Congratulations to Vince
- Thread starter bbot
- Start date
- Status
marconelly!
Veteran
Hmm, I might be wrong, but wasn't Vince actually predicting a 1TFLOPS Cell? 
Or was that for four PE Cell configuration?
Or was that for four PE Cell configuration?
Well Vince might not know much history or politics, even I sometimes half-way respect his technical knowledge.
Congrats if ya got it right Vince, two points. 8)
Congrats if ya got it right Vince, two points. 8)
http://www.beyond3d.com/forum/viewtopic.php?t=20202
Vince was not the only one predicting this.
Btw, I have seen the same people argue 256GFLOPs, 512GFLOPs, 1TFLOP, etc... for a single CELL. e.g. "I am predicting 256GFLOPs, but I would not be surprised if STI surprised us all with 1TFLOP." Did Vince state unequivocally that CELL will absolutely be 256GFLOPs?
Also Brimstone, version, Shifty Geezer, Titanio, Mythos, and jvd all went on record for 256GFLOPs. Congratulations! Your winners prize is a single free post at B3D—use it wisely!
Vince was not the only one predicting this.
Btw, I have seen the same people argue 256GFLOPs, 512GFLOPs, 1TFLOP, etc... for a single CELL. e.g. "I am predicting 256GFLOPs, but I would not be surprised if STI surprised us all with 1TFLOP." Did Vince state unequivocally that CELL will absolutely be 256GFLOPs?
Also Brimstone, version, Shifty Geezer, Titanio, Mythos, and jvd all went on record for 256GFLOPs. Congratulations! Your winners prize is a single free post at B3D—use it wisely!
It's Broadband Engine that everyone has been predicting to be a TFLOPS, this "prototype" shown today is just a PE.
I find it funny how the patent(and it's a few years old by now) was still on point about the structure and even about many of the performance figures of the prototype revealed today....
I find it funny how the patent(and it's a few years old by now) was still on point about the structure and even about many of the performance figures of the prototype revealed today....
Hmm, I might be wrong, but wasn't Vince actually predicting a 1TFLOPS Cell?
No one predicted that 1 PE would do 1 TFLOPS, no one predict anything on performance, Vince like other people here just went by the patent, the 1 PE configuration is 256 GFLOPS. And today that's pretty official
(Well until I can see the paper myself)Though many were having trouble with the 4 GHz figure.
Anyway, what we were predicting though, how big the beast will be, and whether or not its possible to put 4 PE and 64MB of eDRAM on a single chip using 65nm process. Looking at what's reveal today, the size of Cells and its transistors count as well as how memory heavy the chip is, Single chip Broadband Engine should be possible on 65nm, though it'll be a server class chip, than what you would expect on PS3. My estimate was around 400mm2, if I remember right. And with today info, I would still go with that ballpark figure, for this imaginary 65nm BE.
But hearing about report on power dissapation of Cell chip, single chip BE doesn't look too promising.
Actually, almost everybody from day one agreed on the "fact" that a PE (Or SPE as they seem to call it know) was around 256Gflops. People only argued about the number of PE that will be in the PS3's Cell CPU (1 = 256Gflops, and 4 = 1 Teraflops).Acert93 said:
Edit: Owned by V3!
Wahay! A free post! A free post, a free post. Tra-la-laa, a free post. This is great. I wonder what I'll write. I don't want to waste it. I'd better...hang on - I've just posted this! Doh!Acert93 said:
Spidermate
Regular
Where is Deadmeat now? I remember him predicting the entire architectect only supporting somewhere around 70 GFLOPS with a ridiculous clock speed?
jvd said:
I'm not totally sure he deserves a full apology. Just in my searching he estimated that the PS3 CPU would only have 1PE/8APU design running at 2GHz and having 128 GLFOPS. We're still not sure of the actual PS3 chip yet. I do find it interesting that he was pretty close. Had he suggested 4Ghz, it's possible more people might had taken him seriously. <shrugs>
Tommy McClain
Vysez said:
Not "almost everybody from day one". Many? Maybe. "Almost everybody"? No. Even the recent predictions thread proves this statement wrong.
Btw, if "almost everybody" was claiming CELL would be 256GFLOPs, what is the point of this thread?
Ahhh yes, it is a trolling thread. All I was trying to do was take the edge off the "Vince vs. the World" arguement before the forum broke down into chaos Edit: Owned by V3!
Hello? I never stated Vince said CELL would be 1TFLOPs so I have no idea why you are saying I got owned. I see this thread is going to self destruct.....
Oh shaddap you. Deadmeat had a new version of "Cell" (which still is a fucking ARCHITECTURE and not a processor) for every month of the calendar. Sometimes more. His rock bottom "prediction" was that "Cell" would have lower flops performance than the current PS2. Well, if you count pulling stuff out of his arse as predictions.jvd said:
He predicted cell chips with 2 PUs and 9 APUs and 1PU with 4 APUs running at >1GHz (because 1GHz or god forbid more would make it burn up) and all kinds of crazy stuff.
You're nuts if you think anything that man "predicted" turned out to be true because of anything other than sheer coincidence. Even a blind hen, and all that you know...
Unlike stupid microsoft who went with ATi, oh the IDIOTS?
marconelly!
Veteran
Haha, congrats to Vince and everyone else who made the correct prediction and stood by it.
Deadmeat posts at Opa-Age, if anyone still wants to read his words of wisdom. But, from what I've seen, he's become worse than ever, and I'll just leave it at that. (his last prediction was that Cell will be 64GFLOPS, btw)
Deadmeat posts at Opa-Age, if anyone still wants to read his words of wisdom. But, from what I've seen, he's become worse than ever, and I'll just leave it at that. (his last prediction was that Cell will be 64GFLOPS, btw)
- Status
Similar threads
