Advanced search

Message boards : News : Unboxing of our new GTX470 & 480

Author Message
ignasi
Send message
Joined: 10 Apr 08
Posts: 254
Credit: 16,836,000
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 16511 - Posted: 23 Apr 2010 | 13:35:42 UTC
Last modified: 23 Apr 2010 | 13:35:51 UTC

Here you have some shots taken this morning of the unboxing and installing of the brand new GTX470 and GTX480 donated to GPUGRID by user ftpd. Thanks so much again! http://tinyurl.com/welovefermi

cenit
Send message
Joined: 11 Nov 09
Posts: 23
Credit: 668,841
RAC: 0
Level
Gly
Scientific publications
watwat
Message 16532 - Posted: 25 Apr 2010 | 12:01:33 UTC - in response to Message 16511.

don't you think you're too much biased in nvidia direction?

I hope that you receiveat least some money for this. It is a little ridiculous how you're mixing scientific method and fanboy comments...

Snow Crash
Send message
Joined: 4 Apr 09
Posts: 450
Credit: 539,316,349
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 16533 - Posted: 25 Apr 2010 | 13:01:56 UTC - in response to Message 16532.
Last modified: 25 Apr 2010 | 13:04:51 UTC

don't you think you're too much biased in nvidia direction?...


GPUGrid is continually reviewing all technology. So far they have used PS3, GPU, and CPU to find what is the best to process their science at any given moment in time. Originally they were using cell processor in PS3 and when Sony locked up the OS on those machines, GPUGrid moved on in their investigations. Currently Nvidia/ CUDA is the best processing environment for their applications but if you read up a little you would know that they are continually trying out new versions of their code on ATI cards.

The biggest problem is that the ATI development environment simply is not helping much. Last week GPUGrid, for the second or third time in the last couple oif months, optimized code and tested on ATI. Unfortunately it is still proving to process substantially slower than Nvidia/ CUDA.

They have also investigated multi-cpu versions of their applications and found that to also be substantially slower than Nvidia/ CUDA.

Not just theory, not just blind faith in any given technology, these guys are for real in finding the best and are not afraid to move to any technology that proves to be better at processing their work.
____________
Thanks - Steve

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,834,518,624
RAC: 292,156
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 16534 - Posted: 25 Apr 2010 | 16:57:12 UTC - in response to Message 16533.

NVidia cards cost more to buy and to run, but they are not just designed to play games; hence their use in CAD, MRI, and of course molecular analysis.

cenit, present ATI cards could not possibly compete with Fermi cards here. Leave your judgement until a mature Fermi application has been realised, utilized and analyzed while running GPUGrid tasks - it won't be long.

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1895
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 16547 - Posted: 26 Apr 2010 | 7:07:36 UTC - in response to Message 16532.

Hi Cenit,
no we are not biased towards Nvidia, nor we receive money and not even cards lately.

We simply have been waiting this new chip for months and we were very happy to be able to try it. A new chip happens every couple of years, so it is an important event.

In this case, the card was also donated by user, so we though to show how important this was. We would do the same for ATI, for a chip with equal expectations for compute (maybe their next one).

I can also add that, so far, the performance of Fermi is disappointing but I remind this to the appropriate thread in the forum.
gdf


don't you think you're too much biased in nvidia direction?

I hope that you receiveat least some money for this. It is a little ridiculous how you're mixing scientific method and fanboy comments...

cenit
Send message
Joined: 11 Nov 09
Posts: 23
Credit: 668,841
RAC: 0
Level
Gly
Scientific publications
watwat
Message 16549 - Posted: 26 Apr 2010 | 7:55:26 UTC - in response to Message 16547.

thanks for your reply.
I want to explain better to some of those that wrote in this thread that I'm not a fanboy of anything at all (ATi first and foremost).
The problem I saw is that it seemed to me that someone in the gpugrid team was a little biased. Didn't you see the tinyurl? "welovefermi"?? Poor choice, if I could say!
And also, two news for a card that no-one is able to buy, is hyper-hyped and is deliberately reduced in performances (double-prec, I know it doesn't matter to you)?
And yes, I know the status of gpugrid (and OpenCL in general) on ATi.

I hope you could understand why I wrote that comment.

Profile bloodrain
Send message
Joined: 11 Dec 08
Posts: 32
Credit: 748,159
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwatwatwatwat
Message 17267 - Posted: 24 May 2010 | 22:17:21 UTC - in response to Message 16511.

another thing with ati. is this is their first mass use of the stream tech to do research . its like when the 8800 series card came out. the later stuff was better seeing they knew how to build it correctly and cheaply

Post to thread

Message boards : News : Unboxing of our new GTX470 & 480