[Jawed]

... On Today's GPUs

http://www.techreport.com/etc/2005q3...g/index.x?pg=1

The first page contains a very interesting point about hierarchical Z buffers being too small in 2004-gen GPUs, for resolutions beyond 1600x1200.

Still reading...

Jawed

[Karma Police]

Good article. I get the feeling that when I get a new VC I won't have to worry about stuttering below 30fps for a very, very, very long time.

[RejZoR]

Why the hell do we need such absurd resolutions?
Games look damn nice at 1280 with 2 or 4x FSAA.
But people use 1600 with 6xFSAA because CPUs are the bottleneck anyway...

[PatrickL]

Because 1920 1200 monitor cost only the price of a top graphic card instead of the price of 3 like some years ago.

[neliz]

Originally Posted by RejZoR

Why the hell do we need such absurd resolutions?
Games look damn nice at 1280 with 2 or 4x FSAA.
But people use 1600 with 6xFSAA because CPUs are the bottleneck anyway...

Because it just looks better on 1600x1200 than on 1024x768..

I do it because I can..

[Kombatant]

Originally Posted by RejZoR

Why the hell do we need such absurd resolutions?
Games look damn nice at 1280 with 2 or 4x FSAA.
But people use 1600 with 6xFSAA because CPUs are the bottleneck anyway...

Absurd resolutions? Well, I guess that a few years back even 1280x1024 was called an absurd resolution to play at by many

[Jawed]

I've been running this 1600x1200x85Hz monitor since 1998.

I sure wish I could have had a card that would let me play games at that resolution the whole time I've owned this monitor...

Jawed

[Wunderchu]

Originally Posted by Jawed

... On Today's GPUs

http://www.techreport.com/etc/2005q3...g/index.x?pg=1

The first page contains a very interesting point about hierarchical Z buffers being too small in 2004-gen GPUs, for resolutions beyond 1600x1200.

Still reading...

Jawed

nice

I was interested in more benchmarks using higher resolutions 8)

thanks for the heads up, Jawed

[Geo]

Originally Posted by Jawed

I've been running this 1600x1200x85Hz monitor since 1998.

I sure wish I could have had a card that would let me play games at that resolution the whole time I've owned this monitor...

Jawed

Alas, mine is 16x12 at 72Hz. And 72 just doesn't cut it for me. I prefer 100, but can live with 85.

[Thowllly]

I bought a cheap monitor that could do 2048x1536 5 years ago. At that point almost all of my gaming friends had monitors that could do that res for several years already! I bought the monitor together with a gforce GTS 64MB that was powerful enough at the time to play many games at that res, only problem was, very few games let me select that res. Later the support for it improved, but by that point my gf2 was no longer good enough to run any new games at 2048x1536

For those who have never tried gaming at that res, you owe it to yourself to try, it's amazing how good it looks, even without AA and just low levels of AF.

[HaLDoL]

Interesting that a single 7800GTX is twice at fast as a X850XT PE in half life 2 at the highest resolution.

http://www.techreport.com/etc/2005q3...g/hl2-town.gif

[neliz]

Originally Posted by HaLDoL

Interesting that a single 7800GTX is twice at fast as a X850XT PE in half life 2 at the highest resolution.

http://www.techreport.com/etc/2005q3...g/hl2-town.gif

Interesting that a single 7800GTX is twice as fast as a 6800Ultra SLI in half life 2 at the highest resolution!

(wait, did I just take of my green hat?)

[AlStrong]

I find it more interesting that the X850 keeps up with the 6800U SLI and even beats it in the lower resolutions.

"X850 BEATS 6800ULTRA SLI!" But it's true...in some cases.

[HaLDoL]

Originally Posted by Alstrong

I find it more interesting that the X850 keeps up with the 6800U SLI and even beats it in the lower resolutions.

"X850 BEATS 6800ULTRA SLI!" But it's true...in some cases.

An ATi card beats 2 nv cards in an ATI game. Not interesting.
NV card beats ATi flagship in an ATI game with double the numbers. Interesting at least.

[compres]

Originally Posted by HaLDoL

An ATi card beats 2 nv cards in an ATI game. Not interesting.
NV card beats ATi flagship in an ATI game with double the numbers. Interesting at least.

So basically its only interesting when nVidia wins.

[neliz]

Originally Posted by HaLDoL

NV card beats ATi flagship in an ATI game with double the numbers. Interesting at least.

Yes, saying that NV's new generation GPU is faster than ati's last generation GPU really puts you high on the IQ list.

(note the way you subtly left out "flagship" after NV as well to suggest that this is mainstream performance.)

[trinibwoy]

Originally Posted by compres

So basically its only interesting when nVidia wins.

Nice summary

[Hellbinder]

hello Mr green.

You did read the bit about the buffer size difference between the G70 and the other cards right?

The difference in High resoliution gaming in Hl2 has zip to to with the majority of the architecture and most everything to do with the G70 being properly equipped for running at that resolution.

Why make a ridiculously stupid fanboiz comment about 2x the performance at the highest resolution when the hard data on why it was going to happen was given before you read the first chart of numbers?

Nvnews must be getting boring these days eh?

[Ragemare]

Originally Posted by HaLDoL

Interesting that a single 7800GTX is twice at fast as a X850XT PE in half life 2 at the highest resolution.

http://www.techreport.com/etc/2005q3...g/hl2-town.gif

Are you deliberately trying to derail this thread?

Did you not read the whole bit about the Z buffer? The reason for this massive difference between the NV card ant the ATi card is well explained, thus meaning due to the lack of mystery surrounding it, it's not particularry interesting.

Whats more interesting is what this means for benchmarking using ultra high resolutions i.e., it's a bit pointless untill there is another card on the market that fully supports them.

[Hellbinder]

Originally Posted by Ragemare

Are you deliberately trying to derail this thread?

Did you not read the whole bit about the Z buffer? The reason for this massive difference between the NV card ant the ATi card is well explained, thus meaning due to the lack of mystery surrounding it, it's not particularry interesting.

Whats more interesting is what this means for benchmarking using ultra high resolutions i.e., it's a bit pointless untill there is another card on the market that fully supports them.

Now not really. Benchmarking at high resolutions is good to show people what you *could* ge gaming at with a GTX and a nice monitor.

[Moloch]

I like it better how the X850 XT-PE is faster than the 6800 ultra at 2048x1536.
More than any other test, I don't think nvidia likes that.
Too bad they couldn't test the 512MB x800xl.

[Sxotty]

It did kind of blow me away how the HL2 was so flat though. It was surprising for me. The other tests did not show that at all, it makes me wonder if something was messed up or something.

Anyway I swore off ridiculously overpriced hardware (>$400) so it doesn't matter to me yet anyway :P

[ANova]

I find it more interesting that the X850 is as fast as the 7800 GTX at 12x10 and not terribly slower at 16x12.

[Hellbinder]

What i found most interesting was how close to the 6800's the X850 was in Doom 3. Even passing it a couple of times.

Thats like a *MAMOTHLY HUGE* improvement in drivers at least for that specific application.

It still gets a butt whippin most of the time but its still 94 FPS at the minimum settings they tested with wich is pretty darn good.

[swaaye]

I guess ATi's l33t group of determined haxor ex-'90s-demo-coding drivermofos got a little bothered over that last fall. Good to see they've pulled it off.