Thursday, November 12, 2009

AMD to launch combo graphics-microprocessor chips in 2011

AMD to launch combo
graphics-microprocessor chips in 2011


Advanced Micro Devices released new details on its long-awaited Fusion chips, which combine graphics and microprocessor functions into a single chip in a computing platform that will launch in 2011.

Rick Bergman, senior vice president and general manager of AMD products (right), said at the company’s analyst meeting today that AMD is preparing a whole series of Fusion chips that will be made using a 32-nanometer manufacturing process that will be ready for full-scale production in 2011.

The chips will hit new targets in power efficiency, performance, and size and could be used in powerful desktop computers as well as the smallest netbooks, which are smaller than laptops and are meant for surfing the web.

These chips will have more than a billion transistors on them, Bergman said, but will likely be smaller and less powerful than separate, stand-alone microprocessor or graphics chips that launch in the same time frame.

The point is that they eliminate a chip in the PC system, cutting out costs and processing bottlenecks. The new chips also puts pressure on Intel and Nvidia, which make separate chips.

Samples of these chips will be sent to customers in the first half of 2010. They’re on schedule for production in late 2010 and will likely appear in customer PCs in 2011, Bergman said. That’s a shorter timeline than ever before. AMD has been promising Fusion chips since it acquired ATI Technologies for $5.4 billion in 2006. But it has delayed the launch, due to the difficulty of designing combination chips. Today’s schedule is consistent with the roadmap AMD announced a year ago.

One of the chips under design is code-named Llano. It will have four microprocessor cores, or computing brains, as well as a powerful graphics core that can handle the latest Microsoft DX11 graphics technology. A second chip in the works is code-named Ontario, which will have two microprocessor cores and a graphics core. By comparison, AMD plans to have a 12-core microprocessor on the market by 2010. As you can see, it’s easier to fit more cores on a stand-alone microprocessor than it is on a hybrid chip.

AMD calls these hybrid chips APUs, for accelerated processing units, compared to central processing units (CPUs, or microprocessors) or graphics processing units (GPUs). In other news, AMD executives said they expect to operate profitably from now on, or at least that is the major goal.



..............................................................................................

Saturday, November 7, 2009

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
New 125W TDP For The 965

New 125W TDP For The 965:

Today AMD is releasing a revised Phenom II X4 965 Black Edition processor. The TDP has been lowered from 140W to 125W, and also happens to be the new C3 stepping. Along with this new processor launch the prices for the 965's have been lowered to an expected price of $195 USD, a savings of about $20.

New Phenom II X4 965 Processor:

* TDP = 125W (down from 140W)
* Hardware C1E implemented (faster switching of power states means virtually no impact to performance by power management when BIOS support is properly implemented)
* “Heavy” load support for DDR3-1333 (with proper BIOS implementation, memory controller will now support up to 4 x DDR3 DIMMs @ 1333MHz)

How do you tell the difference? The last letter of the OPN numbers are different:

* 125W = HDZ965FBK4DGM
* 140W = HDZ965FBK4DGI

This processor runs at a slightly lower voltage than the previous model. The 140W 965 has a default core voltage of 1.4v, whereas the new 125W 965 runs at 1.375v. In theory these chips should provide slightly better overclocking (on average).

The Phenom II X4 965 (140W) has been out for a while, so I'm going to go light on the benchmarks in this article. Also the Phenom II X4 965 is only 200 MHz (or 6%) faster than the Phenom II X4 955, which we have a full review here.




As with previous Athlon II / Phenom II releases, a BIOS update will probably be needed to ensure complete compatibility and functionality with existing AM2+ / AM3 motherboards. Most updates that support these chips have been out for a month or more. Worst case scenario if you are trying to run a newer chip on an older AM2+ board without an update is you will probably have to disable CnQ.

AMD Phenom II X4 965 Processor Technical Specifications:

  • Model Number & Core Frequency: X4 965 / 3.4GHz (Black Edition)
  • OPN: HDZ965FBK4DGM (“M” indicates new revision)
  • L1 Cache Sizes: 64K of L1 instruction and 64K of L1 data cache per core (512KB total L1 per processor)
  • L2 Cache Sizes: 512KB of L2 data cache per core (2MB total L2 per processor)
  • L3 Cache Size: 6MB (shared between cores)
  • Total Cache (L2+L3): 8MB
  • Memory Controller Type:ype: Integrated 128-bit wide memory controller*
  • Memory Controller Speed: Up to 2.0GHz with Dual Dynamic Power Management
  • Types of Memory Supported:
    • PC2-8500 / DDR2-1066MHz
    • PC3-10600 / DDR3-1333MHz
  • HyperTransport 3.0 Link: One 16-bit/16-bit link @ up to 4.0GHz full duplex (2.0GHz x2)
  • Total Processor Bandwidth: Up to 37.3 GB/s total bandwidth
  • Packaging: Socket AM3 938-pin organic micro pin grid array (micro-PGA)
  • Fab location: GLOBALFOUNDARIES Fab 1 module 1 in Dresden, Germany (formerly AMD Fab 36)
  • Process Technology: 45-nanometer DSL SOI (silicon-on-insulator) technology
  • Approximate Transistor count: ~ 758 million (45nm)
  • Approximate Die Size: 258 mm² (45nm)
  • Max Temp: 62° Celsius
  • Nominal Voltage: 0.875 - 1.4 Volts
  • Max TDP: 125 Watts
......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
Pricing and Availability

The table below lists most of AMD's newest processors according to prices I found recently at Newegg & AMD's Processor Pricing List.

CPU-Z Info:

This screen shot is with everything set at default (for the processor).


Phenom II X4 965 - 125W TDP
AMD Overdrive 3.1:

Your favorite all-in-one overclocking tool just got better. A new version of AMD Overdrive (version 3.1) should be available for download today.

Some new features / improvements include:

* CPU model-specific profiles which should benefit OC results of some key chips
* Improved compatibility with AMD Cool’n’Quiet
* More profiles, the option to auto-enable Smart Profiles during installation and various fixes and improvements


......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU
Test Setup, Methodology, and Benchmarks Used

Too keep things consistent and try to maintain and "apples to apples" comparison, we try to use as much common hardware across the board. Each system's BIOS was set to "Optimized Defaults", with a verification of CPU & Memory speeds and timings.


We started out with a clean install of Vista and performed all the Windows updates to bring it up to current levels as of when this review was written. Then we installed all out benchmark programs and ran a disk defrag. Afterwards we disabled some un-needed Vista features that could affect results such as SuperFetch, System Restore, Defrag, Indexing, and Remote Assistance. Furthermore we ran the Bapco Auto-Configuration Tool and used its "Level 3" settings which disables a bunch more stuff:

* Disables screen saver
* Disables Windows Update
* Disables Desktop Cleanup Wizard
* Disables User Account Control (UAC)
* Disables the Sidebar
* Disables Windows Security Center warning messages
* Disables Windows Firewall
* Disables incoming Remote Desktop connections
* Disables Windows Error Reporting to Microsoft
* Prevents the Welcome Center from running at startup
* Disables Windows Defender

Each of the benchmarks were run at least five times, with the average being reported.



The following tests were performed:

* Synthetic Benchmarks:
o SiSoftware Sandra - 2009 SP1 (15.60)
o Lavalys Everest Ultimate Edition - v4.60
o WinRAR - 3.80
* 3D Rendering Performance:
o POV-Ray - 3.7 beta 29
o MAXON CINEBENCH - R10
* Video Encoding Performance:
o Tech ARP's x264 Benchmark - v2.0
o Windows Media Encoder 9 (x64 Edition)
o Divx Encoder - 6.6.1 (Codec 6.8.5)
* Math Calculation Performance:
o Wolfram Research Mathematica - v7
o wPrime - v2.00
o PiFast - 4.3
o HyperPi - 0.99b (Contains Super-Pi Mod v1.5)
* Gaming Performance:
o Lost Planet: Colonies Edition
o Far Cry 2
o Crysis Warhead
o Enemy Territory: Quake Wars
o Valve Particle Simulation Benchmark
o Valve VRAD map compilation tool
* Miscellaneous:
o Notfred Folding @ Home Benchmark

......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
Everest CPU Queen's Problem

Everest CPU Queen's Problem:

This simple integer benchmark focuses on the branch prediction capabilities and the misprediction penalties of the CPU. It finds the solutions for the classic "Queens problem" on a 10 by 10 sized chessboard.

At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. For example -- with HyperThreading disabled -- the Intel Northwood core processors get higher scores than the Intel Prescott core based ones due to the 20-step vs 31-step long pipeline. However, with enabled HyperThreading the picture is controversial, because due to architectural bottlenecks the Northwood core runs out of internal resources and slows down. Similarly, at the same clock speed AMD K8 class processors will be faster than AMD K7 ones due to the improved branch prediction capabilities of the K8 architecture.

CPU Queen test uses integer MMX, SSE2 and SSSE3 optimizations. It consumes less than 1 MB system memory and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


Everest CPU PhotoWorxx Processing:

This integer benchmark performs different common tasks used during digital photo processing.

It performs the following tasks on a very large RGB image:

* Fill
* Flip
* Rotate90R (rotate 90 degrees CW)
* Rotate90L (rotate 90 degrees CCW)
* Random (fill the image with random coloured pixels)
* RGB2BW (color to black & white conversion)
* Difference
* Crop

This benchmark stresses the integer arithmetic and multiplication execution units of the CPU and also the memory subsystem. Due to the fact that this test performs high memory read/write traffic, it cannot effectively scale in situations where more than 2 processing threads used. For example, on a 8-way Pentium III Xeon system the 8 processing threads will be "fighting" over the memory, creating a serious bottleneck that would lead to as low scores as a 2-way or 4-way similar processor based system could achieve.

CPU PhotoWorxx test uses only the basic x86 instructions, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


Everest CPU ZLib Compression:

This integer benchmark measures combined CPU and memory subsystem performance through the public ZLib compression library Version 1.2.3.

CPU ZLib test uses only the basic x86 instructions, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


Everest CPU AES Encryption:

This integer benchmark measures CPU performance using AES (a.k.a. Rijndael) data encryption. It utilizes Vincent Rijmen, Antoon Bosselaers and Paulo Barreto's public domain C code in ECB mode.

CPU AES test uses only the basic x86 instructions, and it's hardware accelerated on VIA PadLock Security Engine capable VIA C3 and VIA C7 processors. The test consumes 48 MB memory, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
Everest FPU Julia Fractal

Everest FPU Julia Fractal:

This benchmark measures the single precision (also known as 32-bit) floating-point performance through the computation of several frames of the popular "Julia" fractal. The code behind this benchmark method is written in Assembly, and it is extremely optimized for every popular AMD and Intel processor core variants by utilizing the appropriate x87, 3DNow!, 3DNow!+ or SSE instruction set extension.

FPU Julia test consumes less than 1 MB system memory, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.



Everest FPU Mandelbrot Fractal:

This benchmark measures the double precision (also known as 64-bit) floating-point performance through the computation of several frames of the popular "Mandelbrot" fractal. The code behind this benchmark method is written in Assembly, and it is extremely optimized for every popular AMD and Intel processor core variants by utilizing the appropriate x87 or SSE2 instruction set extension.

FPU Mandel test consumes less than 1 MB system memory, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


Everest FPU SinJulia Fractal:

This benchmark measures the extended precision (also known as 80-bit) floating-point performance through the computation of a single frame of a modified "Julia" fractal. The code behind this benchmark method is written in Assembly, and it is extremely optimized for every popular AMD and Intel processor core variants by utilizing trigonometric and exponential x87 instructions.

FPU SinJulia test consumes less than 1 MB system memory, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.


......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
POV-Ray, x264 HD, WME9, Folding@Home

POV-Ray:

POV-Ray is a free raytracing program. Version 3.7 beta has 64bit support and native SMP capability. We used the built-in benchmark to test multithreaded performance.


x264 HD Benchmark:

x264 is a free library for encoding H264/AVC video streams. Tech ARP has a nifty benchmark app to bring some automation and standardization to the encoder. Simply execute a batch file and it will run three tests using two different encoder versions. We are only reporting the newest version since its scores are a little better. This is multithreaded and maintained very high CPU usage, especially during the second pass.


Windows Media Encoder 9 - x64 Edition:

Windows Media Encoder 9 x64 Edition is free from Microsoft and is yet another multithreaded app. The conversion was actually a two pass process, but WME only reports a final time. The first pass doesn't fully utilize all CPU power, but the second pass comes pretty close to full utilization. For this test we used a 10 Min ~130 MB video to re-encode to the 2Mb/s DVD quality format.


Folding @ Home Benchmarks:

We at EXTREME Overclocking are Folding @ Home nuts! If you fold then you have probably seen the EOC Folding Stats pages. Using the Folding@Home benchmark CD created by notfred, we ran the benchmark64 app to get these numbers. The benchmark doesn't run any of the new SMP work units, but it does give general relative performance of each processor.


......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
Overclocking

Overclocking:

As mentioned initially (and the whole purpose of this article) is that the new 965 model is a lower power 125W TDP. We should expect that overclocking would be slightly better than than the 140W TDP version. I was able to get a pretty decent OC leaving at stock voltage, however past that I wasn't able to squeeze out much more. I tried up to 1.525v but at the rate the system crashed you could tell it was a downhill slope with the increase in voltage. I messed around a little overclocking the memory controller, I was up around 2.5GHz without any issues.

We run Windows Vista x64 edition... If you are still running only a 32-bit operation system then you could expect to see about a 200 MHz increase over these numbers.

* 3.4 GHz - Stock Speed & Voltage (1.375v)
* 3.8 GHz - Max Overclock @ Stock Voltage
* 3.9 GHz - Max Overclock @ 1.425v

3.8 GHz @ Stock Voltage (1.375v)


Max OC - 3.9 GHz @ 1.425v

......................................................................................................

AMD Launches 125W Phenom II X4 965 CPU

AMD Launches 125W Phenom II X4 965 CPU
Product Summary & Final Conclusions


Product Summary & Final Conclusions:

There's really no downside to the new Phenom II X4 965 - 125W TDP model. The chip will overclock right around that 4 GHz zone, but it is just luck of the draw and each system setup will have an affect. We've seen numerous times in the past where the same model chip comes out with a new revision and often a lower TDP rating, both AMD and Intel do it. The new C3 stepping chips I'm sure will eventually trickle into all of the Phenom II's.

I did not get a chance to mess around with AOD 3.1 too much, but it is nice to see the AMD team continually strive to improve its functionality. AMD Overdrive really makes overclocking the Black Edition processors a snap, for both the novice and expert. It is much more convenient to adjust things on-the-fly than constantly rebooting the system to change settings in your BIOS.

......................................................................................................

Tuesday, November 3, 2009

Intel Prototypes New Cooler for Gulftown CPU

Intel Prototypes New Cooler for Gulftown CPU


Intel retail boxed processors always come packed with a heatsink and fan. While they aren't the most elaborate cooling solutions, they are quiet, efficient, and even adequate for some light overclocking.

Of course, those serious about keeping their hot chips cool will opt for aftermarket solutions. For the upcoming six-core Nehalem-based Gulftown CPU, Intel will be packing it with a tower-style heatsink.

PCWorld.fr snapped pictures of a prototype of the cooler that will be paired with Gulftown. Besides just a bigger heatsink with more fins, the proposed cooler also features four copper heatpipes.

....................................................................................................

Intel's SSD firmware brings speed boost, mass death again

Intel's SSD firmware brings speed boost, mass death (again)

Intel's new SSD firmware massively boosts performance with Windows 7-related ATA commands, but has a bricking bug that's lost data for many users. How much longer will Intel's streak of SSD firmware foibles continue, and how much will these problems hold back the SSD revolution?

Intel's new firmware and toolkit for its new X25 line of solid state disks (SSDs) delivered a massive improvement in their already blisteringly fast performance this week, with increased write speed and a near-elimination of usage-induced performance degradation. The update also included support for the Windows 7 "trim" feature, which will boost SSD's speed and lifespan. So with a 40 percent performance boost and extra Windows 7 goodness, what's not to like about this update? Try data corruption, which some users have reported, causing the update to be pulled not long after it was released. This latest in Intel's uninterrupted streak of firmware foibles will probably be resolved as quickly as the prior two, but it may fuel ongoing doubt about the stability of the SSD as a reliable storage device.

Intel's G1 SSD introduced a subsector remap algorithm which massively increased performance against comparable SSDs, seemingly without cost, until a PCPerspective review showed that the new algorithm's remap table would become fragmented over time and cause degradation, sometimes to below the performance of a regular laptop hard drive. Shortly after this was fixed with a firmware update, Intel's G2 SSDs, with brand-new 34nm NAND, appeared on the stage with further improvements. These, though, were plagued by a lockout problem involving BIOS drive passwords, stopping shipments of the new drive until the issue was resolved via a firmware update.

Intel's latest problem, the third, causes unpredictable, irreversible bricking of G2 SSDs that have been updated to the newest firmware. While it's likely to be fixed fast, this new bug underscores a long and disturbing trend of strange bugs in SSDs from Intel and other drive makers. From lifespan problems to the JMicron controller issue which drastically reduced performance of many makes and models of SSDs, to Intel's ongoing problems, evidence is mounting to convince the user that SSDs aren't ready for the prime time, even as the SSD turning point draws near.

More's the pity, because the new firmware and its related utility resolve one of the major problems in SSD performance: the total lack of coordination between the OS and the SSD. When operating systems "delete" files, they don't actually write zeroes to the relevant sectors, but rather declare the data deleted and move on. This is massively faster (as users who have deleted a 100GB porn pirated movie Linux ISO stash in a few seconds are acutely aware), but causes problems for SSDs. NAND flash reads in pages, but writes in much larger blocks, necessitating a "read-modify-write" operation when a partial write occurs to a block already containing data. This means that when a write occurs to the former location of a deleted file, the SSD, jealously guarding its worthless "deleted" data, performs an unnecessary RMW cycle. As the drive ages, and all available sectors are filled with detritus from deleted files, this can lower performance dramatically.

The solution is to allow the SSD and OS to communicate. The latest revisions to the ATA specification allow a command called "TRIM," which tells the SSD that the data has been deleted. While the data remains, the SSD knows it's junk, and won't trigger an RMW cycle when it's partially overwritten. Performance is nearly as good in old drives as in new. Support for TRIM is native in Windows 7, and Intel's new firmware and utility allow support, for Intel-branded SSDs only, in Windows Vista and Windows XP. Other manufacturers of SSDs have released TRIM-supporting SSDs for Windows 7, but none had TRIM support under prior operating systems.

The new update also boosted write performance of the 160GB version of the X25-M G2 by about 40 percent at open-box. The 160GB model had, confusingly, the same performance specs as the 80GB version at launch, and this has been remedied. Intel hasn't announced exactly what changes account for the performance boost.

Ultimately, both the Intel-exclusivity and the bricking problems associated with the TRIM command will dissipate, as SSD options and optimizations improve and the market expands. However, until an SSD launch is as trouble-free as an HDD launch, SSDs will continue to face challenges in the enterprise. The question is, how long will it be until inexplicable bugs stop delaying and blunting SSDs' every success?


...................................................................................................

Friday, October 23, 2009

AMD's Athlon II X3 435

AMD's Athlon II X3 435

A month ago AMD introduced the world’s first quad-core processor to debut at $99. Last week, AMD announced its third quarter earnings for 2009. While the company as a whole lost money, the Product Company (CPU and GPU design) turned a small profit. I don’t want to say that the worst is behind AMD, but things are definitely looking up.

And for the consumer, AMD is providing a ton of value these days. You're getting more transistors per dollar than Intel will give you, and it's not just bloat, these things are fast:

The value train continues with todays introduction of the first triple core Athlon II processors: the Athlon II X3 435 and 425. Clocked at 2.9GHz and 2.7GHz respectively, these processors are simply Athlon II X4s with one core disabled.

They’re also quite affordable. The 435 will set you back $87 while the 425 costs $76. This puts them on par with Intel’s Pentium E6000 series dual core processors, but cheaper than the Core 2 Duo E7500. This has been AMD’s high end dual core strategy for the Phenom’s life: sell three cores for the price of two. And in the past, it has worked.

The X3s AMD is announcing today are clocked high enough that you still have good performance in single threaded applications, and in those that can take advantage of three cores you’re almost guaranteed to have better performance than the Intel alternative.

The real question you have to ask is whether it makes more sense to spend a little more than get a quad-core processor or not.

The Athlon II X3s are 45nm 95W TDP parts and work in both Socket-AM2+ and Socket-AM3 motherboards. As I mentioned before, these are architecturally identical to the X4s just with one core disabled. That means you get a 512KB L2 per core but no L3 cache.

I’ll spoil the surprise for you here: they’re faster than the equivalently priced Intel CPUs in most cases, but that’s not too surprising.

The Athlon II X3 435 is a bit more overclockable than the X4 620. Without any additional voltage we got 3.25GHz on our 620 sample, but our 435 yielded 3.33GHz:


With an extra ~15% voltage we could get 3.63GHz:

AMD is also introducing a slew of energy efficient Athlon IIs as well. They’re all in the table below:

These energy efficient processors are binned for lower voltages and thus have a 45W TDP. Unfortunately you do sacrifice clock speed in some cases as a result. There's also a hefty price premium, at the high end you lose clock speed and pay 44% more for a 45W TDP.

The Test


..................................................................................................

AMD's Athlon II X3 435

AMD's Athlon II X3 435
SYSMark 2007 Performance

Our journey starts with SYSMark 2007, the only all-encompassing performance suite in our review today. The idea here is simple: one benchmark to indicate the overall performance of your machine.

Overall performance under SYSMark is pretty balanced for the Athlon II X3 435. It's faster than the $99 quad-core (620) but slightly slower than the quad core 630. We're slower than the old triple core Phenom II X3 720 though.





................................................................................................