Quantcast

31 August, 2010

Intel's Sandy Bridge CPUs Will Play Blu-Ray 3D


Share
Right now on Westmere chips are integrated graphics along with the CPU sitting on the same silicon, but Sandy Bridge will move them into a single die. This will boost performance in itself, and Intel boasts that the new IGP will be able to decode Blu-ray 3D media. Intel's hasn't yet revealed everything on Sandy Bridge yet, but the company is likely to discuss the upgraded graphics part at the Intel Developer Forum on September 13 to 15. The current graphics part in Westmere chips can already do 1080p Blu-ray decodes, but as we know from the consumer electronics space, newer and stronger hardware is required to render the additional images for 3D.
Intel's best attempt at IGPs yet.

Right now on Westmere chips are integrated graphics along with the CPU sitting on the same silicon, but Sandy Bridge will move them into a single die. This will boost performance in itself, and Intel boasts that the new IGP will be able to decode Blu-ray 3D media.

Intel's hasn't yet revealed everything on Sandy Bridge yet, but the company is likely to discuss the upgraded graphics part at the Intel Developer Forum on September 13 to 15.

The current graphics part in Westmere chips can already do 1080p Blu-ray decodes, but as we know from the consumer electronics space, newer and stronger hardware is required to render the additional images for 3D.

Intel spokesman Nick Knupffer confirmed that Sandy Bridge mobile chips will have the capability for Blu-ray 3D, which hopefully translates into greater 3D graphics performance too.

Source: tomshardware.com

New Core Parts And Pentiums On The Way


Share
Intel has slashed the price of its high-end Core i7 950 processor from $562 to $294 and it has also added several new processors to its pricelist. The most interesting new part is the Core i3 560. It’s clocked at 3.33GHz and it costs the same as a 550, which runs at 3.2GHz. In other words, if you were considering buying a Core i3, you should wait for the new one to show up in stores.
Intel Core i7 950 drops to $294

Intel has slashed the price of its high-end Core i7 950 processor from $562 to $294 and it has also added several new processors to its pricelist.

The most interesting new part is the Core i3 560. It’s clocked at 3.33GHz and it costs the same as a 550, which runs at 3.2GHz. In other words, if you were considering buying a Core i3, you should wait for the new one to show up in stores.

The low end got a slight refresh with two new Wolfdale-based Pentium parts. The E6800 is clocked at 3.3GHz and it’s priced at $86. The 3.0GHz E5700 has also entered the fray and it costs $75, while 2.7GHz E3500 costs $57.

There are no changes in the mobile segment. The only new CPU is the dual-core Atom N550 and it is priced at $86, which really sounds quite steep considering the old Atom 330 dual-core costs just $43.

Source: fudzilla.com

Intel Talks A Little About 10-Core Westmere-EX


Share
At Hot Chips, Intel engineer Dheemanth Nagaraj told everyone in attendance that the next server processor from the company would have 10 cores and be able to process 20 threads. It's the next in the server EX line, the Westmere-EX, which will pick up from the Nehalem-EX using the newest core architecture. What Nagaraj declined to reveal were clock speeds or anything related to performance, other than it'll have two more cores and four more thread capability than Nehalem-EX.
That'd be decacore.

At Hot Chips, Intel engineer Dheemanth Nagaraj told everyone in attendance that the next server processor from the company would have 10 cores and be able to process 20 threads.

It's the next in the server EX line, the Westmere-EX, which will pick up from the Nehalem-EX using the newest core architecture. What Nagaraj declined to reveal were clock speeds or anything related to performance, other than it'll have two more cores and four more thread capability than Nehalem-EX.

According to the Register, Westmere-EX will have an L3 cache – something Intel prefers to call "last level cache" – and each of its 10 cores will share 10 "slices" of this cache, which are accessed over a bidirectional ring bus and can handle five parallel cache requests per clock cycle.

Those running Nehalem-EX systems now can maintain their current platforms, as Westmere-EX will be socket compatible with the current Boxboro-EX platform.

One feature that did not make it into Nehalem-EX due to time constraints but will be in Westmere-EX is Directory Assisted Snoopy (DAS) to improve local memory latency.

When asked why Intel is only going 10 cores when the competition has the 12-core Magny-Cours, Nagaraj said that going with 10 "gave us the sweet spot for performance and time-to-market."

Source: tomshardware.com

Desktop Bulldozer Processors Will Require New Platforms - AMD


Share
Advanced Micro Devices said that its next-generation desktop processors code-named Zambezi will use socket AM3+ platforms, which will be backwards compatible with the firm's existing AM3 products. While the latter is an advantage for the platform, it may be a disadvantage for eight-core processors based on Bulldozer micro-architecture. "The existing G34 and C32 server infrastructure will support the new Bulldozer-based server products.
AMD Zambezi to Use AM3+ Platforms

Advanced Micro Devices said that its next-generation desktop processors code-named Zambezi will use socket AM3+ platforms, which will be backwards compatible with the firm's existing AM3 products. While the latter is an advantage for the platform, it may be a disadvantage for eight-core processors based on Bulldozer micro-architecture.

"The existing G34 and C32 server infrastructure will support the new Bulldozer-based server products. In order for AMD’s desktop offering to fully leverage the capabilities of Bulldozer, an enhanced AM3+ socket will be introduced that supports Bulldozer and is backward-compatible with our existing AM3 CPU offerings," an official from AMD said in an interview with Planet3DNow web-site.

Apparently, it was possible for AMD to make Bulldozer microprocessors compatible with existing AM3 infrastructure, but in order to do that, the company would have to sacrifice certain important features of the new core.

"When we initially set out on the path to Bulldozer we were hoping for AM3 compatibility, but further along the process we realized that we had a choice to make based on some of the features that we wanted to bring with Bulldozer. We could either provide AM3 support and lose some of the capabilities of the new Bulldozer architecture or, we could choose the AM3+ socket which would allow the Bulldozer-base Zambezi to have greater performance and capability," the official said.

The compatibility with older microprocessors allows AMD to simplify transition to the new micro-architecture and process design since the new AM3+ platform will support inexpensive chips from day one. However, such compatibility also means that AMD Zambezi processors will only support dual-channel memory controller. Considering the fact that all modern high-end Intel Core i7 processors with up to six cores feature triple-channel memory controller, it is unclear how AMD plans to "feed" eight cores of Zambezi with dul-channel DDR3 without creating bottlenecks.

AMD Bulldozer-based processor code-named Zambezi will have up to eight cores along with a new TurboCore dynamic acceleration technology. Thanks to the new micro-architecture the chip promises to be faster than existing AMD products.

Source: xbitlabs.com

AMD's Next-Gen Server Platforms To Feature TurboCore Technology


Share
Advanced Micro Devices plans to add dynamic overclocking technology to its server processors code-named Interlagos that are based on the Bulldozer micro-architecture next year. This will be the first time when an AMD server chip will get ability to dynamically boost itself. However, the server TurboCore technology will naturally have some specifics and will most likely receive a new name. "There will be a TurboCore feature for Bulldozer, but there will be some improvements from what you see in “Thuban”.
AMD Enables TurboCore for Servers on Bulldozer Chips

Advanced Micro Devices plans to add dynamic overclocking technology to its server processors code-named Interlagos that are based on the Bulldozer micro-architecture next year. This will be the first time when an AMD server chip will get ability to dynamically boost itself. However, the server TurboCore technology will naturally have some specifics and will most likely receive a new name.

"There will be a TurboCore feature for Bulldozer, but there will be some improvements from what you see in “Thuban”. There are some enhancements to give it more “turbo”. This will be the first introduction of the TurboCore technology in the server processors. We expect that this will translate into a big boost in performance when using single threaded applications, and there should be some interesting capabilities for heavier workloads as well," said John Fruehe, the director of product marketing for server/workstation products at AMD.

AMD's previous-generation and current generation Opteron microprocessors support CoolCore and/or CoolSpeed technologies that optimize power consumption and performance. The former can disable parts of the chip that are not used to trim power usage, whereas the latter can allow reducing p-states of microprocessors for certain periods of time so that to keep CPU operating even when thermal limits are reached.

Many of Intel Xeon processors for dual-processor and multi-processor machines are equipped with Intel TurboBoost technology, which allows to disable some cores that are not necessary under present workload and increase clock-speeds of remaining cores.

AMD's latest desktop processors with six-cores code-named Thuban and officially known as AMD Phenom II X6 also feature TurboCore technology that dynamically disables unused cores and raises the clock-speed of the remaining, which greatly helps to speed up applications that cannot take advantage of multiple cores.

It remains to be seen which differences will bring the second-generation TurboCore technology to Bulldozer processors for desktops and servers.

Source: xbitlabs.com

AMD Drops Support Of 3DNow! Instructions


Share
Advanced Micro Devices introduced the 3DNow! instruction set back in the K6-2 days to perform single instruction multiple data (SIMD) instructions, otherwise known as vectorized instructions. The 3DNow! was meant to greatly improve performance of floating point operations, but not a lot of programs actually took advantage of the instruction sets. As a result, over a decade later AMD decided to pull the plug and its future chips will not support 3DNow!
AMD's Future Chips Will Not Support 3DNow!

Advanced Micro Devices introduced the 3DNow! instruction set back in the K6-2 days to perform single instruction multiple data (SIMD) instructions, otherwise known as vectorized instructions. The 3DNow! was meant to greatly improve performance of floating point operations, but not a lot of programs actually took advantage of the instruction sets. As a result, over a decade later AMD decided to pull the plug and its future chips will not support 3DNow!

"[Since the introduction of 3DNow!], we have added many SIMD instruction sets to our processors, such as the widely used Streaming SIMD Extensions (SSE) instruction set and its successive versions. 3DNow! instructions are being deprecated and will not be supported in certain upcoming AMD processors. In those processors, the 3DNow! Instructions feature flag bit will not be set," wrote Sharon Troia, senior developer relations engineer.

Around the same time as 3DNow! instructions were developed, programmers were accustomed to using a model of ‘try and catch’ to check if a processor supported an instruction or instruction set. This is when the application ‘tries’ to execute an instruction to see if it’s available. If the application receives an Undefined Exception (#UD) from the processor, it believes the instruction set isn’t available. These types of applications may not do well under newer virtual machines. That’s subject for another blog though.

The 3DNow! versions of the PREFETCH and PREFETCHW instructions are now in a class of their own and AMD plans to continue to support them.

It is most likely that mainstream applications have non-3DNow! code path to take, such as an SSE path.

Source: xbitlabs.com

09 August, 2010

AMD's Fusion Processors Are Designed To Be As Fast As Discrete GPUs - Company


Share
A high-ranking executive of Advanced Micro Devices said Friday that while the so-called accelerated processing units (APU), which combine x86 cores, memory controller and graphics processor on the same piece of silicon, will tangibly rise the bar for performance of built-in graphics, they will not destroy the market of discrete graphics processing units (GPUs). "AMD’s APUs are designed to deliver exceptional performance – and we believe this will be readily apparent when compared to integrated graphics products available from the competition. Some of our APUs, by sheer performance alone, are expected to be faster than some discrete GPUs on the market..
Fusion APUs Will Rise Low-Cost Graphics Performance Bar, But Will Not Kill the GPU

A high-ranking executive of Advanced Micro Devices said Friday that while the so-called accelerated processing units (APU), which combine x86 cores, memory controller and graphics processor on the same piece of silicon, will tangibly rise the bar for performance of built-in graphics, they will not destroy the market of discrete graphics processing units (GPUs).

"AMD’s APUs are designed to deliver exceptional performance – and we believe this will be readily apparent when compared to integrated graphics products available from the competition. Some of our APUs, by sheer performance alone, are expected to be faster than some discrete GPUs on the market. As our GPU cores improve, you can expect our APU graphics performance to similarly improve," said Godfrey Cheng, the director of client technology unit at AMD.

In the first half of next year AMD plans to launch the code-named Llano accelerating processing unit (APU) with up to four Phenom II-class x86 cores and with up to ATI Radeon HD 5000-class 480 stream processors. Potentially, Llano offers higher computing performance than code-named named Redwood chip (which has up to 400 SPs), which means that AMD will have to refresh entry-level lineup otherwise Llano will likely stop sales Cedar-based products.

The Redwood chip that powers ATI Radeon HD 5500-series graphics solutions has up to 400 stream processors, meanwhile, Cedar chip that is featured on ATI Radeon HD 5400-series boards only has just 80 stream processors. Underperforming Cedar, incoming Llano and inexpensive Redwood-based graphics cards clearly show that performance bar for low-end and affordable graphics either has to be increased or the low-end graphics will cease to exist.

AMD's Fusion APUs have a substantial caveat in the form of relatively low memory bandwidth. Thanks to the fact that APUs combine CPUs and GPUs on a single piece of silicon, memory and bus bandwidth will not be required for communication between the two types of cores and some other overdraws will be reduced. Still, discrete graphics processing units will continue to exist for quite a while: there is a number of APU and GPU generations in development.

"One of the main reasons people want a discrete GPU is to improve gaming performance. Gamers have remained loyal to AMD and ATI for more than a decade – a decade that has seen some pretty incredible changes in the industry. This segment shows no signs of disappearing anytime soon and we intend to continue to provide enthusiasts with leading discrete graphics solutions. In fact, we have several generations of APUs and GPUs in development right now," stressed Mr. Cheng.

Given the fact that video games for personal computers remain pretty demanding towards performance of graphics cards, it is pretty clear that there is a long life for ATI Radeon GPU ahead. Moreover, there is professional graphics segment that demands cream-of-the-creams and which is willing to pay thousands of dollars per graphics card. Nevertheless, there is a question whether the research and development of future hardware graphics technologies will be funded from dedicated budgets or from convergent CPU-GPU-APU budgets, given the fact that sales of discrete graphics chips will inevitably slowdown once proper integrated solutions emerge.

Source: xbitlabs.com

Manufacturers Plan Ontario Laptops Early Next Year


Share
Notebooks with AMD's first Fusion processors, based on the 40nm ‘Ontario' architecture are going to hit the shops early next year. While we have known that AMD will be shipping to partners before the end of this year it is not clear when we would see some products based on the chips. Digitimes has been talking to its deep throats in the manufacturing industry who told hacks that they will be shipping notebooks based on the new APUs in the first quarter of next year. TSMC, which is helping AMD make the 40nm chips, will ramp-up to full-scale production of the new silicon in early December.
Fusion getting early outing

Notebooks with AMD's first Fusion processors, based on the 40nm ‘Ontario' architecture are going to hit the shops early next year. While we have known that AMD will be shipping to partners before the end of this year it is not clear when we would see some products based on the chips.

Digitimes has been talking to its deep throats in the manufacturing industry who told hacks that they will be shipping notebooks based on the new APUs in the first quarter of next year. TSMC, which is helping AMD make the 40nm chips, will ramp-up to full-scale production of the new silicon in early December. Acer, ASUS and HP apparently should have plenty of time to start manufacturing systems for a launch in the new year.

Although Digitimes used its usual “unnamed sources” for the news, it does line up with last month's leaked AMD roadmap. In that we saw the the Zacate CPU/GPU combo as launching at the tail-end of this year or the start of next year.

It also supports the idea that the APU will be based on the Ontario architecture which is targeted at low-power and ultraportable systems.

More here: fudzilla.com

02 August, 2010

3.2GHz Remains The Fastest AMD In 2010


Share
According to AMD’s most recent plans, it seems that its 3.2GHz six-core remains the fastest AMD CPU in 2010. Its name is Phenom II X6 1090T and this 3.2GHz 125W TDP CPU sells for roughly €250 in the EU. There are no plans to top this speed in the remaining part of 2010 but compared to Intel’s six core offering, this looks like a very affordable product. The good part of this CPU is that with the help of Turbo it overclocks all the way to 3.6GHz, on auto pilot.
From six cores

According to AMD’s most recent plans, it seems that its 3.2GHz six-core remains the fastest AMD CPU in 2010.

Its name is Phenom II X6 1090T and this 3.2GHz 125W TDP CPU sells for roughly €250 in the EU. There are no plans to top this speed in the remaining part of 2010 but compared to Intel’s six core offering, this looks like a very affordable product. The good part of this CPU is that with the help of Turbo it overclocks all the way to 3.6GHz, on auto pilot.

AMD still offers unlocked Black Edition CPUs which means that you should be able to overclock them well over its official clock and many enthusiasts with no so deep pocket will want to go after this six core. The eternal problem how to put all six cores to good use. This will obviously remain the biggest issue for the CPU industry as a whole, especially knowing that eight- and twelve-core CPUs are around the corner.

More is not necessarily better, as you can't easily utilize all the cores. There is still one more six-core to launch. The Phenom II X6 1075T should come out this quarter and with its 3.0GHz clock it will pack a 125W TDP.

Source: fudzilla.com

Early AMD Llano And Ontario Performance Numbers Tip Up


Share
WE'RE STILL AT least a couple of months away from AMD's Ontario announcement and even longer from Llano, but some very early performance figures have tipped up online courtesy of, well, distributed computing. To be more precise details of various distributed computing apps being tested on AMD Llano and Ontario systems have appeared on BOINC or the Berkeley Open Infrastructure for Network Computing if you prefer.
Early AMD Llano and Ontario performance numbers tip up

WE'RE STILL AT least a couple of months away from AMD's Ontario announcement and even longer from Llano, but some very early performance figures have tipped up online courtesy of, well, distributed computing. To be more precise details of various distributed computing apps being tested on AMD Llano and Ontario systems have appeared on BOINC or the Berkeley Open Infrastructure for Network Computing if you prefer.

BOINC is a non-commercial middleware for various distributed computing programs such as SETI@home and several other less famous ones. The site is listing details of AMD engineering sample CPUs which are listed as AMD64 family 18, model 0 stepping 0 and AMD64 family 20, model 0 stepping 0 which is meant to be AMD’s upcoming Llano and Ontario APUs.

The Llano APU is listed as having three CPU’s, although this would be cores rather than actual CPU’s and the Ontario is a dual core. The Llano system is kitted out with 8GB of RAM, but we’re fairly certain the 1MB of cache listing is somewhat off target. The Ontario system has 2GB of RAM, but 230MB of that seems to be taken up by the GPU. Again the cache listing is odd, at 448KB. It’s possible that the cache is per core rather than for the entire CPU, as that makes sense judging by other CPU listings on BOINC and L3 cache doesn’t appear to be taken into account.

It’s hard to draw any real conclusions from the numbers, as the clock speeds for the two CPUs aren’t listed. Without knowing that it’s hard to compare the processors to current products in the market, but the Ontario is looking a lot more “ready” than the Llano. The only numbers on offer is measured floating point and integer speeds in million ops/sec, The Ontario comes out with 1,351 million ops/sec for floating point and 3,047 million ops/sec for integer performance running Windows Server 2008 of all things.

We also managed to dig out some numbers for the same CPU running an unspecified version of Linux with kernel 2.6.31.5 and here the floating point speed is slightly lower at 1,330 million ops/sec, but the integer performance went up slightly to 3,120 million ops/sec. This compares quite favourably to Intel’s dual core Atom 330 which only manages 871 million ops/sec for floating point and 2,249 million ops/sec for integer.

The triple-core Llano processor on the other hand doesn’t seem to be performing nearly as well as it should as it has a floating point performance of a mere 1,196 million ops/sec and an integer performance of 3,711 million ops/sec. It’s worth taking into account that this is most likely a very early sample and isn’t running anywhere near full speed. It’s no point comparing it to anything either, as the numbers aren’t going to impress anyone at this stage.

If nothing else, this does show that AMD has working silicon of both the Ontario and Llano processors, although as mentioned, Ontario seems to be a lot closer to a finished product at this stage than Llano. Ontario should compare favourable to Intel’s CULV mobile processors, although we doubt Ontario will beat Intel’s ultra low power Core iSomethingmeaningless processors on anything but price, but sometimes that’s enough to be a winner.S|A

Source: semiaccurate.com

Random Post