backtop


Print 50 comment(s) - last by mmarq.. on Jul 20 at 3:34 PM

AMD's Giuseppe Amato dispels rumors and misinterpreted statements of "Fusion," GPGPU and the company

In an interview with Italian media, AMD Executive Giuseppe Amato, Technical Director of Sales & Marketing EMEA, discussed AMD's current market position and future products.

In the interview, Amato shed more light on the structure of AMD's upcoming Fusion processors. A misconception that Amato noted is that Fusion processors will not only be available in single-chip flavors, but also multi-chip formats. Two Fusion processors linked together would allow for parallel GPUs. He said that AMD has still not solidified the future plans of Fusion yet, but indicated it would be very likely to see a Fusion processor with a GPU and CPU connected through a CrossFire-like interface -- and have a total TDP of less than 120 Watts.

Amato also praised the flexibility of the Fusion processor in the interview and told Hardware Upgrade that it will allow AMD to "integrate a specific number of GPU and CPU cores depending on the customer and the uses for which they will use the chip." 

"AMD isn't just a microprocessor company anymore", he stated. After the acquisition of ATI, "AMD changed from a processor company to a platform company." This is where Fusion ties in. Its high grade of flexibility will combine GPUs and CPUs into one product. Amato believes that Fusion platforms will be able to specifically match the needs of its customers.

AMD's Fusion processors will also be closely tied to GPGPU. Using a GPGPU platform based on Fusion, AMD will be able to offer HPC systems that can do all kinds of work. Code that is more suited for CPUs will be executed on the CPU part of the Fusion processor, while code more efficiently run on a GPGPU will be run on the GPU portion of the processor. To sum it up, AMD's Fusion processors will be able to do a variety of work, allowing them to better meet the needs of AMD's customers.

Amato also dispelled rumors that AMD will be going completely fabless. He blames the source of the rumor as a misinterpretation of a speech Hector Ruiz gave. However, AMD plans to stick to a fab-less manufacturing model for GPU and chipset products.

The full interview can be viewed at Hardware Upgrade.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

sounds good
By omnicronx on 7/16/2007 3:07:06 PM , Rating: 3
this sounds pretty damn cool, i have been waiting on information on the fusion for a while. I can see how this would be great in the oem market especially, but i wonder, will the chip be integrated into the board, or will it be a slot keeping it upgradeable. I think a core/gpu configuration thats upgradeable would be great.. although gone would be the days of upgrading part by part haha.




RE: sounds good
By Regs on 7/16/2007 3:20:19 PM , Rating: 2
Yes, it's good to hear. I like that we actually heard from someone from AMD's marketing team and not from some unknown source reporting from another unknown source.

It's something AMD should do more often, and to recollect in the past few months, AMD is on the right track about opening up to the public by speaking to them directly.


RE: sounds good
By Khato on 7/16/2007 3:36:53 PM , Rating: 3
... I'd trust some unknown source more than someone from marketing >.>


RE: sounds good
By kenji4life on 7/16/2007 4:25:15 PM , Rating: 5
Don't know why you'd say that here. While the statement has merit, all he is doing here is dispelling rumors, not creating them.

I won't argue that marketing has never spread rumors. But this doesn't seem to be the case here; especially when he's talking theory & r&d at this point. Not promising anything like release dates or benchmarks. Everything that he's saying is entirely plausible and there's no reason as of yet to think it's all hot air. Don't forget how much money AMD sunk into this.


RE: sounds good
By Khato on 7/16/2007 5:14:02 PM , Rating: 1
Eh, the only rumor I see being dispelled here is that of AMD going fabless. In which case he states what most that follow such things already guessed - AMD's going to continue outsourcing the fabrication of graphics chips and chipsets.

As to the rest of the article, eh, AMD marketing promises/statements to generate interest in their 'products' that are under development. One of my favorite quotes from the article, "We prefer not to spread information on operating clock speeds and performance as of right now..."


RE: sounds good
By 91TTZ on 7/16/2007 5:38:22 PM , Rating: 1
quote:
Don't know why you'd say that here. While the statement has merit, all he is doing here is dispelling rumors, not creating them.


You should never believe the "official" word of a company. They have no incentive to tell the truth.

Remember a week before the PS3 price cuts? Sony's official position was that there were "no plans" to cut the price of the PS3. Yet a week later the price was dropped. Things like price changes take time for a company to approve and inform their distributors, so there's no way that Sony didn't have plans when their exec made that statement. They were busy working on the price drop and their exec knew it, yet he still lied through his teeth.


RE: sounds good
By Ringold on 7/16/2007 6:53:34 PM , Rating: 4
Give them a little slack. "Lie" is pretty tough language. Businesses and their exec's are on a leash of a certain length about what information can be disclosed, how it can be disclosed and when, for legal and ethical reasons probably as it relates to the owners(investors). Being careles about spilling the beans could lead someone to jail and if it weren't this way there'd be nothing illegal about leaking information to hedge fund guys before issueing a proper press release. Perhaps the exec in question could've phrased it better, but saying he lied is.. a little stronger of language then I'd use.


RE: sounds good
By TomZ on 7/16/2007 7:01:24 PM , Rating: 5
How can you tell when a marketing executive is lying?

When he moves his lips. :o)


RE: sounds good
By theapparition on 7/17/2007 8:15:29 AM , Rating: 3
If you look at Sony's (or Microsoft, or anynone else for that matter) press releases reguarding price cuts/product announcements, etc. they all go somethink like this.

"We have no plans on announcing a price cut at this time"

It's a carefully crafted factual statement. So, technically no lie. Sony didn't plan on "announcing" the price cut "at this time". They very well may have been planning the actual price cut for several months.

Microsoft has no plans to "announce" a new xbox product "at this time". We all know they have 65nm redesign in the works.

Try it, I think you'll like it.

"I have no plans to announce to my wife that I'm sleeping with her sister."

See, just rolls off the tongue.


RE: sounds good
By JeffDM on 7/20/2007 11:04:32 AM , Rating: 2
"It's a carefully crafted factual statement. So, technically no lie. Sony didn't plan on "announcing" the price cut "at this time". They very well may have been planning the actual price cut for several months."

That still doesn't argue why I should give this AMD guy any credibility. I don't care if it's not technically a lie, it's still a reason why no one should pay attention to them.


RE: sounds good
By Hawkido on 7/17/2007 2:15:59 PM , Rating: 2
quote:
You should never believe the "official" word of a company. They have no incentive to tell the truth.


Incentive, No, no real incentive. However as a publically traded company there are Severe Penalties for Lying to the investors in the form of official press releases.

Forward looking statments are given room for error (such as release dates being off for unforseeable reasons, clock speeds not meeting the benchmarks due to poorer than expected yields on initial batches. So long as the stated goal of releasing the product or eventually meeting the goal happens, no problem. Investor confidence may suffer a bit, for such grandiose statments, but not the core of investors. They realize you can't get rich all at once, you get rich alittle bit at a time.

How much you do you hate your dad if he says "we'll go to the circus", but doesn't? What if he has to postpone it a weekend or two because of surgery or work interferes?

How spoiled are you if you can't be alittle bit flexible for something you haven't even paid for yet?


RE: sounds good
By JeffDM on 7/20/2007 11:00:06 AM , Rating: 2
I don't think it matters, corporate mouthpieces are not going to give out useful pieces of information. I don't care if they aren't technically lies, carefully crafted "tidbits" that are intended to give the opposite impression of what it's really saying is quite snake-like. I hope they don't wonder why people don't trust corporate PR when they do this.


RE: sounds good
By creathir on 7/16/2007 3:27:48 PM , Rating: 3
With SLI/Crossfire, those days are already gone. (If you want to take advantage of those techs)
- Creathir


RE: sounds good
By robrbecker on 7/16/2007 3:52:55 PM , Rating: 2
I think from what I've read at the inquirer and other tech sites their approach will most likely be to have multiple sockets on a board that different types of "processors" plug in to. Think of a mobo with 2 or more AM3 sockets connected with hypertransport. You could put any combination of things into these sockets: one or more dual or quad core AMD general purpose CPUs or one or more AMD (ATI) GPGPU. Workstations or scientific types could put added floating point powerhouses. And the best part is that the system will figure out which processor would perform operations the fastest and delegate the work to the appropriate chip.
It's like a heterogeneous multi-core processor where each core is physically replaceable since it is it's own chip. AMD could put these one the same die to cut cost and power further (and probably will) for mass-market models.

I think this approach will make huge progress for AMD in the mobile space: better graphics with lower power consumption!


RE: sounds good
By cobalt42 on 7/16/2007 4:09:14 PM , Rating: 3
I think you're thinking of Torrenza. Similar in concept, but one of the differences is that Fusion has more emphasis on heterogeneity in a single chip and Torrenza as an approach to plug in to HT sockets, e.g. with miscellaneous accelerators including FPGAs.


RE: sounds good
By GabrielIkram on 7/16/2007 4:27:35 PM , Rating: 4
Well, Fusion can be a part of Torrenza. It will most likely be working as a part of a Torrenza platform. Remember, Torrenza is a new AMD platform, and Fusion will probably be able to fit right into the platform.

A Torrenza platform's motherboard consists of two accelerator sockets and a PCIe accelerator.

http://www.dailytech.com/AMD+Bringing+Torrenza+and...

What this means is that a highly specific system can be built. For example, for one accelerator socket I can decide to use a Fusion processor. For the other socket, I can use a third-party dedicated math coprocessor. Another highly-specific accelerator can be added using the PCI-Express interface. So basically, I can build a complete platform tailored specifically for my needs.

This is what I am seeing the new goal of AMD as, and this is what they will probably be targeting the corporate world with; extreme flexibility. If you need any futher explanation, go ahead and ask. Alternatively, you can also refer to this link: http://www.dailytech.com/AMD+Bringing+Torrenza+and...

or this link:
http://www.hwupgrade.com/articles/cpu/14/the-futur...


RE: sounds good
By Treckin on 7/16/07, Rating: 0
RE: sounds good
By Andrwken on 7/16/2007 11:14:34 PM , Rating: 2
Maybe someone can explain the long term benefits of this to me. I like the sound of the technology, but with 16 core designs already on the roadmap, doesn't this seem stopgap? Isn't the real goal to have multiple cores that can be assigned to doing specifice tasks, ie. video rendering or math? I can't find the article anymore but wasn't there just a write up on a quad core that was handling some pretty intensive 3d rendering on a new game engine. I would think other than the programming being the bottleneck, a 16 core nehalem could probably have half the cores do a pretty impressive job of rendering. 4 cores handling physics, and 4 for regular cpu function. Or are we too far out programming wise for that and the cores will have to be specialized for some time yet?


RE: sounds good
By mmarq on 7/17/2007 7:00:49 PM , Rating: 3
quote:
but i wonder, will the chip be integrated into the board, or will it be a slot keeping it upgradeable


In the first implementation it will be like today CPU + PCIe GPU, more a (hardware circuit + software tiny layer) called CTM, for application specific task stream acceleration.

http://en.wikipedia.org/wiki/Close_to_Metal

Reading the interview Amato said they were able to accelerate a virus scanner to performance not possible with only the CPU. It sure 'can' be done with * the majority * of applications, if software developers so program them, because they all could and surely will have stream specific tasks except the pure Integer ones.

Interview:
quote:
An example that we have shown that uses our previous video architectures is Tarari. By recompiling its antivirus scanner and using an AMD GPU, Tarari was able to reach significantly higher performance compared to what would have been obtainable using only a CPU


One honest comment i have is that thank god we don't have to go up trough to SSE50 anymore, because it will be useless, and not helping software developers because right now they are only passing from SSE2.

Would be a revolution if for having the best performance we wouldn't need to have the best CPU anymore. At least at streaming and FP.

Meaning has more applications go for CTM more irrelevant is witch CPU, Intel or AMD, is a little more general purpose performant than the other.

quote:
I think a core/gpu configuration thats upgradeable would be great.. although gone would be the days of upgrading part by part haha.


No.You would still be able to upgrade part by part

Interview:
quote:
A large error that has been made regarding Fusion is that people are thinking that this type of architecture will only be a single chip package architecture, meaning both the CPU(!) and GPU are to be integrated on the same die.


A) - General stream will come to the GPU or GPGPU by CTM (hardware circuit + software tiny layer) and or other schemes. Here we already have propositions from ATI and Nvidia(tesla) of GPGPU capables of more general computing tasks, like physics acceleration and other tasks.

ATI is more advanced in the game, but Nvidia has also implement a similar thing, and it will be possible to use Intel CPUs with it, either ATI or Nvidia.

B) - Streaming will come to the CPU by means:
__ B1) Making functional units inside the CPU capable of it, as is the case of today SSE units in all CPUs(more or less), only requiring additional logic and a software layer for load balancing with Stream units outside of a particular CPU die, and or for general application stream task acceleration, like in CTM or the the almost only software layer Intel derivative from the project Larabee.

Those functional units can possibly(?) have also more GPU centric tasks like vertex and shading processing, turning the CPU in a GPU like. Only AMD at the moment seems to me, to have any intention at this in a shorter period, because of the advantage of better 'clustering' of their CPU designs(?). It will happen?... we have to see.

Interview:
quote:
Thanks to the availablilty of a higher number of 'registries' General purpose GPU computing will also be made much easier in 2009 when Microsoft releases the DirectX 11 API.
(in that case CPU registries also IMO, because x86_64 can have 32 GPR, double of now, without breaking applications(?))

__ B2) Having inside the CPU die, *units separated from the traditional cores*, like other cores, connected through a Xbar, other link or sharing a L2/L3, and capables of streaming, like is the case of the IBM Cell processor implementation in the PS3, and the derivative in the Xbox.

But for load balancing with Stream units outside of a particular CPU die, and or for general application stream task acceleration, it would still be needed something like CTM or the almost only software layer Intel 'Larabee' derivative.

Again those separated units can surely have also more GPU centric tasks like vertex and shading processing, turning the CPU in a GPU like.

__ B3) Having inside the CPU package in a MCM configuration, like is the case of the C2Quad, a traditional CPU die and a GPU die. In this example a C2Quad(?) would be a C2 core + a GPU core. In here the CPU die and GPU die communicate trough a link or by sharing a L2/L3.

Again for load balancing with Stream units outside of a particular CPU MCM package and or for general application stream task acceleration, it would still be needed something like CTM or the almost only software layer Intel 'Larabee' derivative.

Interview:
quote:
AMD now possesses all of the technologies it needs to develop Fusion architectures. Whether it is a native solution with serveral cores integrated in the same die, similar to what we are using for Barcelona, or a multi-die package (author’s note: the same architecture used by Intel for its Core 2 Quad chips) composed of two separate silicon die installed on the same package, AMD is open to all technological evolutions that the market requires.


Conclusion:

This hardware circuit + software tiny layer is imperative, be it CTM, which seems to me to became the central part of Crossfire 2 for ATI, or Larabee which relays much more on software.

All those GPGPU and CPU/GPUs can be connected all together very effectively, trough a cache coherent protocol, used in mainframes multiparallel machines and clusters, of which IBM, AMD and others have very solid implementations, and which Intel will have when CSI arrives, if it has and how, cache coherency. That would make CPU importance even more irrelevant for the large group of straming capable applications.

More connected to 'us', in the enthusiast market, it is precisely there where AMD with HT3, DC 2, HTX slots like in the next RD790, and the ones after that http://www.amdzone.com/index.php?name=PNphpBB2&fil...
will absolutely rock!!...

I'm not payed to post, nor a fanboy, nor in defense of anyone, but it seems to me that AMD has a clear advantage, and if CSI don't came out relatively soon, Intel will be in trouble if it stays sticked to the same old shared FSB.



RE: sounds good
By indianpunk on 7/18/2007 4:52:56 AM , Rating: 2
i agree with you bro max u upload part by part maybe and just maybe ram but with already 4 gb ram possible and pc upgrades being Possible every 12-15 months (See anandtech article on intel price drops u'll know the trend )

I'll be going for a cheaper amd athlon x2 at the time being but i am sure come next august quad core will be cheaper by dozen and 1 tb's will be common as well and maybe just maybe ddr3 will be free flowing as well


RE: sounds good
By mmarq on 7/20/2007 12:48:40 AM , Rating: 2
"" but i am sure come next august quad core will be cheaper by dozen and 1 tb's will be common as well and maybe just maybe ddr3 will be free flowing as well ""

Without wanting to be ironic as you, but it seems you missed the point completely.

You can have a ready rig for all connected cache coherent GPGPU and CPU/GPUs, this year, and for Vista Ultimate or Linux!!... there are tradeoffs, there are always tradeoffs... (see here)

http://www.dailytech.com/article.aspx?newsid=8066&...

But don't have to wait for 1 tbs rams, free ddr3, quads or other pointless quarks...

** THIS IS ABOUT 10-20X, OR MORE FOR APPLICATIONS CENTERED HEAVILY AROUND STREAMING **

Forget about 16- 80 core cpus, intel, amd or whatever, tera of ram or other bull to make you lighter in the pockets... this is a revolution!

"" By contrast, ad-hoc stream processors easily reach over 10x performance ""

http://en.wikipedia.org/wiki/Stream_processing


"Well, we didn't have anyone in line that got shot waiting for our system." -- Nintendo of America Vice President Perrin Kaplan











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki