Anyone buy that Geforce4 Ti card kit?
Anyone buy that Geforce4 Ti card kit?
I have a 933 tower w/ geforce4 MX, and was thinking about picking up one of these, even though the price is still a bit steep. Anyone pick one up? how big of a performance gain did you see over your old card (specificly an MX) with games, osx, adobe appz, 3d appz, video appz?
thanks
I have a 933 tower w/ geforce4 MX, and was thinking about picking up one of these, even though the price is still a bit steep. Anyone pick one up? how big of a performance gain did you see over your old card (specificly an MX) with games, osx, adobe appz, 3d appz, video appz?
thanks
Comments
I saw a general increase in responsivness, I'm not saying its night and day but there was a difference non the less.
If you're a gamer you'll see a bigger difference.
I also maxed out the ram to 1.5gig's at the same time.
[code]
GeForce 4MX:
\tQuartz Graphics Test\t119.02\t
\t\tLine\t129.67\t3.30 Klines/sec [50% alpha]
\t\tRectangle\t118.16\t8.31 Krects/sec [50% alpha]
\t\tCircle\t127.18\t2.93 Kcircles/sec [50% alpha]
\t\tBezier\t117.77\t1.28 Kbeziers/sec [50% alpha]
\t\tText\t102.34\t1.72 Kchars/sec
\tOpenGL Graphics Test\t125.70\t
\t\tSpinning Squares\t125.70\t87.97 frames/sec
\tUser Interface Test\t120.05\t
\t\tElements\t120.05\t38.42 refresh/sec
\t
GeForce Ti4600:
\tQuartz Graphics Test\t119.69\t
\t\tLine\t125.91\t3.21 Klines/sec [50% alpha]
\t\tRectangle\t119.22\t8.39 Krects/sec [50% alpha]
\t\tCircle\t129.62\t2.99 Kcircles/sec [50% alpha]
\t\tBezier\t118.93\t1.29 Kbeziers/sec [50% alpha]
\t\tText\t104.76\t1.76 Kchars/sec
\tOpenGL Graphics Test\t120.46\t
\t\tSpinning Squares\t120.46\t84.30 frames/sec
\tUser Interface Test\t120.01\t
\t\tElements\t120.01\t38.40 refresh/sec
</pre><hr></blockquote>
As for games, with Quake III at 1280x1024, FPS for the 4MX was 84, the Ti 4600 got 98. At lower resolutions the difference was negligible. For example at 640x480 both cards got 106 FPS. Of course at lower resolutions the graphics card isn't really being pushed.
My advice, save your money and use it for something else like more memory, bigger HD, etc.
[ 12-17-2002: Message edited by: RussS ]</p>
Do you think the Ti is getting limited some how by osx or the hardware. Cuz it really should blow the MX away. The mac upgrade market will really die if the cards cost $400 bucks and don't do squat.
[ 12-17-2002: Message edited by: Ti Fighter ]
[ 12-17-2002: Message edited by: Ti Fighter ]</p>
I get 100+ from mine when I play T2 (this is a PC tho.. eeeee) This concerns me somewhat as I have already decided that I am going back to a mac when I purchase my next computer, and I am a keen gamer.
The Ti4600 is a supiror card on more levels than the MX... Mine has 128Megs of ram, which will help with hires textures etc.
But you might want to wait for the nv35 based chipset. This will push the nVidia range of cards (to be called geForceFX) past the ATI Radion 9700 pro's.
If anyone knows of software to benchmark PC and mac under opengl, I would like to compare some resutls.
[ 12-17-2002: Message edited by: Gargoyle ]</p>
I'm guessing the hardware is limiting the performance somewhat, but that doesn't explain how the Ti 4600 could be slower than the 4MX in the OpenGl test. It should have at least matched it.
Gargoyle: [quote] You only get 98fps on q3 with your ti4600 ???? very odd!!!
I get 100+ from mine when I play T2 (this is a PC tho.. eeeee) This concerns me somewhat as I have already decided that I am going back to a mac when I purchase my next computer, and I am a keen gamer. <hr></blockquote>
I hear ya! I have a PC, Athlon XP 1700 (1.47GHz) with a Ti 4600. In Quake III with the same settings it gets 153 FPS. No surprise really, it is an extra 537MHz faster than the Quicksilver 933! If we had Macs at least that fast (we can only hope
With all that said, I still prefer to use the Mac for just about everything, I very rarely even turn on the PC.
[ 12-17-2002: Message edited by: RussS ]</p>
<strong>hmmm, that sucks. I had a feeling that would be the case. Although the Ti is supposed to have all those special shaders that the MX doesn't have, so the game should look better then the MX even if the fps is not that much faster, but an older game like Q3 wouldn't take advantage of that technology. Regardless I'm more of a console gamer anyway, so i was looking for an increase in performance in the OS and graphic appz esp illustrator 10's slow ass.
Do you think the Ti is getting limited some how by osx or the hardware. Cuz it really should blow the MX away. The mac upgrade market will really die if the cards cost $400 bucks and don't do squat.
[ 12-17-2002: Message edited by: Ti Fighter ]
[ 12-17-2002: Message edited by: Ti Fighter ]</strong><hr></blockquote>
I don't think one can expect much from a graphics card update with any OS... But Pixel and Vertex Shaders are another thing completely, and add cool new features to games... (Wow, in 2005-2007 Apple will probably actually make use of Pixel and Vertex shaders in Quartz/Whatever.)
Meaning, Geforce 4MX to Geforce 4 TI shouldn't matter that much when Quartz is concerned these days, and thus won't improve your Word Processing, Web Browsing, or iTunes. (Come on, you already know that one...)
- - - -
What will Vertex and Pixel shaders be used for? Well, you'll be able to draw a well lit and shadowed hair ball being affected by dynamic winds... (um, and that's useful how?) Enter stage left, Geforce FX 3.
<strong>
With all that said, I still prefer to use the Mac for just about everything, I very rarely even turn on the PC.
[ 12-17-2002: Message edited by: RussS ]</strong><hr></blockquote>
Ain't that the truth
Is this ok to do this?
You may want to hold off on that nvidia card till Macworld where ATI will most likely announce the Radeon 9700Pro and Radeon 9500Pro. Both of these cards will give you more bang for your dollar as they outclass and outperform anything nvidia.
But it's not the hardware that is the problem here it's software...you see in os x there are no drivers for these cards the MX and Ti....why apple won't make them i don't know but what i do know is that the cards run off the GF MX 2 drivers.....if you search your hd in os x you will see those are the only drivers. Till Apple comes out with the right drivers we will all see slow FPS from these cards as the software wasn't made for them.
Frank_t
At lower definitions the benchmarks are based upon the peformance of the CPU, but at higher resolutions the graphic card become essential.
Your test showed a difference at 1280 per 1024 it will be bigger at1600 per 1200 : you will see a difference (i expect : in the contrary, Apple sucks really). You will notice also that high end graphic card make a big difference only if the CPU is powerfull. Put a radeon 9700 pro on a pentium 800 and it will give you bad results.
These mid-results are due to the lack of power of the CPU : urge to wait until the release of the PPC 970.
In the PC world there is a great difference between a geforce 4TI (even a 4200) and a geforce 4Mx in all benchmarks if the complexity of the scene is enough.
RussS try runing quake 3 in os 9 with that card it should be faster....i Have a new Dual 867Mhz and it comes with the GF4 MX.
But it's not the hardware that is the problem here it's software...you see in os x there are no drivers for these cards the MX and Ti....why apple won't make them i don't know but what i do know is that the cards run off the GF MX 2 drivers.....if you search your hd in os x you will see those are the only drivers. Till Apple comes out with the right drivers we will all see slow FPS from these cards as the software wasn't made for them. <hr></blockquote>
I guess I should have mentioned it before, The Quake III tests were run in OS 9!
As for drivers in OS X I've always heard that Nvidia makes one unified driver that supports all their cards. I know ATI's cards all had different drivers and heard that recently they also switched to using a unified driver. I'm no expert here, just what I've been told.
[quote]Powerdoc wrote:
If the geforce TI is no more performant than a geforce 4 mx : there is only an explanation : mac os X sucks or you did not use a sufficiant complex image. <hr></blockquote>
I'm guessing you are referring to the OpenGL test here. The graphic used was part of Xbench's test. I didn't choose it. Still doesn't explain why the Ti was slower than the 4MX.
[quote]Powerdoc wrote:
At lower definitions the benchmarks are based upon the peformance of the CPU, but at higher resolutions the graphic card become essential.
Your test showed a difference at 1280 per 1024 it will be bigger at1600 per 1200 : you will see a difference (i expect : in the contrary, Apple sucks really). <hr></blockquote>
That's why I said that at lower resolutions the graphics card wasn't being pushed. I should have properly said that at low resolutions it's the CPU that's doing most of the work.
Agree. 1600x1200 would show a bigger increase, but I'm using a 17" Apple LCD and the highest resolution it supports is 1280. A 14 FPS increase with the Ti at 1280 shows it starting to perform.
[quote]Powerdoc wrote:
You will notice also that high end graphic card make a big difference only if the CPU is powerfull. Put a radeon 9700 pro on a pentium 800 and it will give you bad results.
These mid-results are due to the lack of power of the CPU : urge to wait until the release of the PPC 970. <hr></blockquote>
Yeah, I know that the CPU is the limiting factor here. My other post showing the difference in FPS between the G4 933 and the AMD 1.47Ghz shows to me that if we had a G4 in the 1.5 Ghz range that the FPS would be right up there. An extra 500+ MHz would make a big difference!
Note: Since updating to OS X 10.2.3, Xbench has shown a performance increase in the Quartz Graphics tests while strangely showing a decrease in the OpenGL test. I'll post the new results here as soon as I get a chance.
Russ
[ 12-23-2002: Message edited by: RussS ]</p>
OS X 10.2.2
[code]
Quartz Graphics Test\t119.69\t
\t\tLine\t125.91\t3.21 Klines/sec [50% alpha]
\t\tRectangle\t119.22\t8.39 Krects/sec [50% alpha]
\t\tCircle\t129.62\t2.99 Kcircles/sec [50% alpha]
\t\tBezier\t118.93\t1.29 Kbeziers/sec [50% alpha]
\t\tText\t104.76\t1.76 Kchars/sec
\tOpenGL Graphics Test\t120.46\t
\t\tSpinning Squares\t120.46\t84.30 frames/sec
\tUser Interface Test\t120.01\t
\t\tElements\t120.01\t38.40 refresh/sec</pre><hr></blockquote>
OS X 10.2.3
[code]
Quartz Graphics Test\t124.79\t
\t\tLine\t136.64\t3.48 Klines/sec [50% alpha]
\t\tRectangle\t125.45\t8.83 Krects/sec [50% alpha]
\t\tCircle\t137.17\t3.16 Kcircles/sec [50% alpha]
\t\tBezier\t120.65\t1.31 Kbeziers/sec [50% alpha]
\t\tText\t104.06\t1.75 Kchars/sec
\tOpenGL Graphics Test\t117.27\t
\t\tSpinning Squares\t117.27\t82.06 frames/sec
\tUser Interface Test\t122.40\t
\t\tElements\t122.40\t39.17 refresh/sec </pre><hr></blockquote>
Everything tested faster, well except for OpenGL. Don't know what's going on there.
Russ
[ 12-24-2002: Message edited by: RussS ]</p>
<strong>
You may want to hold off on that nvidia card till Macworld where ATI will most likely announce the Radeon 9700Pro and Radeon 9500Pro. Both of these cards will give you more bang for your dollar as they outclass and outperform anything nvidia.</strong><hr></blockquote>
I hope so on both counts