http://www.bit-tech.net/news/hardware/2010/06/02/nvidia-geforce-gtx-460-specs-and-launch-dat/1
£200 then loooooooooool
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Nvidia GeForce GTX 460 Specs and Launch Date Leaked
- Thread starter RavenXXX2
- Start date
http://www.bit-tech.net/news/hardware/2010/06/02/nvidia-geforce-gtx-460-specs-and-launch-dat/1
£200 then loooooooooool
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
He clearly didn't, he's the master of promising things he can't deliver, sounds like classic Nvidia tactics to me.
As mentioned, it probably means 480gtx clock speeds.
Though, does anything thing it sound suspicously like a GT200b, shrunk down with dx11(you would assume) tacked on. 240 Shaders? 32 shaders in a cluster, does not compute, if it maintains Fermi's architecture, just cut down shader clusters, it would be a 256/224 shader part.
Power wise, well, it makes no sense at all, its not a cut down part so should have WAY lower power consumption. When you fused parts of the core off, theres still leakage, and some power wastage, thats why a 352shader 465gtx still uses so much power.
A core thats half the size should have really half the power, maybe even less, 180W is wrong, just completely and utterly wrong for a half Fermi part. At Fermi style clocks a core half the size should use half the power, if its got increased clock speeds, it still shouldn't use anywhere near that much power, its utterly ridiculous.
It would be very odd for them to make a significantly different midrange to top end, though how bad Fermi is from a manufacturing point you can understand it, but even if its a 16 shader cluster instead of 32, with different ratio's of rop's/tmu's per cluster, thats a massive design difference and really would have had to be designed like that WELL before Fermi had its big troubles.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
I hope it's another g90 rebrand
It would be hilarious, but its incorrect, Bindibaghi is a complete twit, in the comments further down he's managed to say he's had the memory amount confirmed but NOTHING else, he GUESSED at 240sp's.
240sp's confirmed would have meant something completely radical, some twit guessing at 240sp's from an architecture with 32sp granularity is just that, some twit guessing something stupid.
Someone else there decided to talk himself into a 460gtx with 384sp's, decided it can be 850mhz clocks and somehow will match the 480gtx, ignoring the large drop in shaders meaning a larger clock speed would be required, and the memory/lack of bandwidth means it wouldn't get close in performance.
Seriously, bit-tech, when they review a few things do ok, but screw up constantly on other things.
Comparing every other generation, pretty much ever, the midrange part will be a half Fermi, it will have 256 shaders(if they get high yield on fully working parts of course), half the memory bus, half the memory, similar clock speeds, half the power, half the rops, half the tmu's, etc, etc, etc.
The confirmed memory being half, which confirms half the bus, pretty much to me confirms its a half Fermi, nothing more, nothing less. It will not get to 480gtx speeds, honestly I'd again put that down to someone misunderstanding something, BADLY, again.