Then... I would suggest the search function. As you can imagine, your thread is not the first on the subject.
How come Xenos dose not get as much publicity as Cell?
- Thread starter Sound_Card
- Start date
Then... I would suggest the search function. As you can imagine, your thread is not the first on the subject.
crazygambit
Regular
inefficient
Veteran
Not sure what you have been reading. But your assumptions are wrong.
GPGPU performs many times SLOWER than the traditional CPUs when doing MOST tasks.
GPGPU performs many times FASTER than the traditional CPUs in SOME tasks.
Not to undermine their accomplishments, but AMD/ATI (or NVIDIA for that matter) did not just invent an amazing new product line that is going to take the world by storm. This seems to be what you seem to believe. More like people are just recently starting to realize that all this parrallel processing power on GPUs might actually be usefull for things other than just running Doom.
So far there is no "killer app" for GPGPUs. I don't think anyone reasonable person expects them to take off in popularity in the near future. Much less replace traditional CPUs. They should get better over time, but a lot of research is still needed. It's in it's infancy!
How so ?
Big investment in developing the thing that needs to be amortized. The thing itself is >200mm^2 which, while not huge, isn't exactly small.
Cheers
Yes, there are, but at this point in time those companies are not high profile consumer oriented companies given the current price and availability of CELL. You'll find it's being deployed within medical and physics research HPC and within the defence industry. You can expect it to take off in a bigger manner when it hits 65 nm and the price goes down.
There is no incentive from MS or AMD/ATI to do the same with Xenos, it's a purely designed for MS's consol.
You think the two chip solution with EDRAM in Xenos is cheaper? The process including EDRAM is not exactly cheap.
I think dantruon is also taking into account that the CELL volumes will not only be based on PS3 sales when it is starting to ge deployed in other consumer products by Toshiba and Sony.
I think it's because Sony itself tends to hype the crap out of there machines as uber-powerful. They did it with PS2, they did it with PS3 the same. E3 2005 example #1.
Secondly, because their fans run with it. Gaf for one thought Cell was the greatest thing since sliced bread early on, and vastly overhyped it. Some of them went really overboard.
Microsoft never positioned the 360 as a technical monster. Nor did they downplay specs like Nintendo. Maybe they thought early on that PS3 would be vastly more powerful so they didn't want to go that route. Who knows.
Also because as cool as Xenos is, in the end it's still comparable performance to a traditional design in RSX. Cell has a theoretical possibility at least, of being many times more powerful than conventional CPU's at certain tasks.
But it's funny I was thinking of making a thread myself in praise of Xenos now that all the competitors cards are on the table. I think the little chip has acquited itself very nicely. Personally I think a lot more highly of it now that we see final PS3 games. Moreso when you consider: fairly cutting edge production timeframe (Xenos had to be being produced around the time of 7800GTX or so, September of 2005 if not earlier). Small size (just 180mm main core, although yeah, there's another 70mm on 2nd die, much of that is EDRAM) and the fact it uses two all new graphics paradigms and seems to have pulled them off without a hitch (USA, EDRAM). And the fact ATI cant get a product out on time to save it's life these days, makes Xenos succesful cutting edge schedule even more unique.
________
Prilosec lawsuites
Secondly, because their fans run with it. Gaf for one thought Cell was the greatest thing since sliced bread early on, and vastly overhyped it. Some of them went really overboard.
Microsoft never positioned the 360 as a technical monster. Nor did they downplay specs like Nintendo. Maybe they thought early on that PS3 would be vastly more powerful so they didn't want to go that route. Who knows.
Also because as cool as Xenos is, in the end it's still comparable performance to a traditional design in RSX. Cell has a theoretical possibility at least, of being many times more powerful than conventional CPU's at certain tasks.
But it's funny I was thinking of making a thread myself in praise of Xenos now that all the competitors cards are on the table. I think the little chip has acquited itself very nicely. Personally I think a lot more highly of it now that we see final PS3 games. Moreso when you consider: fairly cutting edge production timeframe (Xenos had to be being produced around the time of 7800GTX or so, September of 2005 if not earlier). Small size (just 180mm main core, although yeah, there's another 70mm on 2nd die, much of that is EDRAM) and the fact it uses two all new graphics paradigms and seems to have pulled them off without a hitch (USA, EDRAM). And the fact ATI cant get a product out on time to save it's life these days, makes Xenos succesful cutting edge schedule even more unique.
________
Prilosec lawsuites
Last edited by a moderator:
Cell means a few things:
* A high performance, low power architecture.
* As an implementation, the chip used in PS3
* In extended form, Sony also invested in the R&D of hi-speed Cell network.
MS has not marketed Xenos as a platform, but Cell is designed to be one from day 1.
* It exists independently outside PS3
* It supports virtualization via the hypervisor
* You can run general purpose OSes on it to solve a large variety of problems, some of them faster than other CPUs on the market.
* It can run a large library of open source software
* It is supposed to scale up/down (by increasing/decreasing # of SPUs and PPUs).
* It revives supercomputing concepts from the 80s and 90s
A complete cell has 1 PPE (PPU) and 8 SPEs (SPUs).
Mercury Interactive, Yellow Dog Linux and most of the game publishers/devs for a start. Their customers (e.g., Barcelona Supercomputing Center) also invested time and resorces in Cell (buying Cell hardware, developing applications, training their people, etc...).
Last edited by a moderator:
archangelmorph
Veteran
Really? I think I can safely say that you speak for yourself alone in this regard..
I'm not even sure if a GPU-driven (in place of a CPU) driven console could even function at all let alone perform better than a Cell-based effort..
"The problem is..." ...?
wait a minute.. problem to who..?
To you..?
And if it's so much of a problem then why..?
If we want to include customers of these companies, others partnered with them or people just working with the chip, the list gets much longer, and not all of it is public. For example, Raytheon was mentioned again more recently as a company working actively with IBM w.r.t. Cell but they themselves have kept fairly quiet on it for some time. And, of course, you have very high profile customers like Los Alamos, new partners of Mercury's like Mentor Graphics, clients of the Terrasoft/SCE cluser like Los Alamos (again), Lawrence Berkeley National Lab, Oak Ridge National Lab, the Department of Energy and a bunch of universities and research units (not all of which are even officially engaged with IBM or Mercury - I know first hand of researchers working with Cell via simulators and PS3 who haven't official contact with IBM).
If the OP is wondering why Cell is higher profile, that's why, basically - it is in much broader use, being an open platform and not something limited to a single product and use-case. And some of those uses are very high profile (like the Los Alamos supercomputer), and thus attract a lot of attention. Some quick googling should have hinted at this at least.
Last edited by a moderator:
The cell gets more publicity, because Sony are masters of making up cool PR stories.
You never hear Peter Moore talking to the press about the power of the Xenos, and how it makes things possible..
However, you have heard Ken Kutagari talk about how the Cell is so powerful it cures cancer, how [insert random bla bla bla that sounds cool ] is only possible because of the power of the Cell processor.
Thats why.
Because of that, the casuals think that the PS3 is much more powerful than a high end PC, because it got 8 cores and it cures cancer... The media, are all about amout of viewers\readers,and since nobody even knows what the Xenos is, its easier to write about the supercomputer inside the PS3.
Its the same as with the PS2, everybody talked about how extremely powerful it was, and about the emotion engine.. There are many casuals that to this date, belive the PS2 is more powerful than the Xbox.
You never hear Peter Moore talking to the press about the power of the Xenos, and how it makes things possible..
However, you have heard Ken Kutagari talk about how the Cell is so powerful it cures cancer, how [insert random bla bla bla that sounds cool ] is only possible because of the power of the Cell processor.
Thats why.
Because of that, the casuals think that the PS3 is much more powerful than a high end PC, because it got 8 cores and it cures cancer... The media, are all about amout of viewers\readers,and since nobody even knows what the Xenos is, its easier to write about the supercomputer inside the PS3.
Its the same as with the PS2, everybody talked about how extremely powerful it was, and about the emotion engine.. There are many casuals that to this date, belive the PS2 is more powerful than the Xbox.
Just wanted to add...
Another thing good for cell is the redundency factor similar to what nvidia and ati/amd is doing with their GPU.
ppu w/8spu goes to product A (cell blades)?
ppu w/7spu goes to PS3s and maybe other products
ppu w ith less than 7 goes to products...
I vaguely remember an interview I read where Kutagari mentioned cell with 3spu can be used in TVs and or DVD recorders...
Another thing good for cell is the redundency factor similar to what nvidia and ati/amd is doing with their GPU.
ppu w/8spu goes to product A (cell blades)?
ppu w/7spu goes to PS3s and maybe other products
ppu w ith less than 7 goes to products...
I vaguely remember an interview I read where Kutagari mentioned cell with 3spu can be used in TVs and or DVD recorders...
Not to mention all the CELLs that don't make the 3.2GHz cut within the power envelope.
This is the only way CELL is ever going to make economic sense in CE products. Otherwise it would be up against dedicated solutions from Broadcom and others with significantly lower cost and power usage.
Cheers
Is there actually a high-end PC with a higher flop figure than the CELL, what kind of configuration do you have in mind?
Last edited by a moderator:
It's easier to write about Cell because it actually makes news. That's all there is to it. (And no, not the type of news PR agencies make up..in fact, if you look at the news that is out there on Cell, it predominantly does not relate to Sony at all).
Xenon/Xenon are much more internal codewords used for their respectve silicon duriong development. They were never intended for widespread marketing pr. Its like saying conroe or woodcrest sound stupid. They do but its not something you'll see in a magazine add. MS has been careful to market the system as whole based upon its strengths and has been very successful at that.
Sound_Card
Regular
