Winbench 99 Benchmarks
We put our dual Celeron 400 and TMR-006 setup through the Winbench 99 suite of business benchmarks:
| Winbench99 |
| CPUMark99 |
Score |
| Dual PIII Xeon 500, 1024K (NT) |
44.1 |
| Dual PIII Xeon 500, 512K (NT) |
41.4 |
| PIII 500, (NT) |
38.3 |
| Single Celeron 400 TMR-006 (Win98) |
33.1 |
| Single Celeron 400 TMR-006 (NT) |
32.9 |
| Dual Celeron 400 TMR-006 (NT) |
32.9 |
| Dual P2 400 (NT) |
31.5 |
| Winbench99 |
| Business Diskmark99 |
Score |
| Dual Celeron 400 TMR-006 (NT) |
3250 |
| Dual P2 400 (NT) |
3180 |
| Dual PIII Xeon 500, 512K (NT) |
3050 |
| PIII 500 (NT) |
3050 |
| Single Celeron 400 TMR-006 (NT) |
3010 |
| Dual PIII Xeon 500, 1024K (NT) |
2990 |
| Single Celeron 400 TMR-006 (Win98) |
2610 |
CPUMark measures the speed of the processor subsystem. This includes the CPU, its internal cache and external cache, and system RAM. Thus, we didn't expect to see any substantial gaps in scores in this area between similar CPUs, like the dual PII 400 setup and the dual Celeron 400 setup. Of course, the Xeon, with full speed L2 cache in either the 1MB or 512K varieties are powerhouses.
The Business Diskmark scores cover a wider range. Our dual Celeron setup did well here, aided perhaps by the fresh installation of everything on the hard drive, which would have less fragmentation than if it was a hard drive that had more files and applications on it. Additionally, we see a substantial gap between the single Celeron 400 in NT 4.0 and the single Celeron 400 in Windows98, which is due to the difference in the file systems. NTFS, NT's file system, is more efficient, letting it get some points over Windows 98 and the FAT file system.