KK slips CELL performance at decoding HD streams.

[Mulciber]

one would hope that for something like that they would have a different asic with say 1 or 2 SPEs with half the memory on each of them, a castrated memory interface and possibly smaller L2 cache for the PPE.

such a chip would be reletively small, especially on .65 micron, and would still be quite powerful.

Why? They're gonna make millions of PS3 Cell chips, might as well keep making those and stick them in BR players. Having different chips will be more difficult than having only one chip to be used for everything.

i doubt "economics of scale" matters much when you're producing an asic that huge and then wasting it in BD players. they dont have an unlimited foundry capacity and they could easily make a chip 1/4 that size and still retain the exact same basic set of dev tools and functionality. you really thinking designing a second asic for use in more mundane tasks is such an arduous process that Sony is incapable of it?

they already plan to use these in the PS3 and in the server space. they also have intentions to use Cell in BD players, tvs, pdas, and cell phones if you believe them. I cringe at the thought of an 8 SPE cell in a cell phone or pda.

 

[London Geezer]

I cringe at the thought of an 8 SPE cell in a cell phone or pda.

Imagine:

The new Cell-based Sony mobile phones are capable of 9 billion calls per second! You'll never be alone again!!

 

[Titanio]

sony said in the E3 presentation CeLL inside ps3 can decode 12HD streams AT ONCE.

so according to pc-engine , a 3GHz P4 can manage 3 HD MPEG2 streams

so that makes 1 gameconsole has THE POWERZZ of a 12GHZ PENTIUYM4

if thats not impressive i dont know what it is.

and personaly i dont give a shit about this
we need a nextgen benchmark

Indeed, this thread has me laughing.

Attempt to downplay + 4x P4 performance for HD decoding = does not compute

 

[PC-Engine]

That's not very impressive at all considering a 3GHz P4 can manage 3 HD MPEG2 streams. :?

Any pointers to a test or something like that where a 3Ghz P4 actually decodes a MPEG2 HD stream and only uses 33%?

aaaaa00 got 3 streams running on a 2.5GHz P4.

 

[-tkf-]

Well, my 2.7 P4 is having a hard time doing one, and if we just say a 3.0 does 2 which is more realistic (lots depends on the Graphic Card) it would be more like 3.2Ghz Cell = 18GHZ P4!

I guess the Cell really is impressive, good find PCE

 

[_xxx_]

Stupid question: how many of you people will ever be decoding 5+ MPEG2 streams at once? :?

 

[Megadrive1988]

imagine the performance of a single, 2nd-revision Cell (250M trannies) running at ~5 GHz with all 8 APUs...er SPEs, activated. well over 300 Gflops, no?

 

[London Geezer]

imagine the performance of a single, 2nd-revision Cell (250M trannies) running at ~5 GHz with all 8 APUs...er SPEs, activated. well over 300 Gflops, no?

Probably.

 

[PC-Engine]

Well, my 2.7 P4 is having a hard time doing one, and if we just say a 3.0 does 2 which is more realistic (lots depends on the Graphic Card) it would be more like 3.2Ghz Cell = 18GHZ P4!

I guess the Cell really is impressive, good find PCE

Maybe you should get a new computer because there's something definitely wrong with it.

We're talking MPEG2 if you hadn't noticed.

 

[Megadrive1988]

Interesting... it's still pretty powerful tho, but certainly not 1000x the power of PS2 claimed 5 years ago.


I don't knnow if you doubt it, but i always take sony's claims of performance with a grain of salt. Much of what that article created by MS was false (the major nelson article first posted) but they had one thing right. Sony has a habit of over promising on performance and not living up to that promise. Still cell is powerfull enough to do many cool things.

to be fair, Sony under promised with PS2 and over delivered. I'm not talking about the rediculas comments made around PS2's unvieling, but the claims Sony made just after the release of PS1 in Japan. Ken said in 1995 that they'd get 10 million polygons per second peak performance on .25 micron silicon. Sony did beat that by 6+ times.

PS2 was a massive leap over PS1 even if PS2 did not measure up to the insane media-driven hype of 1999-2000.

PS3 is a smaller leap over PS2 than PS2 was over PS1 for a number of reasons.

maybe Sony will make a larger than PS2 ====> PS3 leap with PS4 8)

 

[Mordecaii]

Do we have any numbers for how quickly a 3 GHz P4 can decode a HD stream in software (aka without using the vid card)? Or is 2-3 streams the number for that?

 

[PC-Engine]

Do we have any numbers for how quickly a 3 GHz P4 can decode a HD stream in software (aka without using the vid card)? Or is 2-3 streams the number for that?

A 2.5GHz P4 can almost manage 3 streams all in software therefore a 3GHz P4 can definitely manage 3 in software.

 

[-tkf-]

Do we have any numbers for how quickly a 3 GHz P4 can decode a HD stream in software (aka without using the vid card)? Or is 2-3 streams the number for that?

A quick glance over at AVS and my own experience plus a question out in the office indicates 2 Mpeg2 @ 3.0 Ghz, which fits in nicely with:

http://www.opa-ages.com/?bomba=article&id=32548

As you maybe aware, a full HD MPEG2 stream decode consumes ~40% of CPU utilization on a 3.2 Ghz Pentium4

Some Graphic Cards does lesson the burden on the CPU though.

 

[McFly]

Well, my 2.7 P4 is having a hard time doing one, and if we just say a 3.0 does 2 which is more realistic (lots depends on the Graphic Card) it would be more like 3.2Ghz Cell = 18GHZ P4!

I guess the Cell really is impressive, good find PCE

Maybe you should get a new computer because there's something definitely wrong with it.

We're talking MPEG2 if you hadn't noticed.

PC-Engines TFT displays have no view-angle problems as well, guess he buys his computer parts in a different dimension where everything is better than here ... damn PC, give me the key to that dimension!

Fredi

 

[PC-Engine]

Well, my 2.7 P4 is having a hard time doing one, and if we just say a 3.0 does 2 which is more realistic (lots depends on the Graphic Card) it would be more like 3.2Ghz Cell = 18GHZ P4!

I guess the Cell really is impressive, good find PCE

Maybe you should get a new computer because there's something definitely wrong with it.

We're talking MPEG2 if you hadn't noticed.

PC-Engines TFT displays have no view-angle problems as well, guess he buys his computer parts in a different dimension where everything is better than here ... damn PC, give me the key to that dimension!

Fredi

It's not my computer it's aaaaa00's.

 

[mckmas8808]

This thread kind of makes no sense. You have guys here trying to debunk the CELL because it can do 4 times as many HD streams as a 3 GHz P4. And 1000x power was KK's target. How can you claim something thats not even built.
________
365

 

[passby]

http://pc.watch.impress.co.jp/docs/2005/0521/e304.htm

According to Kutaragi, a full HD MPEG2 stream video results in 70% utilization of single APU, which is enough to run additional audio filter.

I read it. Then I re-read it. There is no such statement in the linked article.

Request to moderators to end this thread please.

 

[PC-Engine]

http://pc.watch.impress.co.jp/docs/2005/0521/e304.htm

According to Kutaragi, a full HD MPEG2 stream video results in 70% utilization of single APU, which is enough to run additional audio filter.

I read it. Then I re-read it. There is no such statement in the linked article.

Request to moderators to end this thread please.

Read the post by one.

 

[passby]

Wrong link. I searched all over the place for it. This is the link.
http://pc.watch.impress.co.jp/docs/2005/0521/e305.htm

 

[hey69]

why am I reading your post as YODA language..

me go home , stress is bad