![](/i/fill.gif) |
![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
device which delivers 4 TeraFLOPS of computational goodness. And if
C1060. (The S1070 is basically four of these in a box. The C1060 is just
a normal PCI card.) This still delivers 933 GigaFLOPS of compute power.
...of course, all of those are *peak* values. So the S1070 delivers "up
to" 4 TeraFLOPS. (Depending on how well your software utilises it.)
All of those are also for *single-precision* floating-point computation.
The C1060 delivers 933 GigaFLOPS peak power for single-precision, but
for *double-precision* it delivers a far less impessive 78 GigaFLOPS.
(And that's still *peak* performance.) The S1070 correspondingly
delivers only 345 GigaFLOPS in double-precision arithmetic.
Hell, if I had any clue how much performance a current CPU delivers, I'd
know if these numbers are impressive or not! ;-)
I don't think I'll rush out and buy a Tesla any time soon though.
GigaFLOPS. That's presumably single-precision again, but it's comparable
to the (vastly more expensive) Tesla C1060, so it's not unreasonable to
expect double-precision to be similar too.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
"Invisible" <voi### [at] dev null> wrote in message
news:496df9eb$1@news.povray.org...
I'll just wait until they start coming with a holographic interface ;-)
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Jim Holsenback wrote:
> I'll just wait until they start coming with a holographic interface ;-)
<Vader> I find your lack of faith disturbing... </Vader>
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Invisible wrote:
> Hell, if I had any clue how much performance a current CPU delivers, I'd
> know if these numbers are impressive or not! ;-)
Well, consider that a current CPU runs at some small multiple[1] of 3GHz,
and it's unlikely to do more than some small multiple[2] of FLOPs per clock
cycle, it sounds pretty fast to me. I suspect the difference between "peak"
and "sustained" for both is going to be based on how fast you can feed them
numbers from RAM.
What I think is a cool factoid is that Intel (or IBM?) is advertising they
have a petaflop processor. Consider the early 70's mainframes, like the ones
that ran the Apollo missions to the moon[3]. Give three of those to each
person on Earth, and you have about a petaflop. :-) Pretty awesome.
[1] Depending on the number of cores, for example.
[2] Depending on the SIMD width of SSE/MMX/etc instructions.
[3] Like the Sigma Scientific Data Processor X560 I used when first learning
to program. :-)
--
Darren New, San Diego CA, USA (PST)
Why is there a chainsaw in DOOM?
There aren't any trees on Mars.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> Hell, if I had any clue how much performance a current CPU delivers,
>> I'd know if these numbers are impressive or not! ;-)
>
> Well, consider that a current CPU runs at some small multiple[1] of
> 3GHz, and it's unlikely to do more than some small multiple[2] of FLOPs
> per clock cycle, it sounds pretty fast to me.
Er, yes.
240 cores running 96 threads each yielding over 12,000 threads? That's
pretty parallel... ;-)
> I suspect the difference
> between "peak" and "sustained" for both is going to be based on how fast
> you can feed them numbers from RAM.
"A supercomputer is a device for turning a compute-bound problem into an
I/O-bound problem."
Of course, a GPU is way faster than a CPU, but that's because a GPU
can't do everything that a CPU can. (I emphasise: *because*.)
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |