AMD announces FireStream 9170, first dedicated stream processor

Bigfoot

Limp Gawd
Joined
Aug 9, 2005
Messages
171
Said to be available in Q1 of '08, "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."

Sounds pretty hot :eek:

http://www.engadget.com/2007/11/08/amd-announces-firestream-9170-first-dedicated-stream-processor/
 
I want to see this thing hit newegg. Open source development here we come (assuming the SDK allows for it).
 
So they are re labeling a ~r600 with some fancy schmancy name,and charging 2k USD for it ?

Cool ! More power to em ! :cool:
 
So thats why their new card needed 320 stream processors? it was going to moonlight as a super computer?
 
Actually there's no way this is just an R600 doing this unless there's some really fancy schmancy software trickery going on. Those chips can't do double precision floating point ops. This was the same method of attack NVidia was supposedly going for in a GPGPU that was announced earlier this year (the 1TF on a card announcement) but I haven't heard a peep about that guy since. Plus the article mentions 55nm so it's gotta be an RV670 derivative.

Wonder how they're doing the double precision on there too as I never heard of that being a feature. So weird that the RV670 actually has fewer transistors than the R600 but has added UVD functionality and now possibly this? Just a note, double precision adds almost nothing to the graphics side of things, but for GPGPU this is huge and will bring these things closer to par with the level of accuracy that supercomputing platforms can provide researchers.

This will be huge for AMD at those kinds of margins. If this indeed does come out before NVidia's offering there's going to be a ton of buyers. That may sound like a lot of money but for the ability to have the computing power of an entire cluster in one card people will pay out the wazoo.
 
I think nvidia has their shit more togehter when it comes to this GPGPU concept, but only time will tell.
 
Yeah obviously its a PCIe card.

Read the last third of that article, wtf are they talking about?

  1. I'd expect to see some form of external PCIe derivatives with multiple cards as well (NVidia also plans on doing this). I'd have to think most of the boards these are going into aren't going to be crossfire capable. With the types of applications this will be used for the addition of multiple video cards would scale almost perfectly with performance increase.
  2. Which topic in particular?
 
  1. I'd expect to see some form of external PCIe derivatives with multiple cards as well (NVidia also plans on doing this). I'd have to think most of the boards these are going into aren't going to be crossfire capable. With the types of applications this will be used for the addition of multiple video cards would scale almost perfectly with performance increase.
  2. Which topic in particular?

Yeah I believe in the future their is going to be some sort of external PCIe, atleast thats what nvidia has been telling us, I believe that THIS card is internal however.

And I was talking about this.
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.
 
That section you quoted is just a disclaimer. They can say that they plan on doing this or that, but in the end if it doesnt work out, they can always point you to the disclaimer and say, hey we told ya so...
 
I think that might actually be something required by the SEC. The stuff in the middle is more than likely AMD's original complaint in filing the suit. I'm sure there's some legal reasoning behind it because the SEC is actually involved in the suit and having to notify investors of the case.
 
I think that might actually be something required by the SEC. The stuff in the middle is more than likely AMD's original complaint in filing the suit. I'm sure there's some legal reasoning behind it as to not mislead shareholders into investing for the purpose of this as yet unreleased technology that has the possibility of never seeing market.

It's actually not required by law or anything like that, however if you want to cover your ass then you better put in a well thought out disclaimer.

In this case AMD says that we will have this product ready at this time. Lets say that AMD didnt have the disclaimer, and they failed to launch at the time this press release says, they can get there asses sued off. However with this disclaimer they can simply say "Hey we told ya so"
 
I think nvidia has their shit more togehter when it comes to this GPGPU concept, but only time will tell.

its been a while, but dont they simply, with their current model, provide an api to do this kind of stuff on their standard graphics cards.

its seem like if ATI is doing something similar, ATI has the upper hand since they're now released full specs for their graphics cards for open-source development. scientific ventures can thus write their own API and interface directly with the hardware to do anything they want. but this is assuming this new card isnt substantially different from their graphics cards...but it seems like it might be
 
its been a while, but dont they simply, with their current model, provide an api to do this kind of stuff on their standard graphics cards.

its seem like if ATI is doing something similar, ATI has the upper hand since they're now released full specs for their graphics cards for open-source development. scientific ventures can thus write their own API and interface directly with the hardware to do anything they want. but this is assuming this new card isnt substantially different from their graphics cards...but it seems like it might be

Actually it looks to me like it is using a 3870 chip (based on the stated clockspeeds and wattage requirements, little higher due to 2GB of RAM I'm guessing) that has been somehow modified to enable double precision floating point operations. I don't know the logistics behind actually implementing that on silicon but apparently it's not too huge of an undertaking if both of the big boys are doing it to their graphics products turned supercomputers.

Has anyone here actually written a program for a GPGPU? How easy is it to program for these things? I'd be really interested in trying to get into that to see what it can do. Hell with that much raw power who wouldn't want to toy around.
 
meh more AMD propaganda.. trying to hang on to investors they put out this shit. They can't even do a regular CPU right, I'm supposed to believe they can create a mini-supercomputer AND a fused GPU/CPU in little more than a year from now?

Give me a break AMD.

Sounds like the Nazis talking about "wonder weapons" as they got the shit kicked out of em.
 
I know several scientists who have been asking me about doing number crunching on a GPU. There certainly is a demand for a product like this; obviously it's somewhat more niche than regular gaming video cards though.
 
They can't even do a regular CPU right, I'm supposed to believe they can create a mini-supercomputer AND a fused GPU/CPU in little more than a year from now?

Give me a break AMD.

lol @ you, the market would have precluded any cpu chip AMD made that was inferior, ergo they must have made something right.

http://www.stuff.co.nz/4267519a28.html
nb the intel advertisment in it

lotta hype around this one, if its making it this far down under i guess.
 
What kind of applications is this product geared for exactly?
 
Analysis of very large sets of data. This is targeted toward the scientific and engineering community.
 
It isn't [H]ardForum without people pissing on any AMD new product threads :D
 
Back
Top