They signed NDA... I really dont know why they broke it...Maybe for some website traffic....
Printable View
they didnt post the card or the specs, just the results, are they trying to get around a loophole?
True, but then again, the prices you have to pay are huge. I can't really see them generating that much money just be being 6 hours earlier :rolleyes:
I'm not worried by the NDA, chiphell should be. :yepp: And I'm glad we finally have some numbers to go by. I was merely wondering about the NDA. Is that such a huge deal? :p:
Warm climates, try frigen hot, here in Perth it hits over 40c+ quite often each summer, so man I hear you when you say the word Heat!! :yepp:
So when do you normaly start to see the water cooled blocks for these new cards start to roll out?
Day of release?
Week later?
Would love one from Watercool.:up:
So they released a new benchmark just before nvidia releases a new card and the biggest difference between 1.0 and 2.0 is extreme tesselation, not post processing effects, or gpu physics, or lighting/shadows.
http://www.xtremesystems.org/forums/...9&postcount=92
Really is the difference between those two screen shots really worth 20fps on a 5970? I was just reasonably happy it was faster on Dirt2 and possibly faster on AvP, but this just stinks of buying a benchmark.
You'd be surprised.
This could have all been pre-arranged with advertisers, and they're going to pay a premium for each hit to the review page. If I were chipshell, I'd arrange for special release date advertising pricing, because of the guaranteed influx of traffic. Recouping that 200k would be peanuts.
Yeah, the extreme tessellation mode uses over-the-top amounts of tessellation. It's not even close to representing a real world game utilizing tessellation at all. It even looks the same as normal tessellation, which in fact, does do a good job in showing how tessellation in games would stress your GPU.
Sounds logical :) Thanks for the explanation.
I guess you're right. Adds is a powerful way of generating money. The money google generates, is for 97% by adds. :eek:
Here's a nice video about google :p:
http://www.dumpert.nl/mediabase/8669...e_is_eng_.html
96C is obviously a short peak before the fan even ramped up and for all we know the fan could be ramped up to 40% or 90% for the SS.
http://img194.imageshack.us/img194/3055/tempmn.jpg
Power consumption is stupidly high those. The performance per watt is the same as the last generation. This cards needs a refresh badly.
Maybe because I set my expectations so low because of Charlie and Saaya who has tended to agree with Charlie lately, but I am really getting excited about these supposed numbers(7%). I said clocks were going to be around 700mhz with around 20% performance advantage, both which were said to be optimistic. Well the clocks are 700mhz, lets see if we can get 20% or greater for performance when the reviews come out. Chiphell this close to the release date tends to be accurate those.
Still when I compare it to my initial expectations of fermi when the specs were first released in october, I was expecting bigger gains. But a loss in shaders by 6 percent and drop in frequency another 14 percent(shaders were initially supposed to be 1600mhz), still make these results disappointing from what they could have been.
20% is a important number because its too big of a gap for AMD to close with a refresh, in addition its bigger than the gap the gtx 280 had over the 4870. 20% would also solidly justify the 100 dollar price difference.
Still those power consumption appears to suck badly and I can imagine a lot of heat. Too bad winter is near over and spring is coming, which can almost be said the same for Nvidia if they can deliver this launch.
Overclocked at 925/950 with 1.25+ voltage I find my 5870 reachs 92 or so sustained with the default fan profile ( makes its way up to 50-55% fan speed which is quite loud ) Under normal use though its at 32-40% which is bearable. The question to be asked is how loud is the 480 under normal conditions and since we can assume that the furmark result is using a stock card, how much headroom are we talking. The thing I'm least happy with is their number for power... nearly 50% more than the 5870... thats alot of extra juice / heat to cope with (especially with regards to watercooling, thats a lot more heatdump for a rad). All this said, I'm excited to see what these beasts do on a smaller process. As for what they'll do today...
At the comment of 9k extreme on a 5870, this is accurate with a highly clocked i7 and 10.3 drivers. However the dirt 2 numbers, assuming they used the ingame benchmark AND the retail version, are off noticeably...
Honestly it sounds like almost everything Charlie said about Fermi has been wrong, with the exception of silicon revisions and release timeframe. He has said 5% faster than the 5870, castrated to 448 SPs for the top bin GPU with 600/1200 (core/sp) clocks, 300W TDP, 70C at idle while running at 70% fan speed, etc... All of that seems to be nowhere near reality.
wow first review
http://vga.it168.com/a2010/0326/865/...65774_20.shtml
Wow, 10.2 and NO power consumption values. And since I read Chinese, those people sound like absolute PR scumbags.
Chiphell's a better reference point because it doesn't resort to shoddy tactics- and has been so since HD3870 vs 8800GT.