NV40 3DMark 2003 scores revealed -theinquirer

[digitalwanderer]

It really wasnt recieved at all. I mean. How many People with Voodoo5 6000s do you know who've been around to complain or apreciate it?

One, but he appreciates it.

I'd say anyone who forked over money for a V5 6000 was hardcore enough to understand and know what they are getting. I really doubt he bought it for the Power Supply Hookups anyway.

Yeah, it seemed like a rhetorical question...but then I realized I actually DID know someone with one and couldn't pass up answering.

 

[kemosabe]

http://www.digitimes.com/NewsShow/NewsSearch.asp?DocID=64638694A79F2B5748256E6E00481A56&query=NVIDIA

OT: RV380 on 110 nm? no mention of low-K alredy used in RV360? Rather strange.

Well, they RV360 isn't even in the chart, so I don't know why it's strange to not show it as 130nm, low-k.

However, the RV380 at 0.11 does surprise me a bit. I expected RV370 on 0.11u, but I was expecting RV380 to be 0.13 low-k.

That Digitimes chart is screwy. I emailed them yesterday asking them to verify their sources and make a correction. RV370 has been the only chip slated to be on 0.11u.

 

[Heathen]

And since NVIDIA strays away from low-k (even Jen-Hsun says low-k is unecessary heh),

they also said 256 bit memory was unnecessary and look how long that lasted.

 

[Pete]

FYI: 3D card power consumption. Joe posted this in a previous thread, and it's still the source for this sort of info.

As for Athlon 64 power usage, here's a real-world comparison in an excellent review. Note the A64 3400+ system is drawing 119W max, with "standard components." I don't think the CPU is drawing 89 of 119W in a standard system, accounting for memory, fans, HD, and PSU (in)efficiency, but it's possible. (Note the 256MB 9800P adds 50W across the board over the 7000. Now imagine a 150W NV40 basically doubling system power use and heat creation--yikes. I'm really doubting the 150W figure, but not so much the fact that nV may creative a very power-hungry chip.)

 

[Lezmaka]

Don't remember seeing anything about a pic and couldn't find anything by searching, so I apologize if this has been posted before.

Although the scores add up to roughly 12510, wouldn't 453fps in game 1 be next to impossible with a standard non-overclocked cpu? I think Dave said game 1 was the only test where CPU had an impact on performance (or something similar)

http://news.mydrivers.com/pages/20040409121531_17315.htm

 

[PatrickL]

Saw that pic on FM forum. too bad that other tests are not up and tthat we have no idea about the pc used. ( I ll let the approved driver problem for later )

 

[Mark]

3Dmark03 score screenshots are so easy to fake it's not even worth the time to post legitimate ones

 

[digitalwanderer]

3Dmark03 score screenshots are so easy to fake it's not even worth the time to post legitimate ones

Agreed, and any time I see a screeny that lossy it just instantly raises some red flags for me.*






*(I am not saying it's a fake, I ain't bothering to look at it that closely....just an observation on a cursory glance.)

 

[Tim Murray]

Could tell if it's legit if they actually posted the scores to the synthetic benches.

 

[AndrewM]

Could tell if it's legit if they actually posted the scores to the synthetic benches.

How? You could make those up too.

 

[Rys]

It looks wrong, even for an overclocked NV40. God knows what CPU they are using (if it's real of course). I'm using something pretty good and I can't get anywhere near 400 fps, but I can still score a fair bit higher than 12500 with stock CPU.

It's a fake I reckon.

Rys

 

[dan2097]

453.5fps in Gt1 is impossible with current cpus

 

[pakotlar]

It looks wrong, even for an overclocked NV40. God knows what CPU they are using (if it's real of course). I'm using something pretty good and I can't get anywhere near 400 fps, but I can still score a fair bit higher than 12500 with stock CPU.

It's a fake I reckon.

Rys

just out of curiosity, how is that 6800ultra? do you like it so far? what about driver stability?

 

[Tim Murray]

Could tell if it's legit if they actually posted the scores to the synthetic benches.

How? You could make those up too.

The odds of them getting the synthetic benches right are very low, considering that the only real public information is the final score and that the synthetic benches don't contribute to that. As it stands, though, I think that's a fake.

 

[digitalwanderer]

453.5fps in Gt1 is impossible with current cpus

Even with one of them fancy new 64-bit jobbies? (I don't know, I'm just asking....why am I so paranoid that I'll be misinterpretted for flaming today? :| )

 

[I.S.T.]

No way in hell could an Athlon 64 or a Prescott(It has x86-64 in it, it's just disabled.)based P4 get a score that high. Edit: Forgot two words. D'oh. Fixed.

 

[digitalwanderer]

The odds of them getting the synthetic benches right are very low, considering that the only real public information is the final score and that the synthetic benches don't contribute to that. As it stands, though, I think that's a fake.

Dumb question, but if we don't know what the synth scores are and if they WOULD have faked them how would we have the slightest idea if they were faked or not? (I guess I'm trying to figure out if there is a way to determine what the rough synths scores should be, I don't know of any.)

 

[digitalwanderer]

No way in hell could an Athlon 64 or a Prescott(It has x86-64 in it, it's just disabled.)based P4 get a score that high. Edit: Forgot two words. D'oh. Fixed.

Thanks, I haven't been keeping up with the new CPU stuff...I won't be getting one for at LEAST a year! (I'm a "rusty edge" kind-o-guy, bang-for-buck and all that. )

 

[dan2097]

453.5fps in Gt1 is impossible with current cpus

Even with one of them fancy new 64-bit jobbies? (I don't know, I'm just asking....why am I so paranoid that I'll be misinterpretted for flaming today? :| )

Ive actually done some more checking, ITS NOT impossible, however it would need atleast a 2.8ghz athlon 64. (derived from the fact a 3.12ghz athlon 64 got 510fps on the test.... in 320x240 )

 

[tEd]

I'm curious how many websites will use 3dmark03 in their nv40 previews