🚀 go-pugleaf

I thought frame rate limits were also goverened by your monitor refresh in 
UT?

Its a false limit I know but thats the number people quote.

Messa


"Folk" <Folk@folk.com> wrote in message 
news:mc0823pp7hc5bn3oo8nmsdgdav9i4kck4h@4ax.com...
> On Sun, 15 Apr 2007 21:20:27 +0200, Bruno <dont.use@this.address.ever>
> wrote:
>
>>Folk wrote:
>>> On Sat, 14 Apr 2007 04:36:24 +0200, Bruno <dont.use@this.address.ever>
>>> wrote:
>>>
>>>> But regardless if you go ATI(AMD) or NVIDIA then gfx card is still more
>>>> important than CPU when gaming.
>>>
>>> Bzzzzzzzzzzzzzzzzzzzzzzz.  Wrong answer.
>>>
>>Not so unless you play in low resolution with low graphics quality
>>settings in the driver control panel.
>>
>>
>>
>>> I've benchmarked both UT99 and UT2004 extensively.  I can say beyond a
>>> shadow of a doubt that the CPU is the bottleneck in virtually all
>>> cases.  The only time graphics would be a problem is if you were using
>>> integrated graphics.  The OP's 9600XT is *plenty* of vid card for
>>> UT99, but his Duron is certainly holding him back.  If he would shove
>>> in the fastest processor his current mobo would allow, he would see
>>> his framerates soar.
>>>
>>As for UT2004 I can't say but the original poster focuses on UT so I did
>>as well. Unless you go really slow in the CPU department or settle for a
>>low resolutions the bottleneck is the GFX power of the PC.
>
> Like I said, I've benchmarked both games extensively.
>
>>Oh and the 9600 XT is not plenty unless we talk low resolutions, okay
>>but NOT plenty!
>
> It's plenty for UT99.  The hot card when UT99 came out was something
> like a Geforce 2 and it ran the game fairly well.  The 9600 XT is head
> and shoulders above that.  It's *plenty* of card for UT99.
>
>>It's very easy to see if the gfx card is the bottle neck or not. Just
>>run a timedemo in 640*480 with low quality settings in the gfx driver
>>and that score will tell you what the CPU + memory speed of the system
>>allows for (unless you gfx card is very slow then it's all ready a
>>bottleneck). Next increase resolution plus driver quality settings and
>>run the timedemo again. If the frame rate scores are aprox. the same
>>then it's not the gfx card which holds UT back if the scores are lower
>>then the gfx card is next to upgrade.
>
> I'm sure the OP would see no difference using your test method.
>
> Two systems ago (Aug of 2004) I was running a P4 at 2.53 (an
> overclocked 1.6... those things would overclock like mad) and a Ti4400
> vid card.  I bought a copy of Doom and my Ti4400 was puking all over
> the place, so I bought a 6800 GT (still using it in my current
> rig!!!).  My Doom framerates jumped nicely and the stutters were
> completely eliminated.  At the time, that was a two generation jump in
> vid cards.
>
> So what did that upgrade do to my UT99 framerates?  Zip.  Nada.
> Absolutely no difference whatsoever!  The only difference was that I
> could now pull the same framerates at 1600 x 1200 as I did at 1024 x
> 768 due to the pixel pushing power of the 6800 GT.  But at the same
> time I could underclock/overclock my processor and watch my UT99
> framerates go up and down almost exactly the same percentage as the
> CPU core clock.  That, my friend, is a perfect example of a CPU bound
> game.
>
> As an aside, hardware enthusiast sites quit using UT99 as a vid card
> benchmark for the same reason and instead used it for CPU benchmarks.
>
> It was a bit difficult to find a benchmark with such old data, but I
> managed to dig this up at Firing Squad:
>
> http://www.firingsquad.com/print_article.asp?current_section=Hardware&fs_article_id=1478
>
> Scroll down to the UT2004 section (UT99 was the same in this regard)
> and notice how the X800 Pro had virtually the same framerates at 800 x
> 600, 1024 x 768 and 1280 x 1024.  Also notice how the framerate
> increase was linear from a 2100+ up through the 3400+.  Also notice
> the comment at the bottom of the graphs:
>
> "We're CPU-bound with the X800 PRO at 800x600 and 1024x768 on the
> 3400+, and CPU-bound at all resolutions with the same graphics card
> and an AMD Athlon XP processor. "
>
> Unless your vid card is pure crap, a processor upgrade for Unreal
> engine based games is *always* the best bang for the buck.  Doom was a
> *completely* different story and really needed a better vid card to
> play well.