Print 6 comment(s) - last by greylica.. on May 5 at 8:13 PM

Nvidia GeForce GTX 690
DailyTech's roundup of hardware reviews from around the web for Thursday

GeForce GTX 690:
@ AnandTech
 @ Benchmark Reviews
 @ Bjorn3D
 @ Guru3D
 @ HardOCP
 @ Hardware Canucks
 @ HardwareHeaven
 @ Hi Tech Legion
 @ HotHardware
 @ Legit Reviews
 @ Neoseeker
 @ OCC
 @ PCPer
 @ Tom's Hardware
 @ XbitLabs

Noontec Zoro Headset @ XtremeComputing

Raidmax Seiran Case @ ThinkComputers
SilverStone Precision PS07 White Case @ Hardware Secrets
SilverStone TJ04-J Evolution Case @ Ninjalane
Zalman Z11 Case @ HardwareLOOK

Thermalright Silver Arrow SB-E Cooler @ Tweaktown
Zalman CNPS10X Optima @ Frostytech

Nexus Value 430 PSU @ Hardware Secrets

Corsair Force GT 180GB SSD @ Hi Tech Legion
Plextor M3P 256GB SSD @ The SSD Review
RunCore Pro V MAX 240GB SSD @ Tweaktown

Video Card
ASUS Radeon HD 7870 @ t-break
PowerColor PCS+ Radeon HD 7970 @ HardwareHeaven

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

Poor to compute...
By greylica on 5/4/2012 1:51:14 AM , Rating: 2
Too bad that the card is 1/64 of the compute power of the GTX 580, with the compensation of more caped processors. Seems that the GTX 480 was good for computing, GTX 580 better for computing (consumers started to like ), GTX 680 - cold water to force Quadro sales again. Yeah Nvidia, you made your test...


RE: Poor to compute...
By TakinYourPoints on 5/4/2012 4:21:57 AM , Rating: 2
If you need compute power then you don't get Kepler, it is simple. These are gaming cards, pure and simple, and removing compute increased efficiency, these are way cooler than the 400 and 500 series cards.

RE: Poor to compute...
By greylica on 5/4/2012 5:13:11 AM , Rating: 2
Die Shrink is the key to be cooler than it's predecessor here, and not a ultra high latency in the cuda cores.
Is the card faster for games using more latency for the processors that will also calculate physics for DX11 ?
Are you nuts ?

RE: Poor to compute...
By TakinYourPoints on 5/4/2012 10:14:31 PM , Rating: 2
Die shrink is one piece of the equation. Devoting more resources towards gaming performance at the expense of compute performance is something you're skipping on.

For gamers it makes total sense, especially given that computation for physics is still not as much a bottleneck practically speaking as other things.

RE: Poor to compute...
By greylica on 5/5/2012 8:13:10 PM , Rating: 2
Anyway, I told my friends that are using Cycles and other Raytracers to wait or move to ATI 7970, and the only one that bought this card for compute, is very frustrated (He was using a GTX 580).

Nvidia also inserted some bottleneck in the GTX 4XX in the GLReadPixels, the problem persisted in the GTX 5XX and 6XX series, where the GTX 2XX series is at least 10 times faster for vertex selection, forcing coders to recode using Cuda. It's clear that they are trying to slowdown Cuda to screw those coders/users again to slowdown GLReadPixels. Now I can see that the problem is interconnected and caused by purpose.

Thank you ;)

Beauty And A Beast
By skeansmith on 5/3/2012 7:04:59 PM , Rating: 3
The card may not be practical but who cares? It is the muscle car of vid cards!

"What would I do? I'd shut it down and give the money back to the shareholders." -- Michael Dell, after being asked what to do with Apple Computer in 1997

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki