WOOOOWWW I NEVER WOULD HAVE GUESSED. THX GOD FOR THE INQ.

/Sarcasm

Well, see kids, thats why first gen card suck! :)

I don't see why something like this would be made up.

heck most cards will see a large improvment from a better procesor but thanks for the heads up capt. obvius :roll:
j/k zek dont kick my a$$ ok

by: PVTCaboose1337

---

Well, see kids, thats why first gen card suck! :)

---

first off, that's not anything that's news, and second of all, it doesn't say anything about the beastliness of the hardware.

we need to get some real benchmarks @ 16*12 4xAA 16xAF and higher to see if this thing is strong


saying a 3Dmark score over 10000 is cpu limited is like saying water is wet. we all know that, that's the way it's been for every iteration of 3Dmark.

it will score even less higher than a good set of x1950's on 3D05, and less high still on 3D03, etc etc

run 3D06 @ higher res (or any 3D app for that matter) to see how it is when the cpu isn't limiting the gpu

oh, and pvt, i'm not necessarily calling you out, but that's the general comment i am responding to, if any

by: strick94u

---

heck most cards will see a large improvment from a better procesor but thanks for the heads up capt. obvius :roll:
j/k zek dont kick my a$$ ok

---

I can edit that post at my leisure :). But I won't because that wouldn't be very nice. Anyways, I was just trying to point out that the VGA bottleneck is going away, and is being replaced by a CPU bottleneck.

Lets be real now, All cards suffer for lower scores from a bottle neck. And Putting that kind of card in a P4 is pretty damn obvious.

I own a X850pro AGP with clocks at 570/600 and score 5.3k when I could be scoreing 6k but I have a Damn P4@3.1ghz. So this is like a "no duh" post by the Inq.

Oh yes, one more thing, I'd also like to point out that 3Dmark06 is supposed to be a VGA test :p.

by: zekrahminator

---

Oh yes, one more thing, I'd also like to point out that 3Dmark06 is supposed to be a VGA test :p.

---

it's supposed to be, and is for most cards, but clearly not for the 8800gtx

just like 3D05 used to be

and 3D03

and 3D01

anyone who thinks this says anything about the card is not that bright:slap:


the cpu will be even more of a bottleneck for the R600 in 3D06 @ default

by: zekrahminator

---

Oh yes, one more thing, I'd also like to point out that 3Dmark06 is supposed to be a VGA test :p.

---

Not that much anymore, remember that CPU score is now an integral part of your overall 3dMark06 score, dunno what's the big news here, same will happen with any card when using '06, the CPU has a big impact on the final score, this benchmark is a lot more dependant on CPU than any previous 3dMark.

by: zekrahminator

---

I can edit that post at my leisure :). But I won't because that wouldn't be very nice. Anyways, I was just trying to point out that the VGA bottleneck is going away, and is being replaced by a CPU bottleneck.

---

I was just kidding
but yes it looks like it will need a big dog cpu behind it and a large case and fast memory fast bus lots of air and a better O/S than vista (did I say that out loud?)
As I said in another Thread all the years we fought with the cpu bottle neck it looks like its returning maybe the Quad Core will be the answer

When have graphics card not been CPU dependent?

take my 7900gto for example

un overclocked 2.0ghz proccy of amd64 venice i get 9000 or so
@ 2.79ghz, i get 11000

same deal

I don't see how quad core is going to help when nearly all games released are not multithreaded... We need more raw processor performance!

by: Exceededgoku

---

I don't see how quad core is going to help when nearly all games released are not multithreaded... We need more raw processor performance!

---

Intel physics engine will use them.

AMD physics engine? Nvidia physics engine? Aegia Physics engine? what a war.

by: 15th Warlock

---

Not that much anymore, remember that CPU score is now an integral part of your overall 3dMark06 score, dunno what's the big news here, same will happen with any card when using '06, the CPU has a big impact on the final score, this benchmark is a lot more dependant on CPU than any previous 3dMark.

---

i agree...in 06, the CPU takes a lot into the final score

What we SHOULD be seeing is the SM2.0 and SM3.0 scores with the two processors...that would give a better indication

I could understand if they used at least reasonable processors. A 2.66GHz P4 will bottleneck a 9800Pro, so of course it will kill a 8800GTX's performance. Come one, at least be a little resonable INQ.

testing anything on a P4 of any speed is just unreasonable, their crap :p

http://hardforum.com/showthread.php?t=1115508&page=1&pp=20

they seem to agree about the real implications of this over at [H]

by: Ketxxx

---

testing anything on a P4 of any speed is just unreasonable, their crap :p

---

thats what i was thinking hell thats worse than sticking it on a s754 (w/ pcie obviosly and they do have those you know)

LOL people don't understand that people that work at the inquirer submit a story and its placed. The author's name is never told lol.

3D mark 2006 is heavily CPU dependant in its scoring, with my old 1800XT on a 3800+ @ 2.8Gig I was getting around 4500 I think , same card, same clocks, same ram but when I had a 4200 x2 @2.8Gig I remember it increased the score by 600 odd points, it really does like dual cores!

by: Tatty_One

---

3D mark 2006 is heavily CPU dependant in its scoring, with my old 1800XT on a 3800+ @ 2.8Gig I was getting around 4500 I think , same card, same clocks, same ram but when I had a 4200 x2 @2.8Gig I remember it increased the score by 600 odd points, it really does like dual cores!

---

Yes, it is very CPU dependent. My single 7900GT scores over 5000 with my 4400+ @ 2.8GHz, however when putting the same card in my Celeron D 320 @ 3.8GHz(which is arguably better then the 2.66GHz P4 INQ used) machine it lost over 1000 points.

by: newtekie1

---

Yes, it is very CPU dependent. My single 7900GT scores over 5000 with my 4400+ @ 2.8GHz, however when putting the same card in my Celeron D 320 @ 3.8GHz(which is arguably better then the 2.66GHz P4 INQ used) machine it lost over 1000 points.

---

Altough I really doubt that even at 3.8Ghz, a Celeron D would be better than an OC'd Core 2 Duo, I agree 110% to your statement about 3DMark06 being very CPU dependant :)