G80 is too hot and power hungry for GX2. They had to downclock heavily even for G71.
Last edited by a moderator:
The LAST R600 Rumours & Speculation Thread
- Thread starter Geo
- Start date
- Status
G80 is too hot and power hungry for GX2. They had to downclock heavily even for G71.
I tried searching in every possible way, but all I could ever come up with was this :
Which titties do we really want to milk ?
I say it's fake
That level505.com site still doesn't show up in WHOIS records. Until we see who is behind it, I call BS. Especially with the slogan "Known for knowledge". I mean, how could they be known for anything when they didn't even exist just a few days ago?!? Is something only a few well respected review sites could claim.
Both Anandtech and Toms Hardware have better 8800GTX results in Oblivion.
Furthermore, the excuse about not being able to physically remove a sticker is plain crap when you can remove anything using Photoshop and the clone brush.
That level505.com site still doesn't show up in WHOIS records. Until we see who is behind it, I call BS. Especially with the slogan "Known for knowledge". I mean, how could they be known for anything when they didn't even exist just a few days ago?!? Is something only a few well respected review sites could claim.
Both Anandtech and Toms Hardware have better 8800GTX results in Oblivion.
Furthermore, the excuse about not being able to physically remove a sticker is plain crap when you can remove anything using Photoshop and the clone brush.
overclocked_enthusiasm
Regular
In context...
"It still seems to me that ATI is so infatuated with the technology that they have forgotten how to run the business. G80 may not be the best architecture for DX 10 or unified shaders but it is in the market and is building market share and most importantly mind share....much like G70 did uncontested for months and months. Something is seriously wrong connecting the dots at ATI from a product launch standpoint."
Perhaps if you read it again it will make sense. I am making the implication that R600 may be superior in both DX 10 applications and the most complete utilization of unified shaders compared to G80. In fact it may be much better...back in the lab. In the meantime, Nvidia launched a product that currently IS the best in both while ATI sits on the sidelines obsessing on the technology. Once again they have missed a product cycle and have given Nvidia uncontested pricing power and market share. Their job is to develop a product and launch it in a timely manner which they have failed to do.
zgemboandislic
Newcomer
It's just bad luck - bad bugs. It is a sad situation, and if they don't manage to get R600 out the door before NV's G80 refresh, possibly on 80nm, I really don't see how they could ever recover. Well they did recover with the R300 that one time. Can they do it again? I really hope so. You don't want to have a microsoft-like monopoly in the gpu industry, hell, it's a monopoly already with all the price fixing roumors, but it could get worse.
zgemboandislic
Newcomer
Well I was about to release my own review of R600 this morning. However, my stove was broken so I decided to fry some eggs on my Intel QX6700 overclocked to 7.3 Ghz while doing the benchmarks, so I don't know whether that had an impact on the final score.
I guess I'll have to do it again. I'll have it ready by end of March.
I call BS
I call BS on that whole site. At least until WHOIS records start showing who is behind it.
Food for thought:
1. It was registered the same day the "review" was published and it only had that single article.
2. Oblivion scores on Anandtech, and on Tom's Hardware are considerably higher.
3. It would be better if they tested with Core 2 Duo X6800 than with QX6700 because there are no Quad Core optimizations yet and if Windows schedules the threads on two cores from different packages it will have penalty of going over FSB for inter-core communication/synchronization.
4. No photos (not even blurred ones), heat, noise, power consumption figures
5. Stickers which can't be removed physically can be killed with clone brush in Photoshop
6. Last update says they experienced CPU overheating and throttling. Now there is even more room for doubts about result correctness.
And last but not least their slogan "Known for knowledge". How can they exist only few (intermittent) days and be known for anything, let alone knowledge? There are only few credible review sites which could claim that, not some newbies who can't even peel a simple sticker off.
All this seems to me like either:
a) Viral Marketingâ„¢ -- influencing people to hold off their 8800 GTX/GTS purchase to harm NVIDIA
b) Scam Of The Yearâ„¢ -- someone earning nice money from AdSense
c) Domain Kickstartâ„¢ -- someone setting up himself a popular domain for selling on eBay
Take your pick.
I call BS on that whole site. At least until WHOIS records start showing who is behind it.
Food for thought:
1. It was registered the same day the "review" was published and it only had that single article.
2. Oblivion scores on Anandtech, and on Tom's Hardware are considerably higher.
3. It would be better if they tested with Core 2 Duo X6800 than with QX6700 because there are no Quad Core optimizations yet and if Windows schedules the threads on two cores from different packages it will have penalty of going over FSB for inter-core communication/synchronization.
4. No photos (not even blurred ones), heat, noise, power consumption figures
5. Stickers which can't be removed physically can be killed with clone brush in Photoshop
6. Last update says they experienced CPU overheating and throttling. Now there is even more room for doubts about result correctness.
And last but not least their slogan "Known for knowledge". How can they exist only few (intermittent) days and be known for anything, let alone knowledge? There are only few credible review sites which could claim that, not some newbies who can't even peel a simple sticker off.
All this seems to me like either:
a) Viral Marketingâ„¢ -- influencing people to hold off their 8800 GTX/GTS purchase to harm NVIDIA
b) Scam Of The Yearâ„¢ -- someone earning nice money from AdSense
c) Domain Kickstartâ„¢ -- someone setting up himself a popular domain for selling on eBay
Take your pick.
Thanks for your post. Very new and insightful.
BTW, who's betting me $100 level 505 wont put the part 2 up soon? Since we're all so sure it's excuses (not sure why they would really need excuses to put up "fake" results..then again not sure why half the time they must be AMD viral marketers, then the other half the time just fake)
Hey thanks for posting it twice too dude. In case anybody missed it the first time two posts prior right?
BTW, you're the first person to do whois on the site. You should go into detective work. Especially since whois usually doesn't show who is behind a site when it's a conspiracy. They charge extra for that though.
I see "I want to believe" is a trendy policy these days...groovy.
I do have my doubts about the legitimity of that review, BTW.It would be a major major leak if some dude somewhere had a working R600 to toy with AND had no NDA to sign and break AND would be inclined to post the results on a fresh from the ground up site. It just doesn`t happen...maybe a teenie weenie leak?Yeah, sure...but a full benchmark?Mehh, not really plausible, now is it.
I do have my doubts about the legitimity of that review, BTW.It would be a major major leak if some dude somewhere had a working R600 to toy with AND had no NDA to sign and break AND would be inclined to post the results on a fresh from the ground up site. It just doesn`t happen...maybe a teenie weenie leak?Yeah, sure...but a full benchmark?Mehh, not really plausible, now is it.
....or they might end up just more or less equivalent with each having their own advantages or disadvantages. But that's still not the point....
How come ATI is "obsessing on the technology" and NV isn't? If anything G80 is the largest departure from an architectural perspective for NVIDIA. We had heard a mighty long time before that NV will arrive earlier with its D3D10 architecture, which means that both IHVs have very different development cycles and it's not just for this product generation but goes IMHO a couple of years back.
NV didn't exactly divert a huge load of resources on RSX development, au contraire to what ATI delivered with Xenos. While NV was chewing endlessly on revamps of the revamps of the basic NV4x architecture both for the PC as the console market, ATI cancelled one design, spent in a relative sense man-hours for R4x0 and was developing in parallel a vastly different USC/Xenos design, a R5x0 for PC and R600 for PC/D3D10. If you count NV4x/G7x and G8x that's 2 vs. 3 vastly different architectures.
NV screwed up royally with NV30 and ever since their basic strategy relies on minimizing expenses and maximizing margins. If RSX would as a somewhat weird example would had been a "crossover" between G7x and G8x, I'm willing to bet good money that we would not have G80 today on shelves.
While obviously releasing R600 later won't come without disadvantages I wouldn't call it a failure either. Even a year or more back most expected NV to arrive earlier and since no one ever hinted on the contrary from ATI's side, I still don't see them failing to deliver.
What I call a failure is NV30 aka FX5800. Something like 9 months late, vastly underwhelming compared to the straight competitor and with the infamous dustbuster onboard. Still NV bathed it out while pondering on low-end/mainstream DX9.0 and minimized the damage as much as possible. None of ATI's recent high end designs were/are even close to that.
ATI arriving 3-4 months later with R600 does not and should not suggest a free ticket to ride for NVIDIA and I'm certain the latter is very well aware of it. What I'm anxious to see in the foresseable future is what ATI has up its sleeve for the mainstream PC segment and R6x0 based laptop designs. I'm still not being as naive to not expect a neck to neck race again down the line.
And yet, arriving 3-4 months late still gives ATI a decided disadvantage. Think about it for just a moment: if you look back at how well each product has done in the marketplace, nVidia has done better than ATI in every generation where they've put their product out first, even if that product, in many ways, was bested by ATI's.
I just admitted that the later arrival comes with disadvantages; I just cannot see anything yet that would suggest anything close to a failure.
I'll flip the coin in just one hypothetical example: assume R520 would had arrived say 4-5 months before G70; would you call the latter a failure?
icecold1983
Banned
i think a big part of this is the fact that nvidias marketing is incredible. imo nvidias offering would have to be far worse than atis for ati to ever outsell nvidia.
- Status