Does Cell Have Any Other Advantages Over XCPU Other Than FLOPS?

[Dave Baumann]

Understand that there is a difference with state and convey.

I was wondering how long it would take you to start participating in the console forum again, and the subsequent length of time it would take you to freely start throwing insults around; I wasn't really very long. I'd suggest you try and curb your instincts if you wish to keep playing.

 

[Vince]

Understand that there is a difference with state and convey.

Most certainly, just as there is between a literal reading and an interpretation. You can only take from these forums what is written, that is all the resolution it allows; what you are doing is adding your own bias, and I can not be fauted for what is in your mind... and what I did not write.

and the subsequent length of time it would take you to freely start throwing insults around

My appologies then. I was merely following Aaron's stated interpretation of what an insult is:

"An insult generally doesn't have truth behind it. Version IS a waste of time." - AaronSpink​

I do suppose that just because you don't threaten or correct him doesn't mean he is, in fact, correct. I won't let it happen again Dave.

 

[nelg]

The point is it can do things that others can't.

No it proves Toshiba will write software that no one else would bother with except Cover Girl.

 

[Dave Baumann]

My appologies then. I was merely following Aaron's stated interpretation of what an insult is:

No, Vince, you were reverting to type. Don't let it happen again.

 

[Vince]

No, Vince, you were reverting to type

I don't quite follow, it may have been lost in translation coming across the pond.

Edit: Nevermind, I comprehend now. And no; I did it for the exact reason which I stated, which would make sense if you'd stop to think about it. Again I'd appreciate it if you'd stop inferring things I did not state, or assigning motive to people you do not know. Converse to questioning my "type," am I to take it from your actions that your threatening isn't equivalent or fairly applied to all?

 

[expletive]

Great but, atleast I, really don't care; I'm sorry.

1. Youre not sorry
2. I figured out you dont care about anyones contrary opinion about 20 posts ago
3. The fact that you dont care about others opnions seems to be at the root of why you get so much flak around here

If you're so intent on discussing it then I suggest you start another thread in the Hardware Forum and discuss DSPs in CE applications there; but in this thread we were discussing possible uses (eg. brainstorming) of Cell in CE applications.

Actually if you want to be a stickler the thread is about advantages of Cell over XCPU. We got on this tangent by discussing advantages of the design overall in other markets in which it will compete. By your logic, we should each start another thread.

This is a technology forum, we discuss technology; not marketplace economics. If we want to dream up potential uses then that's allowable concidering the thread we're in; just like it's fine to talk of how ATIs GPUs kickass in GPGPU work without having somone come in and talk about how nVidia's are better priced.

You can get off the 'this is a tech forum' and 'WE discuss technology here' high horse any time now. No one else seems to have a problem with, and actually most participate in, discussions on the application of Cell, and all the technology, in the context of the real world. Theres cpuntless threads in this forum that talk about developement costs, timelines, chip yields, etc - all REAL WORLD constraints. IMO, pontificating on the possible uses for a processor with no regard for reality is somewhat wasteful since youre tlaking about things that potentially have ZERO chance of happening for reasons we must not discuss !

As I stated before, there is no need for your fallicious goal of balancing supposed hype: reciprocation where none is necessary. Several posters truely act like teenage boys talking about their birthday gifts; ever escalating in intensity and importance: Yeah, but mine has this and that... yours can't do this. Uhh, great. That doesn't change what I have, I like it, it works well, so stop downplaying it already...

I'm not even sure what youre attempting to say here. I'm not balancing anything, i just enjoy discussions that engage different peoples perspectives to try and get to things we can agree on as truth and fact. We may not come up with asnwers but we can try to come up with the framework (facts, observations, hypotheses) in which answers can be discerned at a later date.

Well Vince you certainly are interesting if nothing else. Honestly, I cant tell if youre 15 year-old with a 30 year old vocabulary or a 30 year-old with 15 year-old maturity. But it takes all kinds to make the world go round.

 

[expletive]

And THAT expletive is my and Vince's point the whole time. I and Vince have never made the CELL look like the holy grail of processors. Just that it might or should be able to give the consumer something that is not available to them now.

And that example of Vince's that even you say you're sold on is the point.

Ok you guys win, if we get to the point where theres "Minority Report" style, holographic, targeted advertising on every street corner and the reason we got there is becuase of teh cell, then i will return to admit i am sold on its ultimate, indisputable, power.

Seriously though, i dont think you ever made it out to be the holy grail, and while he may have never used the words 'holy grail', thats the impression one can imply from vince occasionally.

Yes it might be able to give the consumer something that is not available to them now, but what that 'something' is, is still anyones guess. As far as i can tell, that "something" is not even on the horizon, its cooking in some lab or even just jotted down on paper at this point. I just dont know that when that 'something' is unleashed in the CE space, will the Cell processor be the only thing that can deliver it?

I think the problem with a processor with SO much potential is that we start to dream up uses for it, rather than finding out what people NEED and WANT first.

 

[Vince]

By your logic, we should each start another thread.

I don't disagree, this problem with threads has come up many times. Yet there is a logical jump from Cell|XCPU advantages to how Cell's redundancy and computational power can be used in the CE market; especially seeing as though Toshiba's publically stated it'll be used in this role. I don't quite see the logical jump to how DSPs and cost come about outside of trying to diminish Cell -- at the expense of even considering the XCPU!!

IMO, pontificating on the possible uses for a processor with no regard for reality is somewhat wasteful since youre tlaking about things that potentially have ZERO chance of happening for reasons we must not discuss !

Namely what is impossible? Everything I've stated has been demonstrated on Cell or is based off something that has been. Tadashi Okamura, President of Toshiba I think, is the one who has stated their wish to use it in CE equiptment in 2006/2007, how does it have 0 chance of sucess?

I'm not even sure what youre attempting to say here.

That not every post needs a counter-post; yet everytime someone states something positive for Cell, as somone initially did about using it in CE, we have people who feel compelled to dismiss it for reasons that go beyond the thread. The discussion was on Cell's advantages over the XCPU, one of which is redundancy and it's trickle-down use... how did DSPs, ARM and ASIC pricing get into the conversation?

You, yourself stated:

"No, i think what the debate is surrounding is the propagation of hype surrounding what the processor will mean to CE."​

How does that tie in... except that some feel the need to put up a counter-argument on Cell in CE? As mckmas8808 said: Somebody finally said it. Some people feel like they have to fight to their death to prove that CELL is only good for the PS3 and for some reason Mercury Systems.

And thanks for finding me interesting.

 

[expletive]

"No, i think what the debate is surrounding is the propagation of hype surrounding what the processor will mean to CE."​

How does that tie in... except that some feel the need to put up a counter-argument on Cell in CE? As mckmas8808 said: Somebody finally said it. Some people feel like they have to fight to their death to prove that CELL is only good for the PS3 and for some reason Mercury Systems.

And thanks for finding me interesting.

IMO hype, especially in regard to CE, can be construed as claims that are neither technically nor practically possible or feasible. Imo, thats what we were debating; Whether or not yours (or STI's) claims were either technically possible or even practical (both can be equally important but that gets us into the whole 'real-world' problem again).

I dont look at these threads as an exercise in one-upmanship. If ive got something else to say, i respond. If not, i dont or occassionally drop a simple 'agreed' post. Dont take it any other way than that, i'm not concerned with getting the last word or feel compelled to make a counter argument for its sake. This discussion has happened in threads before and ive had the same stance then as I do now, it really is just my opinion, not someting i made up to undermine the potential of some chip.

 

[mckmas8808]

I think the problem with a processor with SO much potential is that we start to dream up uses for it, rather than finding out what people NEED and WANT first.

How do you find out what people want (a company like Sony) and then tell a forum like B3D about it? Obviously the Japanese people at that show liked the CELL made makeup and hair application.

Do consumers have to beg for it on internet chat rooms for CE companies like Toshiba or Sony to make it first? How many people wouldn't want what Vince examples explained a page or two back? It's the price that would be the most troubling. But if the consumers start to pick up on it (demand raises) then the price ought to come down to a reasonable price.

 

[ihamoitc2005]

Strange urge

How do you find out what people want (a company like Sony) and then tell a forum like B3D about it? Obviously the Japanese people at that show liked the CELL made makeup and hair application.

Do consumers have to beg for it on internet chat rooms for CE companies like Toshiba or Sony to make it first? How many people wouldn't want what Vince examples explained a page or two back? It's the price that would be the most troubling. But if the consumers start to pick up on it (demand raises) then the price ought to come down to a reasonable price.

It is usually not possible to know what is wanted before it is sold and sometimes it is impossible because consumer cannot imagine such products to know if they desire it. For example, no regular consumers asked for computers and most said who would want a computer in the house, but now almost everyone has a computer in the house and many have more than one.

But people without vision will not understand this. They will only look for market surveys and make boring products based on surveys which is always inferior to product created and designed by imagination.

 

[mckmas8808]

But people without vision will not understand this. They will only look for market surveys and make boring products based on surveys which is always inferior to product created and designed by imagination.

Thankyou and this 100% backs how me and Vince are thinking about the possibilities of the CELL chip in CE products. The world doesn't have to be screaming for it for companies to use it. All we need are forward thinking people that know how to use technology to their advantage.

It takes vision people, VISION.

 

[nAo]

I was wondering how long it would take you to start participating in the console forum again, and the subsequent length of time it would take you to freely start throwing insults around; I wasn't really very long. I'd suggest you try and curb your instincts if you wish to keep playing.

I agree with you Dave, this should be a place where people can discuss without being insulted.
It would be also very appreciated, and I'm sure I'm not alone on this, if the next time something like that happens you could publicly give this kind of advices even to other people, not just Vince, otherwise it should be better to keep things private. (that's why we have mods and PMs, right? )
To put it straight last time I checked working in some kind of industry doesn't give to anyone any license to freely insult here, imho.
I'm sure you know what I'm talking about and that you understand the issue here.

Thank you,
Marco

 

[aaronspink]

Is it possible that with the right ideas, the cost factor can be mitigated ? Think Apple.

Apple actually has very very low costs in their CE devices. Most of their price is because they can charge a premium related to their very good user interfaces and industrial design.

Secondly, and this is a "thinking out loud" question: As digital content catches on, would the content providers including Sony be interested to invest, control and deploy a pervasive, strong security CE platform ? Something locked deep inside a processor, that can be upgraded safely and on-demand (cannot rely on hardwired logic), plus uniform so that it can be dropped onto all content devices (reduce weakest links in security)... various incarnation of cells perhaps ? Is this possible ?

To do secure digital content, all you need is a hypervisor like system and a cryptographically secure signature engine. These technologies aren't very costly right now and are being integrated in a great many devices. The problem is that they still have a lot of flaws and ability to backdoor related to the fact that the actual communication isn't taking place between two trusted sources but one trusted source and one untrusted source. Its pretty much an NP complete problem, all the content vendors can do is make it harder to crack, but they can never make it impossible to crack.

And don't get me started on the general consumer issues of DRM. Realistically the death of DRM won't be technical, but will be social.

Aaron Spink
speaking for myself inc.

 

[LunchBox]

it's a bit late but since it was addressed to me...

Didn't we know that if the PPC is faulty the entire chip must be binned? So, you need this mighty PPC sucking juice up which is fine in some CE applications but not all. The areas where you can get away with it are where FPGAs can be used, the areas you can't are where ARM has the market tied up.

ummm... i did state that it should be atleast "functional"...
and i assumed that my prior statement would have been interpreted as "to applicable purposes..."
if i gave a slight immpression that i was hinting otherwise, then i apologize.
i'm not downplaying the use of ARM/dsp...
what i'm saying is that if they can use it for other things instead of tossing it out coz they didn't make the cut...
why shouldn't they?

I agree that the idea of a scalable multi purpose Cell architecture could be a good thing for Sony. I think what we're debating is if the Cell we provide any benefit to the consumer.

businesses, in general, don't really think of benefits for the consumers...
they mostly think of how to make the consumer want the product and how to make consumers think that it would benefit them...
it's all about the money they can milk out of the consumers...
if their product or decision ends up benefiting the consumer then it's a great marketing plus for them
coz it makes them look like they actually care...

 

[Shifty Geezer]

OMG! Somebody finally said it. Some people feel like they have to fight to their death to prove that CELL is only good for the PS3 and for some reason Mercury Systems. Why is it that Mercury Systems is the only company that would want the CELL processor?

Only Mercury Systems can charge insane amounts of money for their products. The cost of Cell isn't an issue for them.

Both points in this thread are on the whole valid. Cell can empower more stuff, but will cost more to use. Will the features it adds be worthwhile? And really no-one's got any absolute answers (which is why the forum rules are state you opinions but don't keep arguing them over and over - you never change anyone's mind this way). Perhaps Sony and Toshiba have a product roadmap where they know the features they want to add and are looking to leverage Cell as a high end device pusher? Perhaps something new we'd never think of? Perhaps they have a dream but haven't thought it through well enough, or are taking a gamble, and it doesn't work at, with Cell remaining a fairly niche IC instead of being all pervading? Neither outcome can be proven through discussion of the pros and cons, only people's expected outcomes to be stated. Reiterating the versatility of Cell or the cheapness of commodity components isn't providing anything new, so surely the discussion closed?

 

[Lord Darkblade]

The Cell doesn't have to be that expensive... there is a lot of redundancy on the boards, imagine a 1:4 (thats nearly half the area of a chip dead!) being used in a DVD player, these chips are pretty useless elsewhere (servers, big business etc wants 1:6+ at the moment) and would drop the cost of the wafer if these are "productive chips", heck even a 1:2 cell, the main processor can run the system, the SPEs decode the video streams etc... its probably a fairly usable chip in low end technology.

 

[Shifty Geezer]

That's still a large and expensive chip though. I can't remember the figures, but isn't it something like 500 Cell processors on a 300mm wafer? Out of a run of 1:8 Cells, most should be 1:6's and above. There might be a few more-disabled Cells but they're going to be rare. So using offcuts from the 1:8 yields isn't going to produce many. By comparison a 1:2 Cell manufactured will get maybe 4x as many on a wafer, which is a quarter the price of a 1:8 run. And by comparison, if an ASIC can do the same task in a processor a quarter the size of a 1:2 Cell, that'll be a quarter of the price too. Comparing prices of Cell really needs hard figures for alternative prices. I've no idea whether an ASIC or FPGA that can do an HD decode is $1, $5 or $25, so I can't guess how a 1:2 Cell at 1.5-2 GHz would compare. I think such a Cell would cost about $25, if we accept estimates of a 1:8 being >$100.

 

[Fox5]

Apple actually has very very low costs in their CE devices. Most of their price is because they can charge a premium related to their very good user interfaces and industrial design.




To do secure digital content, all you need is a hypervisor like system and a cryptographically secure signature engine. These technologies aren't very costly right now and are being integrated in a great many devices. The problem is that they still have a lot of flaws and ability to backdoor related to the fact that the actual communication isn't taking place between two trusted sources but one trusted source and one untrusted source. Its pretty much an NP complete problem, all the content vendors can do is make it harder to crack, but they can never make it impossible to crack.

And don't get me started on the general consumer issues of DRM. Realistically the death of DRM won't be technical, but will be social.

Aaron Spink
speaking for myself inc.

If Apple's computers are included under CE devices, they may just be expensive because the chips are a low volume part for IBM. (I don't think apple uses PPC chips in anything besides their computers)

its probably a fairly usable chip in low end technology.

A 1:0 cell is probably still overkill for most low end applications. According to a brief from philips years ago about their interpolation chip, it had about the performance of a 500mhz pentium 2. At 3.2ghz, a Cell is probably far beyond that even without SPEs.

 

[Dave Baumann]

nAo, Vince has consistently displayed the same traits and tendencies, and despite been frequently warned and ultimately removed from the forum, even after a “cooling off periodâ€￾ he still appears to want to treat people that disagree with his views in the same manner; that is a warning that he will be removed again if it continues.

While Aaron displays a confrontational and sometimes abrasive character, abuse does not appear to be the consistent element – from what I’ve read of this thread that only manifested after people have misunderstood what he’s trying to convey, possibly because others don’t necessarily like what he putting across, regardless of whether it is accurate or not. I think people have learnt a lot more some of Aaron’s points than 90% of the arguments that occur here.