![[H]ardOCP](/images/blank.gif){kind=small}
- Date:
- Tuesday , March 11, 2003
- Author:
- Morry Teitelman
- Editor:
- Kyle Bennett
- Google +1
MSI 655 Max Review
This is one of the first boards to market to support dual channel memory mode for the Intel P4 family. Find out how it stacks up against the other fierce competitors...
Subsystem Testing
Audio – CPU Utilization
The sound subsystem is one of the most critical built-in components on a motherboard. A good audio subsystem can turn a good board in to a sought after board, while an under par audio subsystem can cause a great board to be put back on the shelf. One of the better ways to measure the quality of the subsystem is through CPU utilization. The less an audio subsystem uses the CPU, the more power the CPU can focus on other system critical tasks like pushing polygons. In order to best measure the real world CPU utilization of the audio subsystem, we use Ziff Davis’ Audio Winbench.
Overall, the results show that the CPU utilization is < 4% when using the audio subsystem. However, in some cases, the usage spikes above 5%. It seems that MSI picked an adequate audio solution for the board.
Audio – Subjective Listening
A CPU utilization test can tell you a lot of things about the audio subsystem, except how it actually sounds. For this, you need a subjective audio test that pushes the subsystem to its limits of subtlety and intensity. I dug deep in to my Heavy Metal archives and found a recent album to exercise this board, the album Sinner from Drowning Pool.
I was pleased with the sound coming from the board. All 6 channels performed admirably, with no stuttering, pausing or crackling detected at all. I’ll go so far as to say that you would probably be happy using this as your primary audio solution for gaming.
USB 2.0 / IEEE 1394
In order to adequately test the capabilities of the on board USB 2.0 and IEEE 1394 connections, we chose to use an ACOMDATA HD060U2FE-72-USB 2.0/FireWire HDD connected first to the USB port and after to the IEEE 1394 port in conjunction with TCD Labs’ HDTach program.
The results from this read/write test are not superb, but they do not disappoint either. In both cases, the RAT times come in just below 20 ms as expected. The IEEE 1394 shows it’s strength during the read tests, while both connection types show about equal results during the write tests. The read tests are a bit of a surprise, since the theoretical maximum bandwidth of USB 2.0 is higher than that of IEEE 1394. All in all, I am satisfied with MSI’s implementation of both connection types.
IDE/ATA Performance
The IDE performance of a system comes in a close second to memory performance on a scale of criticality. Miserable IDE performance can drag an otherwise fast system in to the dumps, just as a fast IDE controller can boost the performance of a mediocre system. In order to test the IDE performance of this board, I used TCD Labs’ HDTach program. My test bench currently has three Maxtor 40Gb ATA 133 model 6E040L0 hard drives installed. Two of the drives were used to test the following configurations – Raid 0, Raid 1, IDE1 slave. Furthermore, the RAID 0 tests were conducted with two drives on the same cable using the IDE3 port and two drives on separate ports, one on IDE3 and the other on SATA1. RAID 1 tests were also conducted with the IDE3/SATA1 port configuration. Due to a limitation of the board, RAID 1 was not available when running 2 drives from the IDE3 port.
Both RAID 0 tests yielded very similar results. I attribute this to the fact that in the mixed mode test, the IDE port set the maximum performance threshold since it was the slower interface of the two. In both tests, the drives achieved a respectable RAT time of just under 15ms, with a RBS speed of just over 100 MB/s. Note that in the mixed configuration, the write tests were not done due to an incompatibility between the SATA converter and the drive being used.
The RAID 1 results mimic those of the RAID 0 same cable tests. Both the RAT time and the RBS speed are solid, coming in at below 15ms for RAT and just over 100 MB/s for RBS.
Out of all the read/write tests, the IDE slave test results trailed behind the RAID results as expected. That is not to say that the results were bad though. With an RAT of a bit over 15ms and an RBS speed a bit under 97 MB/s, the drive performed well. This is especially true given the extremely low CPU utilization.
Network Utilization Tests
Hagel Technologies’ DU Meter software was used in conjunction with Windows Task Manager to measure the performance of the onboard Broadband Gigabit NIC. DU meter was used to measure bandwidth, with Windows TaskMan to monitor the CPU utilization on the test system. For the test itself, a 760Mb archive file containing various sized .WMA audio files for the large file transfer test and a 760Mb worth of various sized .WMA audio files for the small files transfer test were used in conjunction with an Intel Gigabit NIC on the host system, and a crossover cable to connect the host system to the test system. A crossover cable was used to rule out any possible bandwidth losses due to hub or switch passage.
The throughput on this board when using gigabit connections is astounding. The CPU utilization during file download was a bit high, averaging about 40%, but given the rate that packets were flying between the machines, I’m willing to overlook this fact.
The large file tests results mimic those of the small file results. You see the average CPU utilization jump to 40% with downloading, while it stays at 20% during upload. The transfer rates are a bit higher than the small files results, but not by much.
Test Systems
The following system configurations were used for the Sandra memory benchmark graph, as well as all graphs listed under the Application and Gaming Benchmark sections:
MSI 655 Max (SiS 655): Intel Pentium 4 2.53 GHz CPU (clocked at 19x133 for 2.53 GHz); 2 x 256MB Corsair XMS3500 @ DCDDR400; ATI Radeon 9700 Pro w/ ATI Catalyst 2.3 drivers; 40 GB Maxtor ATA133 HDD; Vantec 520w PSU; WindowsXP w/SP1
VIA P4PB Ultra Mainboard; Pentium 4 2.53B CPU; Intel HSF; ATi Radeon 9700 Pro; 2x Corsair 256MB XMS3200 DDR; Maxtor 40GB ATA/133 7200RPM HD; Windows XP Professional; WindowsXP w/SP1, ATI Catalyst 2.3., Hyperion v4.45 (166 & 200MHz memory bus tested)
MSI 845PE Max2 (i845PE): Intel Pentium 4 2.53GHz CPU(clocked at 19x133MHz for 2.53GHz clock speed); 512MB Corsair XMS3200; ATI Radeon 9700 Pro w/ ATI Catalyst 2.3 drivers; 40 GB Maxtor ATA133 HDD; Zalman 300w PSU; WindowsXP w/SP1
ASUS P4G8X (Granite Bay): Intel Pentium 4 2.53GHz CPU (clocked at 19x133MHz for 2.53GHz clock speed); 2 x 256Mb Corsair XMS3200; ATI Radeon 9700 Pro w/ ATI Catalyst 2.3 drivers; 40 GB Maxtor ATA133 HDD; Zalman 300w PSU; WindowsXP w/SP1
ASUS A7N8X (nForce2): AMD AthlonXP 2400+ CPU (clocked at 12x166 for 2.0 GHz CPU speed); 2 x 256MB Corsair XMS3200; ATI Radeon 9700 Pro, onboard nForce2 10/100 NIC; 40 GB Maxtor ATA133 HDD, Zalman 300w PSU. WindowsXP w/SP1, ATi Catalyst 2.3.
Note: The following 2 test systems were used only for comparison in the ZD Business 2002 Winstone and ZD Content Creation 2003 Winstone.
GIGABYTE GA-7VAXP (KT400): AMD Athlon 2400+ CPU (clocked at 12x166 for 2.0 GHz CPU speed); 2 x 256Mb Corsair XMS3200; ATI Radeon 9700 Pro w/ ATI Catalyst 2.3 drivers; 40 GB Maxtor ATA133 HDD; Allied 400w PSU; WindowsXP w/SP1
MSI K7N2G-ILSR (nForce2): AMD Athlon 2400+ CPU (clocked at 12x166MHz for 2GHz clock speed); 2 x 256MB Corsair TwinX XMS3200; ATI Radeon 9700 Pro; 40 GB Maxtor ATA133 HDD; Vantec Stealth 520w PSU. Windows XP w/SP1; ATI Catalyst 3.0; NVIDIA nForce 2.03 Driver Package; DirectX 9; BIOS v1.2
SiSoft Sandra Memory Bandwidth Benchmark
The 655 Max stomps the competition in this memory benchmark, due to its dual channel memory implementation and a very wide Pentium 4 bus that will not bottleneck the system like we see on AMD based boards.