Test Setup
DFI LanParty UT 915P-T12: Intel P4 530, 2 x 512 MB Corsair TWINX XMS4000, MSI PCX5750-TD128, 120 GB Seagate PATA 7200rpm (ST3120026A), Windows XP Professional SP1, NVIDIA ForceWare 61.77
Intel D915PSY: Intel P4 530, 2 x 512 MB Corsair TWINX XMS4000 Pro, Albatron PCX5750, 120 GB Seagate PATA 7200rpm (ST3120026A), Windows XP Professional SP1, NVIDIA ForceWare 61.77
The software used to benchmark the 915P-T12 follows:
SiSoft Sandra 2004
Business Winstone 2004
Sysmark 2004
PiFast
CDex 1.51
TMPGEnc 2.521
Unreal Tournament 2003
Quake 3: Arena
Return to Castle Wolfenstein
The stock Intel cooler was used to cool the processor in all the tests. Corsair DDR1 XMS4000 running with 3-4-4-7 memory timings. All scores were run three times, and the average (arithmetic mean) was used for final benchmark scores (except for the SiSoft Sandra test). Intel's Performance Acceleration Technology was set to Ultra.
The comparison board will be an Intel branded D915PSY, armed with an Albatron PCX5750 and 1GB of Corsair TWINX4000 ProSeries ram. While we try to maintain an apples to apples test environment, we feel the different brands of comparable products should have minimal impact on the final scores.
CPU Arithmetic Benchmark
|
|
Whetstone iSSE2
|
Dhrystone ALU
|
|
DFI 915X-T12
|
6258
|
8287
|
|
Intel D915PSY
|
6178
|
7986
|
CPU Multimedia Benchmark
|
|
Integer iSSE2
|
Float iSSE2
|
|
DFI 915X-T12
|
21130
|
28106
|
|
Intel D915PSY
|
21111
|
28079
|
Memory Benchmark w/DDR-1
|
|
Int Buffered iSSE2
|
Float Buffered iSSE2
|
|
DFI 915X-T12
|
4673
|
4679
|
|
Intel D915PSY
|
4670
|
4676
|
The numbers came out right about where they should have for a 3.0 GHz Pentium 4, the memory bandwidth shouldn't matter much between different motherboards running the same RAM, but the CPU benchmarks will usually vary noticeably between motherboards. While the MMX benchmarks are fairly close, the Arithmetic benchmarks are clearly won by the LanParty.
Memory Benchmark DDR-1 vs DDR-2
|
|
Int Buffered iSSE2
|
Float Buffered iSSE2
|
|
DDR-1
|
4673
|
4679
|
|
DDR-2
|
4775
|
4781
|
Using some Mushkin DDR-2, we ran a quick benchmark at DDR533. While there is a 133MHz speed advantage with the Mushkin, there wasn't as much difference as we thought there would be.
ZD Business Winstone 2004
ZD Business Winstone is a synthetic benchmark created by Ziff-Davis to benchmark how well a computer will fare while using business-level applications.
|
|
Score (Higher is better)
|
|
DFI 915X-T12
|
20.2
|
|
Intel D915PSY
|
19.8
|
Not much difference between the two boards here, though the Intel board fails to crack the 20 mark in each of our test runs.
Sysmark 2004 Office Productivity
BAPCo's Sysmark 2004 benchmarks a system by installing professional CPU-intensive applications such as Microsoft Word, Macromedia Dreamweaver, 3D Studio Max, Adobe Aftereffects, as well as many other applications. Each of the applications is then run, BAPCo's automation system takes care of creating content and timing each operation. Depending on the speed of each operation (for example, a Gaussian blur in Adobe Photoshop 7,) a score will be determined for each run.
|
|
Score (Higher is better)
|
|
DFI 915X-T12
|
156
|
|
Intel D915PSY
|
154
|
Sysmark 2004 Content Creation
|
|
Score (Higher is better)
|
|
DFI 915X-T12
|
192
|
|
Intel D915PSY
|
187
|
We're seeing similar results as the earlier tests here, where the LanParty is a bit faster than the Intel board, but only by a little. Given the DFI board offers more options in tweaking, this comes as no surprise.
PiFast
A good indicator of CPU/Motherboard performance is PiFast version 4.2, by Xavier Gourdon. We used a computation of 10000000 digits of Pi, Chudnovsky method, 1024 K FFT, and no disk memory. Note that lower scores are better, and times are in seconds.
|
|
Time in Seconds (Lower is better)
|
|
DFI 915X-T12
|
52.39
|
|
Intel D915PSY
|
53.08
|
PiFast is very memory and CPU intensive, 52.39 seconds is quite respectable for a 3.0 GHz processor.
Keep in mind that the next two tests, CDex and TMPGEnc are bound both by CPU and data transfer speeds.
CDex Audio Conversion: Wav to MP3
CDex was used to convert a 414MB Wav file to a 320kbs MP3. LAME version 1.30 (which is included with CDex.)
|
|
Time in Minutes:Seconds (Lower is better)
|
|
DFI 915X-T12
|
1:20
|
|
Intel D915PSY
|
1:22
|
With the SATA converter, hard drive reads were slower overall, and so were CDex compression times.
TMPGEnc 2.521
We used an Animatrix file, titled The Second Renaissance Part 1, and a WAV created from VirtualDub. The movie was then converted it into a DVD compliant MPEG-2 file with a bitrate of 5000. Times are in minutes:seconds, and lower is better.
|
|
Time in Minutes:Seconds (Lower is better)
|
|
DFI 915X-T12
|
4:08
|
|
Intel D915PSY
|
4:15
|
Another close one, but not as close as you may think if you increase the size of the working video file to something DVD authors may be working with.
NEXT