AMD announces FireStream 9170, first dedicated stream processor

The true computing geeks in the house are about to need adult diapers: AMD's launching a dedicated stream processor, the FireStream 9170, and an accompanying SDK today, which will be available in Q108. Granted, this 55nm "world's first Stream GPU with double-precision floating point technology" will run you two grand, but you'll be getting 500 GFLOPS at under 150 watts power draw, but you'll get all the insane math-crunching benefits of stream processing that your current number-crunching rig isn't seeing, despite the eight 8800s you've got jury-rigged into the thing. Press release after the break. Read more about stream processing here (because we're reall in no position to lecture on its benefits).

==PRESS RELEASE==
AMD Delivers First Stream Processor with Double Precision Floating Point Technology

AMD FireStream 9170 and supporting Software Development Kit unlock Stream-based Accelerated Computing

SUNNYVALE, Calif. - November 8, 2007 -AMD (NYSE: AMD) today announced the AMD FireStream 9170 Stream Processor and an accompanying Software Development Kit (SDK) designed to harness the massive parallel processing power of the graphics processing unit (GPU). AMD leveraged its unique collective expertise in both GPUs and CPUs to deliver the first integrated hardware and software development solution that meets the needs of the demanding high-performance computing (HPC) market. AMD plans to deliver the FireStream 9170 and supporting SDK to market in the first quarter of 2008. With this launch AMD expects to achieve another important milestone on the path to Accelerated Computing by delivering the first in a series of next-generation heterogeneous compute architectures.

"With a broad range of customer engagements underway, notably customers in the oil and gas, financial and engineering analysis industries, AMD is delivering on its vision of Accelerated Computing with breakthrough benefits for our enterprise customers," said Rick Bergman, Senior Vice President and General Manager, Graphics Products, AMD. "Leveraging the immense graphics processing capabilities acquired from ATI and the HPC domain expertise of AMD, we are developing strong relationships with system vendors and the supporting technology eco-system to deliver processing innovation through an open platforms approach."

AMD FireStream 9170
The AMD FireStream 9170 will be the world's first Stream GPU with double-precision floating point technology tailored for scientific and engineering calculations. Competitively priced at an MSRP of $1999 USD, it features up to 500 GFLOPS of compute power, rivalling many of today's supercomputers, and providing dramatic acceleration for critical algorithms. This second generation Stream Processor is built with 55 nm process technology and consumes less than 150 watts of power - delivering an exceptional performance per watt. In addition, the reduced heat dissipation allows it to function in dense design configurations. The FireStream 9170 is a single card solution with two GB of onboard GDDR3 memory to compute large datasets without CPU traffic. The asynchronous direct memory access (DMA) ensures data can flow freely without interrupting the stream processor or CPU.

"GPUs have long been known for their immense parallel processing performance but many challenges still remain in driving widespread customer adoption for general purpose compute," said Jon Peddie, President, Jon Peddie Research. "Leveraging its unique capabilities in high-performance CPU and GPU technologies, AMD is well positioned to drive an integrated hardware and software proposition that can deliver the best of both processing worlds to its HPC customers."

AMD FireStream SDK
The AMD FireStream SDK is designed to deliver the tools developers need to create and optimize applications on AMD Stream processors. Built using an open platforms approach, the AMD FireStream SDK allows developers to access key Application Programming Interfaces (APIs) and specifications, enabling performance tuning at the lowest level and development of third party tools. Building on AMD's Close to the Metal (CTM) interface introduced in 2006, the Compute Abstraction Layer (CAL) provides low-level access to the GPU for development and performance tuning along with forward compatibility to future GPUs. For high-level development, AMD is announcing Brook+, a tool providing C extensions for stream computing based on the Brook project from Stanford University. In addition, AMD also plans to support the AMD Core Math Library (ACML) to provide GPU-accelerated math functions, and the COBRA video library accelerates video transcode. Also available are third-party tools from top industry partners including RapidMind and Microsoft.

In addition, AMD is now a charter participant in HP's new HPC Accelerator Program, offering HP customers best practices and guidance for these technologies, and ensuring that accelerator hardware and software is qualified for HP servers running HPC applications.

"As innovative new HPC technologies like Stream Computing emerge, it is imperative we work with our partners to ensure an open systems approach to enable new levels of processing efficiency and performance," said Winston Prather, vice president and general manager of HPC at HP. "As part of HP's new HPC Accelerator program, we're working closely with AMD and our customers to deliver an optimal mix of hardware innovation and open, collaborative development environments to ensure delivery of best-in-class HPC platforms."

About AMD
Advanced Micro Devices (NYSE: AMD) is a leading global provider of innovative processing solutions in the computing, graphics and consumer electronics markets. AMD is dedicated to driving open innovation, choice and industry growth by delivering superior customer-centric solutions that empower consumers and businesses worldwide. For more information, visit www.amd.com.
*1 *Based on 320 stream cores and engine clock rates of 775-800MHz
*2 *Refers to maximum TDP of AMD FireStream 9170

Cautionary Statement
This release contains forward-looking statements concerning, among other things, product and technology introduction schedules and the extent and nature of engagement with customers and technology eco-system participants, which are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Forward-looking statements are commonly identified by words such as "would," "may," "expects," "believes," "plans," "intends," "projects" and other terms with similar meaning. Investors are cautioned that the forward-looking statements in this release are based on current beliefs, assumptions and expectations, speak only as of the date of this release and involve risks and uncertainties that could cause actual results to differ materially from current expectations. Risks include the possibility that Intel Corporation's pricing, marketing and rebating programs, product bundling, standard setting, new product introductions or other activities targeting AMD's business will prevent attainment of AMD's current plans; AMD will require additional funding and may not be able to raise funds on favorable terms or at all; customers stop buying AMD's products or materially reduce their operations or demand for its products; AMD will be unable to develop, launch and ramp new
products and technologies in the volumes and mix required by the market and at mature yields on a timely basis; global business and economic conditions will worsen; AMD will be unable to transition to advanced manufacturing process technologies in a timely and effective way; and AMD will be unable to maintain the requisite level of investment in research and development and capacity. Investors are urged to review in detail the risks and uncertainties in AMD's Securities and Exchange Commission filings, including but not limited to the Quarterly Report on Form 10-Q for the quarter ended September 29, 2007, which are available on www.sec.gov. AMD, the AMD Arrow logo, AMD FireStream, AMD Torrenza and combinations thereof, are trademarks of Advanced Micro Devices, Inc. Other names are for informational purposes only and may be trademarks of their respective owners.

[Michael @ Nov 8th 2007 1:20AM]

hmmm...maybe this is why everything else AMD is falling behind...I can't wait for programmers to catch up to this kind of processing power

[half&half @ Nov 8th 2007 1:26AM]

sex

better than it?

heh
thats opinion.

[JLTate @ Nov 8th 2007 10:13AM]

Sex = overrated. :(

[darren @ Nov 8th 2007 1:38AM]

i think i wet myself a bit. i knew AMD would pull through!

[Alvin @ Nov 8th 2007 1:56AM]

The above image is an in-game render using said stream processor.

True story.

[Taylor @ Nov 8th 2007 2:05AM]

I thought it was a photo.

I'm an nvidia fan but if AMD can catch up and surpass Nvidia with this, then I'm all for it!

[Alvin @ Nov 8th 2007 2:26AM]

I was just joshing. :-)

But really, this is cool tech. I really hope such innovation will soon find its way into CPUs so that Intel can finally start scrambling again. It's really a win-win for everyone.

[Sirocco @ Nov 8th 2007 1:57AM]

..about to need adult diapers? Or change brands.

[Nathan Bouscal @ Nov 8th 2007 1:59AM]

I'm going to have to sue AMD for this one... for damages to my monitor after spitting carbonated liquid all over it. Seriously though, about time for them to reassume their rightful place in the world. Bravo, AMD.

[RedBull Runner @ Nov 8th 2007 2:06AM]

I'm wet with anticipation...

and urine!

[Nirav Patel @ Nov 8th 2007 2:22AM]

NVIDIA already has a working solution and API...

http://www.nvidia.com/object/tesla_gpu_processor.html
http://developer.nvidia.com/object/cuda.html

How does that make AMD to be the "first dedicated stream processor" creator???

[457R4L @ Nov 8th 2007 2:58AM]

First with double-precision floating point technology.

[Brennan @ Nov 8th 2007 2:23AM]

holy crap!! 2GB of GDDR3 RAM n over 500 GFLOPS? This can surely beat the 8800 Ultra!

now if only it could come to the consumer market.......

[Blackster @ Nov 8th 2007 5:26AM]

you know that it'S not a graphiccard? ;)

[shanky @ Nov 8th 2007 2:37AM]

Is this different from IBM cell processor??

[JD @ Nov 8th 2007 3:42AM]

Yes, It's a shit ton more powerful and if I recall, the cell isn't a stream processor, at least not without software support (says wikipedia) so yeah, it's pretty different. 500 GFlops is a ton of power. The Cell has been shown to do 100Gflops on 8 cores in testing, you can check on the wikipedia site if you want.

[Randomness @ Nov 8th 2007 3:04AM]

But how many FPS can I get in Crysis with it?

[athousandleaves @ Nov 8th 2007 3:13AM]

I want to name my first child after Wikipedia.

[bombastinator @ Nov 8th 2007 7:27AM]

Here that Ladies? Breed with this guy and he'll name the kids after wikipedia.

You have been warned.

[Justin @ Nov 8th 2007 8:44AM]

Who said that's a guy? ;)

[Andir3.0 @ Nov 8th 2007 10:58AM]

On the internet, everyone's a guy, and anyone that looks like a girl is actually a guy... or something like that.

[paccoast2010 @ Nov 8th 2007 5:50AM]

Will this supercede multi-core chips for mainstream PC users - that is, most of us? The only things I do that run my CPU to the max are video related like Xvid encoding. Everything else I don't need extreme power for.

[Michael @ Nov 8th 2007 6:16AM]

Just think what this could do for the F@H project!

Then again, you have to feel sorry for the PS3 owners - if something better at folding comes along what will their machines be good for then?

[Rex @ Nov 8th 2007 10:55AM]

GPUs have been out-performing the PS3 for a while now.

[jordn @ Nov 8th 2007 7:11AM]

ha! take that, intel!

[Steve Jobs @ Nov 8th 2007 8:49AM]

But really isn't the most important question we should be asking here...Will it play doom?

[Justin @ Nov 8th 2007 8:44AM]

Yes. A hundred thousand Dooms at once.

[L.Cyphre @ Nov 8th 2007 11:10AM]

Forget DOOM, this thing'll play Duke Nukem Forever!

[Jerod @ Nov 8th 2007 9:42AM]

Does it blend? Hopefully my company will pony up a few bucs for one or two of these bad boys when they release. Currently all the integrated assembly models we are handling crash our most powerful machine every couple of minutes. Not fun.

[palehorse @ Nov 8th 2007 9:51AM]

Wake me up if/when this hits mainSTREAM in the form of a CPU, under $1000, that I can put in a rig at home...

but oh yes, it does sound nice!

[tekdroid @ Nov 8th 2007 10:56AM]

i'll be interested at the sub-100 pricepoint :)

[Technophile @ Nov 8th 2007 2:23PM]

Realtime raytrace anyone ?

[Andir3.0 @ Nov 8th 2007 5:16PM]

I was thinking this too... You don't need DirectX or OpenGL with ultra fast and scalable stream processors.
However, with 320 streams each stream would have to process:
(x * y * fps) / 320 / 1000 (sec to ms)
(for 1600x1200 @ 30fps) 180 pixels per millisecond per stream
(for 1280x1024 @ 30fps) 123 px/ms/str
(for 1024x768 @ 30fps) 74 px/ms/str ( I want to trademark PiXelMaSSTR now for some reason )

I'm not sure what the average per pixel rate is on a 775-800Mhz core, but it doesn't sound like it would be too far off to be getting around 100 per millisecond.

[jimmyfinch @ Nov 8th 2007 9:05PM]

Faster than the speed of light anyone?

[Clark H. @ Nov 9th 2007 3:16PM]

I reall don't get it... :D

^_^

[gantlord @ Jan 10th 2008 10:34AM]

Watch out for that 500 GFLOPS figure! It's classic bait and switch. They talk about double precision then mention 500 GFLOPS. The 500 GFLOPS is a single precision figure. Not that impressive when you're likely to see as much as 1 TFLOPS from the series-9 Tesla chips (and they may well have additional DP functionality).

P.S. Not a fanboy of either Chimpzilla or Graphzilla