System setup and notes
Hardware and Software
Test Platforms
| ECS PF5 Extreme System | Intel i955X Express System | MSI P4N Diamond System | Intel i945G System | Athlon 64 S939 PCI-Express System | |
| Processor(s) | Intel Pentium 4 660 | Intel Pentium 4 660 | Intel Pentium 4 660 | Intel Pentium 4 660 | AMD Athlon 64 4000+ |
| Mainboard | Elitegroup PF5 Extreme | Intel D955XBK | MSI P4N Diamond nForce4 SLI I.E. | Intel D945GTP | ABIT Fatal1ty AN8 nForce4 |
| Memory | 1GByte (2x512MB) Corsair XMS2-5400UL | 1GByte (2x512MB) Corsair XMS2-5400UL | 1GByte (2x512MB) Corsair XMS2-5400UL | 1GByte (2x512MB) Corsair XMS2-5400UL | 1GByte (2x512MB) Corsair XMS3200XL | Memory timings | 3-2-2-8 @ DDR533 | 3-2-2-8 @ DDR533 | 3-2-2-8 @ DDR533 | 5-4-4-10 @ DDR500* | 2-2-2-5 1T @ DDR400 |
| Graphics Card | ATI RADEON X850 XT PE - PEG16X - CATALYST 5.3 | Disk Drives | 160GB WD IDE & 36GB SATA Raptor | ||
| BIOS Version | Version 1.0 - 26/04/05 | BK95510J.86A.1452.EB | BIOS 1 - 19/04/05 | NT94510J.86A.0897 | BIOS 13 -25/04 05 |
| Operating System | Windows XP Professional, SP2 | ||||
| Mainboard Software | Intel INF Update Utility 7.0.0.1019 | Intel INF Update Utility 7.0.0.1019 | nForce4 7.02 platform driver | Intel INF Update Utility 7.0.0.1019 | nForce 6.39 platform driver |
Benchmark Software
HEXUS.in-house Cryptography Benchmark
HEXUS Pifast Benchmark
ScienceMark 2.0 (7th February 2005)
Realstorm Raytracing 2004
CINEBENCH 2003 multi-CPU render
HEXUS.in-house MP3 Encoding Benchmark using LAME 3.97a (Intel HT compiler) - 701MB WAV
picCOLOR 32-bit v4.0
KribiBench v1.1
Microsoft Movie Maker 2.1
Simpli Software HDTach
3DMark2001SE b330
DOOM 3 b1282
Far Cry v1.3
Notes
As mentioned before, the board suffered from an inability to run a 1GB pack of Corsair XMS2 5400UL memory at their 3-2-2-8 native timings.
The above CPU-Z shot shows RAM running at ~DDR2-500 with 5-4-4-10 timings. That's a direct result of inputting the DDR2-533 setting from within BIOS. A comparative lack of bandwidth and poor latencies will undoubtedly impact upon performance. Let's now see just how much.
The running speed of the boards was as follows:
3600.2MHz - Intel D945GTP - Intel i945G - Intel Pentium 4 660
3600.2MHz - Intel i955XBK - Intel i955X Express -Intel Pentium 4 660
3599.4MHz - MSI P4N Diamond - nForce4 SLI I.E. - Intel Pentium 4 660
3591.1MHz- ECS PF5 Extreme - Intel Pentium 4 660
2411.2MHz - ABIT Fatal1ty AN8 SLI - nForce4 SLI - AMD Athlon 64 4000+
Overclocking
FSB overclocking was conducted by using a semi-unlocked Pentium 4 570J. The multiplier was knocked down to 14x, memory speed reduced to run synchronously to the FSB, CPU and NB voltage raised to 1.5v and +0.15v, respectively, and FSB speeds raised until system failure set in. We were able to reach a stable FSB speed of 312MHz (1248MHz QDR). An overall clock speed of 4368MHz may have been the limiting factor here, and it's clear, on this sample at least, that ECS' claim of FSB1200, etched right on to the PCB, holds true.