The cell is all hype.

[slash196]

Now, this is just imo, but I think all this talk about the Cell chip revolutionizing PCs is hype. We've heard this kind of stuff time and time again, and not once has the revolution occurred. The funny thing about revolutions: they aren't predicted. They just happen. When you say something will be a revolution, you don't actually mean it, you're just trying to generate press. Opinions?

 

[Zebo]

psst.. buy my stock so I can cash out my options.

 

[xbdestroya]

Hype is hype - every company hypes it's stuff. Just because the boy keeps crying wolf doesn't mean the wolf won't come one day. Itanium was going to change the industry, but just because it didn't, doesn't mean it couldn't have.

I think the same thing with Cell right now. Whatever the case, whatever you want to believe or not believe, this chip is pretty serious. Now, what happens to it's future next, and what happens to the industry, is kind of like a choose your own adventure book. There will be a lot of chances for a lot of company's to edge each other out or screw up. For myself, I don't think Cell will unseat x86, but I think it will do well for itself.

 

[slash196]

I think the PS3 will be a sweet system, but it's not gonna revolutionize anything, and it sure as hell won't replace the PC.

 

[91TTZ]

Remember when the PS2 came out? They said it was going to have the performance of a supercomputer, and its Emotion Engine would give characters artificial intelligence and be able to render Toy Story like animation in real time.

Then they showed some clips of its performance and it appeared to live up to the hype... until the system actually came out. Once it came out, everyone realized that those amazing video clips were merely pre-rendered scenes that the PS2 was playing like a DVD player. In real time it couldn't render anything like that.

Don't believe the hype. You can't sell a revolution, it just happens. And Sony is trying to sell a revolution.

Besides, money talks. If this chip really could deliver more performance than many multicore Athlon64's, they wouldn't be selling them for cheap and putting the chips in $300 PS3's, they'd be selling the chips for $900 a piece and using them in high-end equipment.

Expect more of a stripped down, cost efficient CPU that's similar to embedded RISC processors or GPU's that Nvidia and ATI make.

 

[xbdestroya]

91TTZ unfortunately whether you're right or you are wrong, your logic is totally whack. Besides the fact that these chips indeed ARE going to be used in high end equipment, saying don't believe that Cell is 'real' because Sony over-hyped Emotion Engine just doesn't make sense. Sony could have lost all credibility in your eyes but still be telling the truth.

It'll probably be somewhere in between. But the crazy flops they're talking about, even if it ends up being way theoretical, is still a good deal beyond what todays processors are churning out.

 

[ReiAyanami]

definetly "overhyped" but if they say its 1000x more powerful than today's PCs then all it needs is 2x or 3x more powerful to "wipe out" AMD/intel. so if it were just .003% of its hype it'd still be pretty powerful

with a collaboration of IBM + Toshiba + Sony + Rambus + NVidia u'd think they'd churn out something good

although it seems like it would need an extensive fiber and worldwide wireless network to back it up. SUPER-SETI!

 

[91TTZ]

Originally posted by: xbdestroya
91TTZ unfortunately whether you're right or you are wrong, your logic is totally whack. Besides the fact that these chips indeed ARE going to be used in high end equipment, saying don't believe that Cell is 'real' because Sony over-hyped Emotion Engine just doesn't make sense.

You sound like a young kid who lacks the real life experience to make a statement like that.

Right now all we can go by is:

1. Their press announcements about the Cell
2. The real life performance of their previous products compared to the press announcements leading up to them.

Going by their history, which is all we can go by right now, it would seem that they have a history of over-hyping and under-delivering. I expect the cell to be the same case.

I'm sure in a decade or so you'll begin to see things my way. You can only get excited over "revolutionary" product announcements so many times before you wise up and realize it's all hot air.

Remember BitBoys? Elbrus2k? Transmeta? Each of them generated so much hype in their day, but reality wasn't quite as exciting.

 

[Eug]

Yes, way overyhyped.

However, while I'm still not entirely clear what Cell is, it is still quite intriguing. It seems to be quite a change from traditional design:

Power based - 64-bit, with 2.5 MB of on-chip memory, and possibly SMT
9 units on chip - 1 Power core and 8 "synergistic processing units"
Rambus XDR. (I don't like Rambus the company, but what can you do.)
> 4 GHz at introduction (on 90 nm)

 

[xbdestroya]

Originally posted by: 91TTZ

Originally posted by: xbdestroya
91TTZ unfortunately whether you're right or you are wrong, your logic is totally whack. Besides the fact that these chips indeed ARE going to be used in high end equipment, saying don't believe that Cell is 'real' because Sony over-hyped Emotion Engine just doesn't make sense.

You sound like a young kid who lacks the real life experience to make a statement like that.

Right now all we can go by is:

1. Their press announcements about the Cell
2. The real life performance of their previous products compared to the press announcements leading up to them.

Going by their history, which is all we can go by right now, it would seem that they have a history of over-hyping and under-delivering. I expect the cell to be the same case.

I'm sure in a decade or so you'll begin to see things my way. You can only get excited over "revolutionary" product announcements so many times before you wise up and realize it's all hot air.

Remember BitBoys? Elbrus2k? Transmeta? Each of them generated so much hype in their day, but reality wasn't quite as exciting.

And you sound like someone who's insecure in their own arguments if you need to rely on a crutch like supposed age and experience to argue my points.

I'm sure Cell will not be everything Sony, IBM, and Toshiba claim it to be. I'm also sure it will be a lot more than you are giving it room to be. Let me give you a basic example. How many computers has Transmeta gotten their chips into? How many PS3s will Sony get their Cell chip into? How many cell phones? How many TV's? How many Blu-Ray players? Etc, etc, etc...

Listen, Cell already has the base for it's 'sales' installed. It already IS a success - even before launch; it's destiny is sealed. It WILL survive this generation, and if there are shortcomings, they will be ironed out for round two. I'm no zombie, but nor am I a diehard cynic. Take what you will from what I'm saying. If you are as mature as you claim, you will recognize the err of your previous logic.

 

[91TTZ]

That article is interesting, although it does seem like they're trying to promise everything with this chip.

"While the PlayStation 3 is likely to be the first mass-market product to use Cell, the chip's designers have said the flexible architecture means that Cell will be useful for a wide range of applications, from servers to cell phones. Initial devices are unlikely to be any smaller than a game console, however--the first version of the Cell will run hot enough to require a cooling fan, Kahle said."

To me, this seems unrealistic. The type of chip you would want in a server would not be very well suited to be used in a cell phone, even if it's a cut down variant of it. They say it requires a heat sink, but they think it'll work in a cell phone? When you design a CPU you have to make tradeoffs. Do you want the most speed possible at the expense of heat/power consumption? Or do you want an efficient chip that runs cool and doesn't use much power? Even Intel uses different chips for their laptops, and that's still a very power hungry device compared to a cell phone.

I'm sure the chip was designed with something in mind, so I don't really think it would excel at both extremes.

 

[91TTZ]

Originally posted by: xbdestroya

I'm sure Cell will not be everything Sony, IBM, and Toshiba claim it to be. I'm also sure it will be a lot more than you are giving it room to be.

I will define what I think it will turn out to be just so we understand my position. Regardless of the hype about the Cell chip being put in the PS3, with its claims of being 4x as fast as an Opteron, I do not think this will be the case. I think the Cell will compare with low end budget processors from Intel and AMD, much like the CPU's in the current batch of consoles compared to PC processors at the time they were introduced.

The Blachford.info article states that "Intel could put 10 cores on a chip and they'll match neither it's performance or price.". I think this is completely false. Not only will the Cell not be able to match the performance of a 10 core Intel chip, it will not even match the performance of the current high end P4's.

Another one of the claims: "Even with a single Cell it will outgun top end multiprocessor PCs many times over. That's gotta hurt, and it will hurt, Cell is going to effectively make traditional general purpose microprocessors obsolete."

Nonsense.

And the last line of drivel I'll comment on:

"Cell represents the largest threat the PC has ever faced. The PC can't use it's traditional advantage of software because the Cell can run the same software."

I think once the Cell comes out and falls short, we'll all have a good laugh about these bogus claims from enthusiastic idiots. Also, the Cell will not be able to run the same software that PC's can because it's not using x86 architecture. It would have to run emulation which is always slow.

To sum it up- Lots of hype, little substance. When it comes out we'll think it's a neat little CPU but doesn't measure up to the high end CPU's from Intel or AMD.

 

[bobsmith1492]

Cell looks like a GPU. I'm sure Nvidia and ATI's latest stuff would stomp all over Cell.

 

[Schmeh]

Originally posted by: bobsmith1492
Cell looks like a GPU. I'm sure Nvidia and ATI's latest stuff would stomp all over Cell.

The Cell is not a gpu. NVidia is designing the GPU for the Playstation 3.[/quote]

Originally posted by: 91TTZ
I will define what I think it will turn out to be just so we understand my position. Regardless of the hype about the Cell chip being put in the PS3, with its claims of being 4x as fast as an Opteron, I do not think this will be the case. I think the Cell will compare with low end budget processors from Intel and AMD, much like the CPU's in the current batch of consoles compared to PC processors at the time they were introduced.

The Blachford.info article states that "Intel could put 10 cores on a chip and they'll match neither it's performance or price.". I think this is completely false. Not only will the Cell not be able to match the performance of a 10 core Intel chip, it will not even match the performance of the current high end P4's.

According to CNet News.com, the cell, which is a cpu with a 64-bit Power Processor and 8 "synergistic processing units" can perform 256GFLOPS. By comparison a Pentium 4 3Ghz chip can perform 6GFLOPS. Yes, the Cell can not only meet the performance of Intel and AMD's CPUs but it can outperform them by quite a large margin.

Now this is not to say that the Cell will take over the computer industry, but I do think that since it takes a couple hundered Opterons, Xeons, or Itaniums to get 1TFLOP and it would only take 8-15 Cells to get the same performance, the cell has a great chance at being dominant in the Super Computer market, as well as the workstation market for graphical development.

 

[MisterChief]

Now here's a good question: HOW MUCH WILL IT COST?! With the kind of performance quoted above, the damn thing should cost about as much as a PC!

EDIT: AND TELL ME YOU KNOW THE CELL IS A GROUP OF PROCESSORS!!! If you need proof, I'll scan the article. You can probably find the CELL DIAGRAM on Sony's site. Or not...

 

[zakee00]

According to CNet News.com, the cell, which is a cpu with a 64-bit Power Processor and 8 "synergistic processing units" can perform 256GFLOPS. By comparison a Pentium 4 3Ghz chip can perform 6GFLOPS. Yes, the Cell can not only meet the performance of Intel and AMD's CPUs but it can outperform them by quite a large margin.

Now this is not to say that the Cell will take over the computer industry, but I do think that since it takes a couple hundered Opterons, Xeons, or Itaniums to get 1TFLOP and it would only take 8-15 Cells to get the same performance, the cell has a great chance at being dominant in the Super Computer market, as well as the workstation market for graphical development.

total BS, they is simply no way that it can hit 256GFLOPS. there is obviously a huge piece of info we are missing. if it was right now physicly possible to make a chip that did that you would be hearing alot more about it.
cell is diffenately gunna kick alot of ass, but its just another CPU. that just means that games will be limited by Nvidias GPU that they put in the PS3. Big deal.
Nick

 

[JWade]

another processor that was suppose to revolutionize the x86 world is/was the Rise-MP6 processor. socket 7. i have seen 200mhz to 266mhz ones, not sure about any other speeds. it like other hype as already stated was more hype than anything else.

 

[imported_kouch]

if cell was as powerful as they claim, why would you need a GPU to run the graphics code for PS3. I mean something 1000x powerful then a dual opteron server could easily emulate directX or openGL faster than a 6800 could run it. LOL, it is able to emulate all the current PC software faster then any pc chip every made, but it can't even emulate Direct X for a petty console.

 

[zakee00]

i just dont see what the big deal is as far as gaming goes, games are not often CPU limited right now (providing you have a nice graphics card and run with eye candy). i would rather have some kind of cell-based graphics card (or any sort of huge performance jump). that would be more useful IMO; higher AA levels, able to play UE3 based games, etc.
i bet UE3 will work fine with my processor, but i know for sure it will rape my graphics card.
not sure if that made alot of sense,
Nick

 

[Schmeh]

Originally posted by: zakee00

According to CNet News.com, the cell, which is a cpu with a 64-bit Power Processor and 8 "synergistic processing units" can perform 256GFLOPS. By comparison a Pentium 4 3Ghz chip can perform 6GFLOPS. Yes, the Cell can not only meet the performance of Intel and AMD's CPUs but it can outperform them by quite a large margin.

Now this is not to say that the Cell will take over the computer industry, but I do think that since it takes a couple hundered Opterons, Xeons, or Itaniums to get 1TFLOP and it would only take 8-15 Cells to get the same performance, the cell has a great chance at being dominant in the Super Computer market, as well as the workstation market for graphical development.

total BS, they is simply no way that it can hit 256GFLOPS. there is obviously a huge piece of info we are missing. if it was right now physicly possible to make a chip that did that you would be hearing alot more about it.
cell is diffenately gunna kick alot of ass, but its just another CPU. that just means that games will be limited by Nvidias GPU that they put in the PS3. Big deal.
Nick

The reason we aren't hearing "that much about it" is that today was the first time IBM, Sony or Toshiba has officially talked about it in public. They did just that at ISSCC today and will be saying more in the next couple of days as the conference wraps up.

 

[zakee00]

Originally posted by: kouch
if cell was as powerful as they claim, why would you need a GPU to run the graphics code for PS3. I mean something 1000x powerful then a dual opteron server could easily emulate directX or openGL faster than a 6800 could run it. LOL, it is able to emulate all the current PC software faster then any pc chip every made, but it can't even emulate Direct X for a petty console.

agreed

 

[Zebo]

Yup all you need to buy is a CELL CPU with software emulated everything, modem, nic, video card, controllers etc etc etc and it would still be 500X faster than top of line machines today costing $4000.

"revolutionize" indeed.

 

[zakee00]

The reason we aren't hearing "that much about it" is that today was the first time IBM, Sony or Toshiba has officially talked about it in public. They did just that at ISSCC today and will be saying more in the next couple of days as the conference wraps up.

ok lets put it like this:
cell is not going to run at 256GFLOPS. period. maybe "theoriticaly" or something, but not real world. cell is 75% hype.

 

[imported_kouch]

Its funny how people say it can run as many as 10 threads at a time compared to only 2 by a p4. Don't make me laugh, how many applications have 10 simultaneous threads running. Parellel computing is a good concept but just very hard to implement since the human mind and the physical world are sequential entities (i.e. you do not know the future values of variables so you have to wait till you get them before you can proceed and that a lot of times depends on user input etc.)

 

[Schmeh]

Originally posted by: zakee00

The reason we aren't hearing "that much about it" is that today was the first time IBM, Sony or Toshiba has officially talked about it in public. They did just that at ISSCC today and will be saying more in the next couple of days as the conference wraps up.

ok lets put it like this:
cell is not going to run at 256GFLOPS. period. maybe "theoriticaly" or something, but not real world. cell is 75% hype.

Of course it will not hit 256GFLOPS in real world applications, that is the theoritical peak. But even if it only hits half of that, 128GFLOPS, it will still be far more powerful than any cpu out right now.