contact LGA sockets, although the name of these has changed from Socket F to Socket L1FX. But it doesn't use NVIDIA's nForce Professional 3000-series chipset, instead calling on a variant of NVIDIA's consumer-oriented 600 series. This isn't such a new idea, as there are dual Opteron Socket 940 workstation boards on the market which use NVIDIA's nForce 4, such as MSI's K8N Master2-FAR. The 600-series chipset also means Quad FX can use regular DDR2 memory rather than registered ECC DIMMs, opening the platform up to mainstream high-performance DIMMs and overclocking.
NVIDIA's nForce 680a SLI chipset is not really that far off the nForce 680i SLI recently released for Intel processors, either. The biggest difference revolves around Intel's use of a Front Side Bus versus AMD's incorporation of an on-die memory controller. Two distinct chips are used in nForce 680i SLI. The SPP integrates a memory controller and connects to the Intel CPU via a Front Side Bus. This hosts one of the PCI Express 16x slots and the auxiliary PCI Express slots. It then connects via HyperTransport to the MCP, which hosts the second PCI Express 16x slot, the third PCI Express 16x slot which only runs at 8x electrically, and all the peripheral connections such as PCI, SATA, USB, Ethernet LAN and integrated audio. This configuration is basically the same as the traditional Northbridge/Southbridge chipsets of the last decade. The SPP takes the role of the Northbridge and the MCP is essentially a Southbridge, except that the second and third graphics slots also hang off the MCP, as we just described.
Since AMD's processors don't use a Front Side Bus, have a memory controller built in, and already communicate with the outside world via HyperTransport, they don't need a Northbridge. All they require is an MCP for a basic system without SLI. For more PCI Express lanes, a couple of chips can be daisy-chained together, which is what a number of Nvidia chipsets already do. The new nForce 680a SLI does this too, using two virtually identical MCP-like chips. Both drive two PCI Express 16x slots each, with one of each pair operating at 8x electrically (making a total of four). They join forces to provide support for four Gigabit Ethernet connections, up to 20 USB 2 ports, and 12 SATA-2 connections running at 3Gbits/sec. One chip takes care of legacy PCI slots, whilst the other handles any non-graphics PCI Express.
However, if you look closely at the architectural diagram for nForce 680a SLI you will notice that rather than running its two chips serially, as with 680i SLI, each talks directly to one of the processors via its own dedicated HyperTransport link. This is one of the reasons why AMD has based Quad FX on its Opteron platform. The AMD64 core has been built with three HyperTransport links since launch, but in a consumer single-processor configuration only one is used. The trio of links only come into their own with a multi-processor Opteron configuration.
Aside from two of those links allowing the two 680a chips (and their attached graphics cards) to be addressed independently, the third has special characteristics. It's called a coherent HyperTransport link because it can be used to transfer cache coherency information using the MOESI system which AMD operates in all its 64-bit processors. This is a very important feature, and is one of the capabilities which still gives AMD an advantage over Intel.
Since every AMD chip has its own memory controller, if you are running two of them you have two memory controllers. Like any self-respecting dual-Opteron motherboard, the Quad FX platform has two memory banks - one for each chip, as you can see above. Populate both, and you can theoretically take advantage of twice the memory bandwidth. Using the coherent HyperTransport link, one processor can access the memory controller on the other processor, and access data from its bank of DDR2. But it can also request data from the L1 and L2 caches on the other processor's two cores, which will be quicker than a call to main memory.
In contrast, Intel's Core 2 Quad incorporates two dual-core dies which can't access each-other's Level 2 cache directly. Although each dual-core die has a shared Level 2 cache pool, the two separate dies can only exchange data slowly via main memory, despite sitting next to each other.
The dual independent memory controllers and memory banks of Quad FX turn it into a Non-Uniform Memory Architecture (NUMA). For this to work properly, however, you need your operating system to be properly aware. On Windows XP SP2, a /PAE switch is required in the boot.ini for support. What this does is prevent an application from loading data into the memory bank of one processor, then try to access it from the other, which is clearly not going to give you the best performance. However, Windows XP is still considered a bit of a hack where NUMA is concerned, and only Windows Vista has full native support.
This leaves a slight question mark over AMD's strategy, thanks to Microsoft's product positioning of Vista. The benefits for NUMA of running Vista won't be such an issue for enthusiasts, who will probably be leaping onto the new OS as early adopters anyway. More worrying is the fact that Windows Vista Home Premium only supports a single physical processor, so you will have to go for the much more expensive Ultra version. In contrast, Intel's Core 2 Quad will be fine with Premium, because its four cores still count as one physical processor for Microsoft's licensing scheme.
However, other cost worries about Quad FX seem unfounded. AMD has told us that the platform will not be more expensive than Intel's quad-core. Since both nForce 680i SLI and 680a SLI use similar chipsets, there won't be much difference in motherboard pricing. More importantly, AMD explained that the new Quad FX processors will be retailing in packages of matched pairs, the most expensive of which (the FX-74) will correspond in price to Intel's QX6700. So for around the same money you could either buy one Intel processor with four cores, or get two AMD ones with a pair of cores each.
In other words, it's all down to the performance. Here, Quad FX should at least put in a better showing than the top Socket AM2 part, the Athlon FX-62. The latter runs at 2.8GHz, but the flagship Quad FX Athlon FX-74 pushes the clocks to 3GHz, which should give it a little more of an edge. There are cheaper FX-70 and FX-72 variants which run at 2.6GHz and 2.8GHz respectively. But all are still 90nm parts, not the 65nm versions which are now starting to trickle out.
NVIDIA's nForce 680a SLI chipset is not really that far off the nForce 680i SLI recently released for Intel processors, either. The biggest difference revolves around Intel's use of a Front Side Bus versus AMD's incorporation of an on-die memory controller. Two distinct chips are used in nForce 680i SLI. The SPP integrates a memory controller and connects to the Intel CPU via a Front Side Bus. This hosts one of the PCI Express 16x slots and the auxiliary PCI Express slots. It then connects via HyperTransport to the MCP, which hosts the second PCI Express 16x slot, the third PCI Express 16x slot which only runs at 8x electrically, and all the peripheral connections such as PCI, SATA, USB, Ethernet LAN and integrated audio. This configuration is basically the same as the traditional Northbridge/Southbridge chipsets of the last decade. The SPP takes the role of the Northbridge and the MCP is essentially a Southbridge, except that the second and third graphics slots also hang off the MCP, as we just described.
Since AMD's processors don't use a Front Side Bus, have a memory controller built in, and already communicate with the outside world via HyperTransport, they don't need a Northbridge. All they require is an MCP for a basic system without SLI. For more PCI Express lanes, a couple of chips can be daisy-chained together, which is what a number of Nvidia chipsets already do. The new nForce 680a SLI does this too, using two virtually identical MCP-like chips. Both drive two PCI Express 16x slots each, with one of each pair operating at 8x electrically (making a total of four). They join forces to provide support for four Gigabit Ethernet connections, up to 20 USB 2 ports, and 12 SATA-2 connections running at 3Gbits/sec. One chip takes care of legacy PCI slots, whilst the other handles any non-graphics PCI Express.
However, if you look closely at the architectural diagram for nForce 680a SLI you will notice that rather than running its two chips serially, as with 680i SLI, each talks directly to one of the processors via its own dedicated HyperTransport link. This is one of the reasons why AMD has based Quad FX on its Opteron platform. The AMD64 core has been built with three HyperTransport links since launch, but in a consumer single-processor configuration only one is used. The trio of links only come into their own with a multi-processor Opteron configuration.
Aside from two of those links allowing the two 680a chips (and their attached graphics cards) to be addressed independently, the third has special characteristics. It's called a coherent HyperTransport link because it can be used to transfer cache coherency information using the MOESI system which AMD operates in all its 64-bit processors. This is a very important feature, and is one of the capabilities which still gives AMD an advantage over Intel.
Since every AMD chip has its own memory controller, if you are running two of them you have two memory controllers. Like any self-respecting dual-Opteron motherboard, the Quad FX platform has two memory banks - one for each chip, as you can see above. Populate both, and you can theoretically take advantage of twice the memory bandwidth. Using the coherent HyperTransport link, one processor can access the memory controller on the other processor, and access data from its bank of DDR2. But it can also request data from the L1 and L2 caches on the other processor's two cores, which will be quicker than a call to main memory.
In contrast, Intel's Core 2 Quad incorporates two dual-core dies which can't access each-other's Level 2 cache directly. Although each dual-core die has a shared Level 2 cache pool, the two separate dies can only exchange data slowly via main memory, despite sitting next to each other.
The dual independent memory controllers and memory banks of Quad FX turn it into a Non-Uniform Memory Architecture (NUMA). For this to work properly, however, you need your operating system to be properly aware. On Windows XP SP2, a /PAE switch is required in the boot.ini for support. What this does is prevent an application from loading data into the memory bank of one processor, then try to access it from the other, which is clearly not going to give you the best performance. However, Windows XP is still considered a bit of a hack where NUMA is concerned, and only Windows Vista has full native support.
This leaves a slight question mark over AMD's strategy, thanks to Microsoft's product positioning of Vista. The benefits for NUMA of running Vista won't be such an issue for enthusiasts, who will probably be leaping onto the new OS as early adopters anyway. More worrying is the fact that Windows Vista Home Premium only supports a single physical processor, so you will have to go for the much more expensive Ultra version. In contrast, Intel's Core 2 Quad will be fine with Premium, because its four cores still count as one physical processor for Microsoft's licensing scheme.
However, other cost worries about Quad FX seem unfounded. AMD has told us that the platform will not be more expensive than Intel's quad-core. Since both nForce 680i SLI and 680a SLI use similar chipsets, there won't be much difference in motherboard pricing. More importantly, AMD explained that the new Quad FX processors will be retailing in packages of matched pairs, the most expensive of which (the FX-74) will correspond in price to Intel's QX6700. So for around the same money you could either buy one Intel processor with four cores, or get two AMD ones with a pair of cores each.
In other words, it's all down to the performance. Here, Quad FX should at least put in a better showing than the top Socket AM2 part, the Athlon FX-62. The latter runs at 2.8GHz, but the flagship Quad FX Athlon FX-74 pushes the clocks to 3GHz, which should give it a little more of an edge. There are cheaper FX-70 and FX-72 variants which run at 2.6GHz and 2.8GHz respectively. But all are still 90nm parts, not the 65nm versions which are now starting to trickle out.
AMD gets more aggressive, announces next-gen CPU cores
If there is change going on at AMD that is visible to the outside world then it clearly is a decisively more open communication about future products. We have long waited for this change, which began to show up first at the firm’s CTO Summit a few weeks ago and today continued at the company’s financial analyst meeting. The company released code-names as well as early product features that provide an outlook from today to 2009 and early 2010. So let’s have a look.Mobile and ultra-mobile
This segment is targeted by AMD with processors that will consume between 1 and 10 watts of power in the ultra-mobile segment as well as slightly more for notebooks. We already knew that 2007 will see a refresh of the Kite platform, which consists of Turion X2 processors with virtualization support, the M960 (SB600) chipset and ATI Radeon X1200 series graphics (or dedicated graphics processors) such as the Mobility Radeon HD. In 2008, AMD will release the Puma platform with the Griffin dual-core CPU, in combination with the RS780M chipset and the SB700 southbridge, which will support DX10 in integrated graphics. There will also be a new 55 nm discrete mobile GPU, codenamed M8X.
We have been waiting for more details on Fusion and the first offspring of this integrated CPU/GPU approach will be Falcon, a quad-core mobile processor based on the "Bulldozer" core with DX10 or DX11 capability.
In the ultra-mobile segment, AMD will offer the Bobcat processor, which will be offered against Intel’s Silverthorne CPU. Bobcat will aim to attract customers in the 1-10 watt segment, which includes products such as ultra mobile PCs or mobile Internet devices. AMD did not say when this CPU will be available.
Mainstream desktop 2007, 2008
The dual-core mainstream and entry-level segment will see the Pinwheel platform in 2007 (Athlon X2 CPU, Hypertransport 3, 690 chipset). The processor will carryover to 2008 into the Cartwheel platform with a 45 nm Athlon X2 in the second half of the year. The chip will integrate 1 MB L3 cache, run in combination with the RS700 chipset, support PCIe2 and is likely to be offered in PC systems with a 55 nm “DX10+” graphics processor.
Performance desktop 2007, 2008
The Barcelona-based Phenom processor will dominate AMD’s roadmap for performance desktops in 65 nm versions in 2007 and in a 45 nm version in 2008. The 2007 Spider platform will include dual-cores (X2) and quad-cores (X4) with 2 MB L3 cache, while the 2008 Hardcastle platform will see an upgrade of the processor to 45 nm and 6 MB L3 cache.
Concerns about the scalability of the upcoming Phenom processor were addressed by the company with a 3.0 GHz Phenom X4 CPU.Server and desktop beyond 2008
Following Barcelona, AMD will release the Bulldozer core in 2009. The first server processor to be released with this core is code-named Sandtiger and will debut with 8 – 16 cores, according to AMD. This next-generation platform will include DirectConnect2, chipsets developed by AMD, HT3, four HT links, DD3 memory support, the recently announced G3 memory extender, as well as support for DDR3 memory modules.
Bulldozer will also make its way into mainstream and performance desktops in 2009.
For mainstream desktop systems, Copperhead will bring a “native quad-core” 32 nm processor in an AM3 package. This is particularly interesting, as AMD intends to release 32 nm CPUs only one year after the release of 45 nm chips to catch up with Intel, which also plans to release 32 nm processors in 2009. The Copperhead platform will be AMD’s first desktop platform to be compatible with DDR3 memory.
Also based on the Bulldozer core will be the Python platform, which will include a 32 nm Fusion-based processor in an AM3 package to replace Phenom.
This segment is targeted by AMD with processors that will consume between 1 and 10 watts of power in the ultra-mobile segment as well as slightly more for notebooks. We already knew that 2007 will see a refresh of the Kite platform, which consists of Turion X2 processors with virtualization support, the M960 (SB600) chipset and ATI Radeon X1200 series graphics (or dedicated graphics processors) such as the Mobility Radeon HD. In 2008, AMD will release the Puma platform with the Griffin dual-core CPU, in combination with the RS780M chipset and the SB700 southbridge, which will support DX10 in integrated graphics. There will also be a new 55 nm discrete mobile GPU, codenamed M8X.
We have been waiting for more details on Fusion and the first offspring of this integrated CPU/GPU approach will be Falcon, a quad-core mobile processor based on the "Bulldozer" core with DX10 or DX11 capability.
In the ultra-mobile segment, AMD will offer the Bobcat processor, which will be offered against Intel’s Silverthorne CPU. Bobcat will aim to attract customers in the 1-10 watt segment, which includes products such as ultra mobile PCs or mobile Internet devices. AMD did not say when this CPU will be available.
Mainstream desktop 2007, 2008
The dual-core mainstream and entry-level segment will see the Pinwheel platform in 2007 (Athlon X2 CPU, Hypertransport 3, 690 chipset). The processor will carryover to 2008 into the Cartwheel platform with a 45 nm Athlon X2 in the second half of the year. The chip will integrate 1 MB L3 cache, run in combination with the RS700 chipset, support PCIe2 and is likely to be offered in PC systems with a 55 nm “DX10+” graphics processor.
Performance desktop 2007, 2008
The Barcelona-based Phenom processor will dominate AMD’s roadmap for performance desktops in 65 nm versions in 2007 and in a 45 nm version in 2008. The 2007 Spider platform will include dual-cores (X2) and quad-cores (X4) with 2 MB L3 cache, while the 2008 Hardcastle platform will see an upgrade of the processor to 45 nm and 6 MB L3 cache.
Concerns about the scalability of the upcoming Phenom processor were addressed by the company with a 3.0 GHz Phenom X4 CPU.Server and desktop beyond 2008
Following Barcelona, AMD will release the Bulldozer core in 2009. The first server processor to be released with this core is code-named Sandtiger and will debut with 8 – 16 cores, according to AMD. This next-generation platform will include DirectConnect2, chipsets developed by AMD, HT3, four HT links, DD3 memory support, the recently announced G3 memory extender, as well as support for DDR3 memory modules.
Bulldozer will also make its way into mainstream and performance desktops in 2009.
For mainstream desktop systems, Copperhead will bring a “native quad-core” 32 nm processor in an AM3 package. This is particularly interesting, as AMD intends to release 32 nm CPUs only one year after the release of 45 nm chips to catch up with Intel, which also plans to release 32 nm processors in 2009. The Copperhead platform will be AMD’s first desktop platform to be compatible with DDR3 memory.
Also based on the Bulldozer core will be the Python platform, which will include a 32 nm Fusion-based processor in an AM3 package to replace Phenom.
AMD Demos 45nm Native Quad-Core Processors for Server, Desktop
MD (NYSE: AMD) demonstrated at the CeBit electronics exhibition its first 45nm quad-core chips running multiple operating systems and a range of processing intensive applications. The processors were produced in Dresden, Germany, in AMD’s Fab 36 300mm manufacturing facility, using an advanced 45nm process co-developed with IBM.
AMD 45nm transistors are engineered to enable greater performance-per-watt capabilities in AMD processors and platforms. AMD combines new processes and materials with leading edge technologies, such as immersion lithography and AMD’s fourth-generation strained silicon, for a highly-manufacturable, highly-efficient production process.
This important milestone is the first of many as AMD moves toward delivery of 45nm products later this year. The first 45nm chips demonstrated by AMD include the “Shanghai” product for server and “Deneb” for desktop platforms.
AMD 45nm transistors are engineered to enable greater performance-per-watt capabilities in AMD processors and platforms. AMD combines new processes and materials with leading edge technologies, such as immersion lithography and AMD’s fourth-generation strained silicon, for a highly-manufacturable, highly-efficient production process.
This important milestone is the first of many as AMD moves toward delivery of 45nm products later this year. The first 45nm chips demonstrated by AMD include the “Shanghai” product for server and “Deneb” for desktop platforms.
AMD Duron 1.2GHz vs. Intel Celeron 1.2GHz
In terms of bleeding-edge performance, the Duron vs. Celeron bout has never been in the heavyweight class. The Don King electric PR frizz is AWOL, creating lower-profile PR sparring. But this fight is no less important in terms of market share, that's for sure. The value business sector sales far outweigh our beloved enthusiast performance market, and that's also where the real money is made. With the release of Athlon XP, it's easy to forget that AMD also caters to the value (both business and consumer) sector with Duron steppings. The latest is the Duron 1.2GHz, upping the previous Duron 1.1GHz by 100MHz.
The players. So in the green corner is AMD's new challenger, weighing in at 0.18-micron, sporting the Morgan core, with 64KB of L2 cache trim and green pants (I made that bit up), the 1.2GHz Duron. Running at 1.75 volts, the Duron 1.2GHz is socket 462-based with a 12x multiplier and 100MHz frontside bus, which has support for PC-133 SDRAM or 2100 DDR RAM platforms. In the blue corner, weighing in at 0.13-micron, sporting the Tualatin core, with 256KB of L2 cache, is Intel's Celeron 1.2GHz (out for a while now).
Morgan is a stripped-down version of the Athlon XP Palomino core. The major difference is in the size of the L2 cache. Morgan comes with 64KB as opposed to the 256KB on Palomino. Unlike the Athlon XP, the Morgan-based Duron is still packaged with an older CPGA (Ceramic Pin Grid Array) ceramic case. The Morgan core improves upon the previous Spitfire core (used on sub-1GHz Durons), lowering power consumption by 20% and adding data prefetch and 3DNow! Professional instructions, making the new Duron SSE-compliant.
Performance. Benchmarking both contestants under WinXP Pro, with 256MB and a GeForce3, the winner was crystal clear. The Duron 1.2GHz sweeps in Quake III, SYSmark2001 (a close call), and 3DMark2001. The winning scores don't touch an Athlon XP 1900+ or Pentium 4 2GHz, but obviously these budget CPUs, aimed at the corporate space, are not supposed to. Nevertheless, it's reassuring that you can play games respectably on a Duron 1.2GHz.
Overclockers should be content with either CPU. By joining the L1 bridges via a pencil, the Duron 1.2GHz bounced up to 1.35GHz, an extra 135MHz. By increasing the voltage and utilizing more eccentric cooling techniques, it could likely go higher. The Intel Celeron 1.2GHz actually becomes more attractive than its Duron counterpart with overclocking in mind. The Celeron's 100MHz FSB needs to be upped to 125MHz, which really allows the Celeron to party at around the 1.5GHz mark (thanks to its 0.13-micron core). This implies that, should Intel ship 133MHz FSB Celerons, they would be far better equipped to duke it out with Durons. s it the name? Despite the Duron outperforming the Celeron, the corporate space still seems to pass over AMD's chip, whether because of historical stability concerns, lack of brand recognition, misinformation, or perhaps just benign neglect. In the OEM sector, where most Celerons are sold, performance doesn't seem to be the priority. The Intel name seems to be what matters, so clearly AMD has its corporate back against a wall, albeit with a better performing CPU and at a more attractive price. Intel's marketing department wins again as the company does have a slightly wider spectrum differential between their high-end Pentium 4 part and their lower-end Celeron part. The difference in the performance delta between the Duron and Athlon is much less, as is the difference in price.
Final word. If you read this magazine, chances are you're an enthusiast (just like yours truly), and hence the processor is not a place to be cost-cutting. Still, the performance lead in the value sector goes to AMD's Duron 1.2GHz, and although neither processor particularly floats my boat, I find the Celeron 1.2GHz less buoyant. As an enthusiast, you are going to need a bigger boat (hence the relatively mediocre ratings), and I urge you to find a way to spend the extra cash on a pricier Athlon XP for its superior performance. For those who can't, choosing the Duron 1.2GHz over Celeron 1.2GHz is a better bang for the buck, as it's cheaper. Better still: Sit on the fence until AMD's die-shrunk 0.13-micron Appaloosa core is released in mid/late Q1 2002, according to AMD's latest roadmap, rumored to be entering at the 1.5GHz level. I might sail in one of those. . . .
The players. So in the green corner is AMD's new challenger, weighing in at 0.18-micron, sporting the Morgan core, with 64KB of L2 cache trim and green pants (I made that bit up), the 1.2GHz Duron. Running at 1.75 volts, the Duron 1.2GHz is socket 462-based with a 12x multiplier and 100MHz frontside bus, which has support for PC-133 SDRAM or 2100 DDR RAM platforms. In the blue corner, weighing in at 0.13-micron, sporting the Tualatin core, with 256KB of L2 cache, is Intel's Celeron 1.2GHz (out for a while now).
Morgan is a stripped-down version of the Athlon XP Palomino core. The major difference is in the size of the L2 cache. Morgan comes with 64KB as opposed to the 256KB on Palomino. Unlike the Athlon XP, the Morgan-based Duron is still packaged with an older CPGA (Ceramic Pin Grid Array) ceramic case. The Morgan core improves upon the previous Spitfire core (used on sub-1GHz Durons), lowering power consumption by 20% and adding data prefetch and 3DNow! Professional instructions, making the new Duron SSE-compliant.
Performance. Benchmarking both contestants under WinXP Pro, with 256MB and a GeForce3, the winner was crystal clear. The Duron 1.2GHz sweeps in Quake III, SYSmark2001 (a close call), and 3DMark2001. The winning scores don't touch an Athlon XP 1900+ or Pentium 4 2GHz, but obviously these budget CPUs, aimed at the corporate space, are not supposed to. Nevertheless, it's reassuring that you can play games respectably on a Duron 1.2GHz.
Overclockers should be content with either CPU. By joining the L1 bridges via a pencil, the Duron 1.2GHz bounced up to 1.35GHz, an extra 135MHz. By increasing the voltage and utilizing more eccentric cooling techniques, it could likely go higher. The Intel Celeron 1.2GHz actually becomes more attractive than its Duron counterpart with overclocking in mind. The Celeron's 100MHz FSB needs to be upped to 125MHz, which really allows the Celeron to party at around the 1.5GHz mark (thanks to its 0.13-micron core). This implies that, should Intel ship 133MHz FSB Celerons, they would be far better equipped to duke it out with Durons. s it the name? Despite the Duron outperforming the Celeron, the corporate space still seems to pass over AMD's chip, whether because of historical stability concerns, lack of brand recognition, misinformation, or perhaps just benign neglect. In the OEM sector, where most Celerons are sold, performance doesn't seem to be the priority. The Intel name seems to be what matters, so clearly AMD has its corporate back against a wall, albeit with a better performing CPU and at a more attractive price. Intel's marketing department wins again as the company does have a slightly wider spectrum differential between their high-end Pentium 4 part and their lower-end Celeron part. The difference in the performance delta between the Duron and Athlon is much less, as is the difference in price.
Final word. If you read this magazine, chances are you're an enthusiast (just like yours truly), and hence the processor is not a place to be cost-cutting. Still, the performance lead in the value sector goes to AMD's Duron 1.2GHz, and although neither processor particularly floats my boat, I find the Celeron 1.2GHz less buoyant. As an enthusiast, you are going to need a bigger boat (hence the relatively mediocre ratings), and I urge you to find a way to spend the extra cash on a pricier Athlon XP for its superior performance. For those who can't, choosing the Duron 1.2GHz over Celeron 1.2GHz is a better bang for the buck, as it's cheaper. Better still: Sit on the fence until AMD's die-shrunk 0.13-micron Appaloosa core is released in mid/late Q1 2002, according to AMD's latest roadmap, rumored to be entering at the 1.5GHz level. I might sail in one of those. . . .
Subscribe to:
Comments (Atom)
