The Benchmarks

The week 21 Celeron 366 was tested with the following:

• Abit BP6 Motherboard

• AMK LAN0334 17" Case

• Diamond Viper V770 Ultra

• 128 Megs Micron PC-133 SDRAM

• Acer 50X CD Rom

• Maxtor 17.2 Gig Hard Drive

• Sound Blaster Live Value

• SmartFast 100 Base Network card

I used Quake 2, Quake 3, Unreal and 3D Mark 99 Max as the benchmarks. All tests were ran 3 times and the average score is shown. All drivers were the latest ones from the manufacturers. The Viper 770 was running at 185/243 using NVIDIA 2.08 drivers. Just for comparison I included the scores from Project 620.

Vsync was turned off. No tweaks or special cfg files were used. 3D Mark 99 Max was ran at 800x600, 16 bit color, 16 bit Z-Buffer, Triple Frame Buffer.

The Quake 3 timedemo was ran on the default "high quality" setting found in the graphic options, which are:

Resolution: 800x600
Graphic mode: HIGH QUALITY
GL Driver: DEFAULT
GL extensions: ON
Color : 32 BIT
Full screen: ON
Lighting: LIGHTMAP
Geometric detail: HIGH
Texture detail: 3
Texture quality: 32 BIT
Texture Filter: TRILINEAR

The game options are:

Marks on walls: YES
Ejecting brass:  YES
Dynamic lighting: YES
Light flares: NO
Identify target: YES
High quality sky: YES

Benchmark Scores

With the exception of 3D Mark 99, the Celeron running at 605Mhz went toe to toe with the Pentium 3 at 620Mhz. This I'm sure we can all live with since we play games and not benchmarks.  However the 3D Mark score does go to show the potential increase in performance by having the SSE instruction set. Unreal ran faster on the Pentium 3 because the game codes of more tuned for Pentium than Celerons.

When you think about it, it is truly amazing that an $85 CPU can be just as fast as a $450 CPU. Also remember that if I wanted to beat the P3-620 at Quake 3 all I have to do is add the second Celeron 366, run Windows NT and get dual 605.

Next Page: Conclusion