BRIGHT SIDE OF NEWS About | Advertise | Contact BSN USER Login
| Register 		SUBSCRIBE Newsletter | RSS Feeds
Sunday, May 26, 2013
Email this to a friend.
Your friend's e-mail:
Your Name:
Your e-mail:
Message subject:
Comments
Aesma : what is progress ? by: Anonymous on 1/19/2010
The story is wrong because there are dams around Oak Ridge (and the lab is involved in local energy experiments so they care about the topic).

But I also question the subtitle. Isn't forcing people and organisations to use renewable energy progress ? A progress far more important that yet another supercomputer ?

That supercomputer is in fact supposed to be used for climate and energy research, so powering it with fossile energy would seem a little strange.
A Hydro Electic Pant 5 miles Away !! by: Anonymous on 1/7/2010
The reason I question the excuse for the 'shelving' of the project is that there is a Hydro Dam 5 miles away from ORNL, it's far closer than all the coal plants, so likely it supplies all the power to ORNL on the grid it shares. Plus there are a bunch of other nearby too.

I think there's more to it than the type of electricity being provided to the building that houses the machines.
CES: Answers or More Questions by: Greg442 on 1/6/2010
Charlie predicted in May 09 that Fermi that didn’t have a chance in hell to be released until Q1 2010, when so many other tech sites were towing the NVidia PR line of 2009 launch. Wood screws anyone? Love the guy or hate him, he was “right” on Fermi when so many others missed the boat completely. Fudzilla anyone? I generally find Theo post to be reasonable, the powercooler 5870 post was very surprising and honest, so I don’t think he’s carrying water for NVidia, but this Oak Ridge thing smells a bit funny to me. What sticks out like a sore thumb for me is this; from conception Oak Ridge (who’s well versed in the super computing world) would’ve designed a draft of this supercomputer on paper long before seeking to award contracts. They would’ve known both the power requirements and limitation and tentatively chosen components to fall within their design criteria. Once a “final design” was reviewed and approved, they’d seek partners to produce those components to specification. This article makes the people at Oak Ridge look like complete buffoons, and we’re lead to believe that Oak Ridge haphazardly designed a supercomputer in total disregard to buildability, i.e. cost, power, heat, cooling etc. And only realized they didn’t have the go ahead to build the thing until after the deal with NVidia was consummated. Sorry Theo, this is kind of hard for me to believe. But even if I accept your article as facts, it still doesn’t disprove a power issue with Fermi, nor does stating that the new gen Tesla (448SP cut down) has the same power requirements as the previous generation inspire confidence. Because isn’t Fermi a die shrink from the 200 series? And as such shouldn’t improved power requirements be expected? Also considering the vast power requirement improvements we see with the AMD/ATI 5000 series isn’t the pressure now on NVidia to make a similar showing? Now I’m sure there’s a good excuse…cough… I mean a good reason for the 448SP cut down and since this is the eve of the CES show, where Fermi is rumored to be unveiled… let’s see if some of these question are laid to rest once and for all.
I luv Charlies bullshit more than Theos' crap by: Anonymous on 1/6/2010
To the ignorant semi-accurate hater, do you even read the article before commenting ? I dont see any exposure , if anything , there is more BS in this crappy article than anything I read on this site. There sure was tons of investigation , FBI be damned, that was conducted just to write a one page article ( two if you count the pictures. So what if the rumors said they cancelled the super computer. Fuggin nvidia should get off its ass and do something to actually PRODUCE a produce, instead they might be a whole year late with a lame ass technology that no one will want in the end.
supercomputing gossip by: Anonymous on 1/6/2010
Wow -- this guy really doesn't know anything. As someone who works for one of the labs mentioned, I'm honored we have our own version of the National Enquirer! This must be what celebrities feel like! (P.S. Could you at least proofread the article and fix the grammar mistakes?)
nothing than hot air by: Anonymous on 1/6/2010
What a crap story with no content. The truth is, soon ATI will take over the client market of discrete GPUs from nVidia. GPGPU will definitely have a future. But at the moment it's nothing than just excuses for poor nVidia hardware. ATI has a smarter and much more efficient design today. And they will be even better when time is right for GPGPU. That is nVidia's real doom, built on mismanagement and arrogance.
by: Anonymous on 1/6/2010
I don't know who your sources are but if you look at a map you will see ORNL is in the middle of the TVA and is literally surround by hydroelectric plants. I would guess 80-90% of their power comes from renewal energy sources.
Blue Waters? by: Anonymous on 1/6/2010
Blue Waters is an NSF machine being built at the University of Illinois. What does it have to do with the politics between DOE labs?
by: Anonymous on 1/6/2010
to the guy below me , you are a real dick , you know that ? instead of trying to get the real story you nitpick about minor info , it doesn't make a difference if it were 225w or 500w , the point here is that the deal was off because of power regulations not because of Fermi hardware issues , Supercomputers are designed with huge power consumption in mind .

If you still believe that it is a Fermi problem , then I suggest you crawl back to Semiaccurate's forums and bitch there as you like , that place is full of biased and blind people like you , leave the rest of us yo try and see the real truth .

Thanks Theo for exposing charlie's bullshit .
Correct by: Anonymous on 1/6/2010
They could even write less than or equal to 500W which is fully compatible with 160W and 170W, right? :D
If Tesla C2070 is ~170W do you think they still say : "less than or equal to 225W"?

“Discussion is an exchange of knowledge; an argument an exchange of ignorance.” - Robert Quillen
@What a BullShitNews :( by: Anonymous on 1/6/2010
The nVidia paper just specifies that the power dissipation is "less than or equal to 225W" - which is fully compatible with 160W and 170W.
x by: Anonymous on 1/6/2010
Since Valich is an extremist leftist, in essence he helped destroy progress and Nvidia. How ironic.
x by: Anonymous on 1/6/2010
Anyways, electricity usage is not allowed by extreme leftists like George Bush who cowered before the leftists and their Kyoto treaty, their goal is to stop progress. So lets be realistic, folding at home and HPC, let alone GPU's, are in danger.

If California can regulate big screen TV's to use very little power as they recently did (effecively banning plasma displays in the process), how can GPU's be far behind?
x by: Anonymous on 1/6/2010
Enviromentalist extreme leftist liberals destroyed Nvidia GF100 sales through the Kyoto treaty lol. How ironic.
What a BullShitNews :( by: Anonymous on 1/5/2010
"Upcoming Tesla C2050, C2070 and more importantly, the S2070 4GPU-server [the base for the 10+ PFLOPS rig in ORNL] uses roughly the same amount of power as the present generation of Tesla products. Even the 6GB GDDR5-memory carrying Tesla C2070 consumes 190 Watts, in the range of Tesla C1060 [65nm - 170W, 55nm - 160W]"

I recommend that before you do this big FBI style investigation, just read the following:
http://www.nvidia.com/docs/IO/43395/BD-04983-001_v01.pdf

Board Power Dissipation: <= 225W
© 2009 - 2011 Bright Side Of News*, All rights reserved.
© 2009 - 2013 Bright Side Of News*, All rights reserved.