The inqurier posted up a comment saying there were supposed to be drivers out with a "30% perfomance increase". I wanna know cause thats the difference between Nvidia being king or a high priced flop.
Anyone seen any 6800 benches with the new drivers?
- Thread starter Ages120
- Start date
Yeah I've seen a couple people with 6800U cards been checking thier threads haven't seen anything about new drivers though. The memory overlcocks are pretty impressive but that complex core holds em back. Lord Arse or Shamino or whatever his name is did same crazy overclocks with a cascade cooling system though.
That seems to be the case with the X800s as well. Overclocking the memory seems to have little effect.
edmundoab
Diamond Member
Nvidia always comes out with super duper drivers that improves benchmarks by a decent margin.
Again the optimization factor....
Which some people will call Cheating.
Should be better then whats available now especially when bugs are fixed.
Again the optimization factor....
Which some people will call Cheating.
Should be better then whats available now especially when bugs are fixed.
I am not sure why but I am having extreme trouble trying to figure out which way to go. Damn ati and nvidia they have to release such good products. Why couldn't one of them made a flop and make it easier. Fx 9700 days were bliss. I seriously think the extra effort Nvidia put into thier architecture will pay off with smarter more refined code pushing the 6800 beyond the x800. Not to mention thier army of programmers. I thought this when I first read thier specs they just have so much freedom with thier hardware I knew they would make performance gains in the software. Where as with the X800 is to a point where we already know what they can do with it.
Never heard of it
Don't forget their 220 M transistor count which will make it harder for them to reach higher clock rates, they also have the heat dissipation problem.
I tend to agree with you for the most part. But today it has turned into who can write the best acceptable optimization/cheat and make the other guys card "suck". I personally would go with NV on my decision (if i were to actually fork out that much cache, which i wont) basically because their archetecture seems to be better then the aged R300 core and therefore seems more future proof in terms of POTENTIAL "optimizations" and also ps3.0 is included (more POTENTIAL gains), as well as their history with drivers that are a little better (opinion of John Carmack) and also the archetecture just being outright new. Thats just my take on it all. That is considering that the cards are even priced. Otherwise buy the one that costs $20+ less. Be glad, atleast u cant go wrong on what u buy
http://www.theinquirer.net/?article=9172
"The new driver is to be called Detonator FX and it will result in five to 50 percent of "ballistic" performance increases depending on the application while the realistic rather than the ballistic indicates the number will be between 20 to 30 per cent."
http://www.theinquirer.net/?article=9486
"3Dmark 2001SE jumped from 7295 to magnificent 10870 which is about a 33 percent increase in speed, while the famous nature test was scoring 32.1FPS and with Detonator FX 44.03 scores 66.1 and so is 50 percent faster than with 43.45 that we used.
In 3D mark03, the score improved from 2197 to 2737 and 600 marks really makes a difference here while the mother nature test result increased from 11.6 to 13.1 FPS.
Quake result has slightly decreased, but just by 1 to 2 FPS while Aquanox jumped in 16x12 from 24 FPS to 26.8 FPS."
And then another supposed wonder driver
http://www.theinquirer.net/?article=10426
"Despite the mammoth performance jump these new drivers deliver in 3DMark03, we see no such corresponding increases in any so-called "real-world" tests"
How many driver revisions will it take for people to stop truly believing in the driver percentage game nvidia and the inq play with every release, then fail to actually deliver in places where people care
"The new driver is to be called Detonator FX and it will result in five to 50 percent of "ballistic" performance increases depending on the application while the realistic rather than the ballistic indicates the number will be between 20 to 30 per cent."
http://www.theinquirer.net/?article=9486
"3Dmark 2001SE jumped from 7295 to magnificent 10870 which is about a 33 percent increase in speed, while the famous nature test was scoring 32.1FPS and with Detonator FX 44.03 scores 66.1 and so is 50 percent faster than with 43.45 that we used.
In 3D mark03, the score improved from 2197 to 2737 and 600 marks really makes a difference here while the mother nature test result increased from 11.6 to 13.1 FPS.
Quake result has slightly decreased, but just by 1 to 2 FPS while Aquanox jumped in 16x12 from 24 FPS to 26.8 FPS."
And then another supposed wonder driver
http://www.theinquirer.net/?article=10426
"Despite the mammoth performance jump these new drivers deliver in 3DMark03, we see no such corresponding increases in any so-called "real-world" tests"
How many driver revisions will it take for people to stop truly believing in the driver percentage game nvidia and the inq play with every release, then fail to actually deliver in places where people care
If it weren't for the ForceWare series 50 drivers which did exactly that, you may have had a point. Just because they're stupid and don't realize that driver cheats don't achieve anything besides bad press and a bad reputation, doesn't mean that they can't also gain completely legitimate - and rather significant - performance gains at the same time.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 20K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter