Blogcast Audio of Tech Engineers(?) on PS3 vs Xbox360
- Thread starter Megadrive1988
- Start date
- Status
Reminds me of another forum in which the thread author thought he'd start a simple conversation, not very contentious at all. Came back a day later had one look at the mayhem he'd unleashed and typed one simple line with a picture.
"Dance my puppets!"
But I digress, some points were presented fairly well, it IS a homogeneous programming environment that's less complex, but they utterly ignored a lot of the ways SPE's operate. They don't sit around all day being spoonfed and churning out answers like simple execution units in a core, they're capble of running mini kernels, operating directly on a shared memory space, transfers between them and chained up for larger tasks. It's more complex, but potentially the reward is there.
On the PPE core, it's dual thread, so is the each core of Xenon, but both are going to take branch prediction hits, I think Cell just has more redundancy threadwise that way, the PPE core can trip up once in awhile but it does not necessarily have to affect what the SPE's are doing. Both chips will be capable of a lot, its just up to how they're managed.
"Dance my puppets!"
But I digress, some points were presented fairly well, it IS a homogeneous programming environment that's less complex, but they utterly ignored a lot of the ways SPE's operate. They don't sit around all day being spoonfed and churning out answers like simple execution units in a core, they're capble of running mini kernels, operating directly on a shared memory space, transfers between them and chained up for larger tasks. It's more complex, but potentially the reward is there.
On the PPE core, it's dual thread, so is the each core of Xenon, but both are going to take branch prediction hits, I think Cell just has more redundancy threadwise that way, the PPE core can trip up once in awhile but it does not necessarily have to affect what the SPE's are doing. Both chips will be capable of a lot, its just up to how they're managed.
Still waiting for the 256GB/s to main memory explanation...
PC-Engine said:Get real man. Comparing 6.2 GFLOPS of EE to 3 GFLOPS of P3 isn't misleading???
Yes it is, I've said it is. Your smiley here makes me wonder if I'm taking you too seriously - I'm just saying there are differing degrees of spinning your specs. And whilst some will just see this as more of the same, I definitely feel that NV FLOPS and shader ops are on the unacceptable side of spec spinning.
PC-Engine said:Andy said:PC-Engine said:Get real man. Comparing 6.2 GFLOPS of EE to 3 GFLOPS of P3 isn't misleading???
Saying the Nvidia GPU in XBox did 80GFlops isn't either?
It doesn't matter if they used 80 GFLOPS or 40 GFLOPS. The fact is Xbox came out after PS2 and the fact it's more powerful when adding the TnL power of the GPU with the CPU.
The GCN had a total of IIRC 11 GFLOPS so the PS2 was the weekest of the bunch when talking about GFLOPS, but they kept on comparing CPU only GFLOPS to mislead people to believe it was more powerful. Xbox is more powerful so it doesn't matter what numbers were used.
That's all well and good, but we were talking about misleading figures here, and not about which consoles does more GFlops. It is easy to say that PS2 is the weakest console of the 3, but then again it also came out a lot earlier than either XBox or Gamecube, so youd expect them to be superior to PS2. Now that we got 2 consoles launching within a timeframe of 6 months or less, you are gonna expect that they will be somewhat similar, both having their strengths and weakness, and both having areas that are superior and inferior to each other.
What you've quoted, I agree with. I think their design is well balanced. What I said I didn't agree with...PC-Engine said:
Remember the discussion when MS announced the 1 teraflop targetted system performance? No-one took it as valid. So I guess we were all wrong as Xenos has MORE than that, because one teraflop was a 'very conservative figure'?
Both massaged the numbers. It'd be nice if MS didn't then try to fob those numbers off as 'very conservative', unless they can provide figures to back that idea up.
For the record, I don't believe Sony's nonsense of 2 teraflops either. They're both prattling like schoolboys in a playground over who's dad has the better job/better car/whatever.
And as I said, I liked this quote of theirs...
gmoran said:PC-Engine said:Get real man. Comparing 6.2 GFLOPS of EE to 3 GFLOPS of P3 isn't misleading???
Yes it is, I've said it is. Your smiley here makes me wonder if I'm taking you too seriously - I'm just saying there are differing degrees of spinning your specs. And whilst some will just see this as more of the same, I definitely feel that NV FLOPS and shader ops are on the unacceptable side of spec spinning.
But it doesn't matter because Xbox is more powerful than PS2. PS2 is not more powerful than Xbox so why pretend your CPU's 6.2GFLOPS vs your competiton's P3's 3GFLOPS makes your PS2 more powerful???
It would be totally different if Xbox was less powerful, but it's not.
PC-Engine said:DiGuru said:PC-Engine, did anyone ever tell you, that your sig doesn't make any sense whatsoever?
My sig is supposed to be funny aka joke, not to be taken seriously and not to be totally true.
Obviously you don't get the joke while others do.
You mean, that you cannot overclock a quantum computer at all, because all calculations are done simultaneously, at the same time? That it has no clockspeed at all, only an amount of simultaneous units and a datasize? Or that the existing prototypes happily run at room temperature, inside test tubes?
You are totally right, I don't get it.
First it's assuming it can be overclocked which it can't since there is not clock to begin with.
Second it wouldn't be stable at anything above above 0K since it's a quantum computer which only works at that temperature.
Third, since overclocking and stability is senstive to heat and all, it wouldn't be possible even if it was overclockable which it's not.
Second it wouldn't be stable at anything above above 0K since it's a quantum computer which only works at that temperature.
Third, since overclocking and stability is senstive to heat and all, it wouldn't be possible even if it was overclockable which it's not.
PC-Engine said:
DUH, sorry I thought you meant the original specs Sony put out that compared PS2 to a Pentium PC. I didn't realise you were on about some comparison to the Xbox.
I don't recall that as a published spec, however I don't doubt some sony exec would have said it at some point. But it is still presentation rather than a lie, it is not entirely disengenuous to compare a CPU to another, even if the comparison isn't entirely fair. It's rather like how Streve Ballmer says cell can't be more powerfull than XB360's as cell only has one core, he's not exactly lying, its about presentation. And given the fact that MS's specs when they launched XB were as disengenuous as you could get (300M micro polygons/sec, 150M polys/sec, fillrate 4G pixels/sec), I don't see why you are so hung up on it.
The truth is that MS chose the weapons for this match: TFLOPS, Shader Ops; not Sony. I'm not happy that Sony are following suit, but I don't see how you lay it all back on Sony's door step.
I think MS hit the nail on the head with this odd viral marketing campaign. They're pandering info to 1st semester CS students and internet pseduo-geeks who are going to get all hyped up about misinformation they have no educational or experiential knowledge to disassemble for validity. Thus, they go off and spout it to their friends and parents who treat them as tech experts and that sells units.
Brilliant marketing, but highly deceptive on MS's part. Beside, if Major Nelson's quotes really had any validity, wouldn't MS post it somewhere on MSDN? They're not going to, because engineers know better.
All this said, Shifty touched on most of my major gripes. Props to you, Shifty.
Ugh - Major Nelson, Major Healy and Dr. Bellows have all been hitting up Jeannie for too much crack.
Brilliant marketing, but highly deceptive on MS's part. Beside, if Major Nelson's quotes really had any validity, wouldn't MS post it somewhere on MSDN? They're not going to, because engineers know better.
All this said, Shifty touched on most of my major gripes. Props to you, Shifty.
Ugh - Major Nelson, Major Healy and Dr. Bellows have all been hitting up Jeannie for too much crack.
PC-Engine said:SONY started using GFLOPS as a performance metric. MS is only doing it in return. An eye for an eye.
1TFLOPS isn't being honest, 2 TFLOPS isn't either. It's just funny that SONY is stuck in this GFLOPS/TFLOPS game that they started.
I think the only thing everyone can agree on is that no one is being honest - unfortunate...
PC-Engine said:gmoran said:
SONY started using GFLOPS as a performance metric. MS is only doing it in return. An eye for an eye.
1TFLOPS isn't being honest, 2 TFLOPS isn't either. It's just funny that SONY is stuck in this GFLOPS/TFLOPS game that they started.
at least hey have a bigger number
PC-Engine said:gmoran said:
SONY started using GFLOPS as a performance metric. MS is only doing it in return. An eye for an eye.
1TFLOPS isn't being honest, 2 TFLOPS isn't either. It's just funny that SONY is stuck in this GFLOPS/TFLOPS game that they started.
Thats not quite true. 6.2 GFlops as far as I know are genuine, whereas 80GFlops in the XBox GPU aren't, likewise XBox360 and PS3 total GFlops performance (CPUs aside) aren't genuine either, so you could say Microsoft started representing false GFlop measurements this generation and continued to the next generation, Sony not to be outdone followed suit.
- Status
Similar threads
