Inno3D Home Page Inno3D Home Page

FAQ News Search Archive Forum Articles Tweaks Technology Files Prices SETI
Visit NVIDIA's home page.
Favorite Pics
Click to Enlarge
Articles/Reviews
OCZ Tech Titan 3
Absolute MORPHEUS
1.0GHz Pentium III
eVGA MX Shootout
nForce Preview
AMD AXIA CPU
VisionTek GeForce3
2001 Spring Lineup
GeForce3 Preview
eVGA TwinView Plus
VisionTek GF2 Ultra
OCZ GeForce2 Pro
Winfast GeForce2 MX
GeForce2 vs Quake3
Linksys Cable Router
GF2 FSAA Shootout
GeForce2 MX Preview
Benchmarking Guide
Quake 3 Tune-Up
Power DVD Review
Live! Experiences
Sponsors
Memory from Crucial.com


FastCounter by bCentral

 Visitors Are Online
Powered by Perlonline.com
Drivers/FAQ
NVIDIA
3D Chipset
Gamers Ammo
Reactor Critical
GeForce FAQ
Associates
Dayton's Misc.
G-Force X Sweden
Maximum Reboot
Media Xplosion
NVchips-fr
nV Italia
Riva Station
3D GPU
nV News Home Page

OCZ GeForce2 PRO 64Meg Review - Benchmarks and Gaming

By: Brian Gray - December 10, 2000

Skip To:

Test System

Here is the rundown of the gear used in testing.

The Grays are coming... Alienware Area 51:Aurora (Modified)

  • AMD Duron 1000m (OCZ)
  • Abit KT7 KT133 motherboard (OCZ)
  • 256MB PC133 Virtual Channel CAS2 SDRAM
  • 15GB ATA66 7200rpm HD
  • Sound Blaster Live! X-Gamer
  • Linksys 10/100 NIC
  • MidiLand S4 8200 5.1 Sound System
  • Martian Red and black color scheme  :þ

The formal benchmarks include Quake3 Arena and 3D Mark 2000. They are used so you have a metric to compare this card's raw performance to benchmarks in other reviews. More importantly, I will share experiences of running several other games using the OCZ Pro, as opposed to an Ultra.

Also keep in mind that all benchmarks are at 235MHz core and 445MHz memory clocks.

OpenGL

Quake3 Arena continues to be the most widely accepted benchmark for OpenGL. Why derail the train if its on course so...

Quake3 Arena Settings

  • Geometry and Textures - Max Settings
  • Texture Compression - Disabled
  • V-Sync - Off
  • Trilinear Filtering - Enabled
  • Anisotropic Filtering - Enabled
  • High Quality Sound - Enabled

Quake3 Arena Results

No FSAA 16bpp 32bpp
1024x768 100.0 80.8
1280x1024 74.2 59.7
1600x1200 55.4 39.2
1.5x1.5 FSAA    
800x600 82.1 58.3
1024x768 57.2 38.8
2x2 FSAA    
800x600 55.0 40.9

Enabling texture compression will increase scores by a little less than 10% on average, but with 64MB of RAM on board, I found myself ignoring S3TC and enjoying the impressive visuals. Since 1600x1200x32bpp is not a viable resolution for gaming, even with S3TC enabled, I found myself using 1280x1024x32bpp the most, followed closely by 800x600x16bpp with 2x2 FSAA.  Here are a couple of screens.

Quake3 Arena: 1280x1024x32bpp

Full Size 167KB

I had a field day sniping people on this Railgun only server. The hardware advantage here definitely makes up for my lack of skill when using the Railgun.

Quake3 Arena: 800x600x16bpp 2x2 FSAA

Full Size 91KB

16bpp color precision is helped at this setting by the four samples taken per pixel. I have not verified this with NVIDIA, but myself and many others on the web believe that NVIDIA no longer writes 16bpp 2x2 FSAA to a 16bit buffer. Instead, they leave the Z-Buffer wide open to allow additional color precision of taking four samples to remain, around 22 to 24bpp. Is it 32bpp? No, but it looks quite a bit better than 16bpp normally does.

I also played quite a bit of Half-Life and MDK2 with the OCZ card. Half-Life on the Pro, as on the Ultra, looked best at 1024x768x32bpp with 2x2 FSAA with constant framerate around 50fps with dips to the 30's but nothing lower, extremely playable despite the high resolution with FSAA enabled.

MDK2 runs very wide open as well. The T&L engine does a fine job of boosting visuals through more complex lighting, and at 1600x1200x32 the Pro pulled down 49fps average capped by fillrate limits at the top end with very few dips into the 20's, acceptable in MDK2's less frenetic gameplay.

Direct 3D

For Direct3D, I have a couple of 3D Mark 2000 benchmark runs. Again, I can not stress enough to ignore the composite score and focus on individual numbers like fillrate and framerate to compare results. Use the CPU Mark to compensate for your CPU's speed when comparing the individual scores.

3D Mark 2000 - 1024x768x16bpp

3D Mark 2000 - 1024x768x32bpp

The fillrate numbers easily surpass what the standard GTS can do, and land just below the Ultra at default clocks. While my Annhilator2 Ultra beats the OCZ Pro by 800 points in the composite score, the actual framerate scores are fairly close, with 5-10%. The overclocked OCZ Pro nips at the heals of the Ultra.

I also tested the OCZ Pro while playing Unreal Tournament in Direct3D. The Pro took on the 1280x1024x32bpp resolution and color depth with all settings maxed and played on par with the Ultra, coming within 2fps of the Ultra's timedemo, 29.1fps vs.31.8fps, in heavily CPU dependent UTBench demo. During normal play I could not "feel" any difference between an Ultra and the OCZ Pro. FSAA performance in UT was not the same story. The fillrate advantage of the Ultra is just too much to overcome as 800x600x32 with 2x2 FSAA did bog down to the low twenties on the Pro more often than on the Ultra.

In both Test Drive 6 and Mercedes Truck Racing, I could use just about any setting with fillrate requirements similar to 1600x1200x32bpp and play very smooth. 800x600x32bpp with 2x2(special) FSAA or 1024x76x32bpp with 1x2 FSAA both looked impressive.

By the way, the TV-out does work and works well, however I found 640x480 converts better.

Sorry about the quality of the picture...

The 3D Mark 2000 demo is impressive on the big-screen.

Enough Playing Around

Time to put all of this in perspective. Ultra or Pro? Pro or another upgrade?

While I have enjoyed my time with the OCZ Pro, its time to get to the point.

Next Page: Conclusion

Skip To:

 
Last Updated on December 17, 2000

All trademarks used are properties of their respective owners.