Date:
Tuesday , September 14, 2004
Author:
Morry Teitelman
Editor:
Kyle Bennett
Google +1

MSI K8T Neo2-FIR

MSI has chosen to lead with the much touted K8T800Pro chipset for their AMD Athlon64 motherboard. We test its stability, speed and enthusiast features.

continued...

Subsystem Testing

Audio – Subjective Listening

One of the easiest ways to determine the quality of the audio subsystem is via a subjective sound test. Ideally, a sound test requires audio covering the entire spectrum, from subtle to intense. For this test, I chose to listen to selected tracks off of the Evanescence album Fallen.

I was impressed with the K8T Neo2’s audio reproduction. All sound produced was crisp and clear, without any detectable distortion.

Audio – Microphone Port Testing

The MIC-IN input was tested using a Labtec Desk Mic 524. Spoken word was recorded and played back using Microsoft Sound Recorder, with the Microphone Boost option disable and enabled. The Microphone Boost option is found within the Advanced menu under the Microphone section within the Volume Control menu.

The microphone pickup tests were good, as well. The quality of pickup seemed to be a bit better with Microphone Boot enabled, though. As long as your MIC record volume is set correctly, you will have no problems with this solution.

The thing to keep in mind here is that an onboard sound system is simply going to use more resources than a stand alone sound card as with the onboard sound the CPU is doing a lot of work in relation to sound.

Audio – In Game Testing

In addition to CD or MP3 playback, users most often rely on the audio subsystem for gaming, whether it be for stand alone first person shooters or online death matching. To adequately test the quality of the audio subsystem during game type scenarios, we took benchmark measurements with sound enabled and disabled using the following benchmarks: Jedi Knight 2 and Comanche 4.

Article Image

After the superb audio recording and playback tests, the in game performance felt a bit lacking. In both the games, performance suffered by at least 15%, with the performance in the OpenGL based Jedi Knight 2 suffering just shy of 20%. This will not affect game play too much, as long as your FPS (frames per second) count stays above 40. All in all, MSI did their best to package the K8T Neo2 with a well rounded audio subsystem.

Drive Performance

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. SATA and IDE drive tests were performed using a Maxtor 40Gb ATA 133 model 6E040L0 hard drives on the IDE headers and Seagate 80 GB Barracuda SATA hard drives on the SATA headers. The SATA drives were used for testing in RAID 0 configurations on the VIA VT8267 and Promise RAID controllers. SATA and IDE drives were also used for mixed RAID 0 configurations using the Promise RAID controller. Testing was also conducted using stand alone SATA drives on both the VIA VT8267 and Promise controllers, and IDE drives connected in a primary slave configuration and in standalone fashion on the Promise controller. All drive benchmark was done using the open source Iometer program. Note that that Promise RAID controller supported a minimum block size of 32k, not the desired 16k size.

Article Image

Article Image

Article Image

In a surprising turn of events, the Promise based SATA RAID 0 array with 32k block size outperformed the VT8327 based SATA RAID 0 array with 16k block size across the board. The surprising fact is that in practice, a large RAID 0 block size usually translates in to slower overall performance. I suspect that if the Promise controller supported 16k block size, the RAID 0 array’s performance would have been tough for the VIA controller to match at all. Both external drives performed on par with expectations, with the IEEE 1394 drive slightly besting the USB 2.0 connected drive. All in all, MSI did a masterful job in selecting and implementing the drive controllers for the K8T Neo2.

Network Utilization Tests

Hagel Technologies’ DU Meter software was used in conjunction with Windows Task Manager to measure the performance of the onboard VIA Gigabit NIC in the rear panel. DU meter was used to measure bandwidth, with Windows TaskMan to monitor the CPU utilization on the test system. For the test itself, a 750Mb archive file containing various sized .WMA audio files for the large file transfer test and a 750Mb worth of various sized .WMA audio files for the small files transfer test were used in conjunction with an Intel PRO/1000 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.

Article Image Article Image Article Image Article Image

The small file transfer tests results were lower than expected, with an average download speed of just under 25 MB/s. The upload speed lagged the download speed by approximately 5 MB/s. The CPU utilization averaged close to 10% during uploads and 25% during downloads.

Article Image Article Image Article Image Article Image

The large file transfer results were a duplicate of the small file transfer results, with similar lackluster upload and download speeds and CPU utilization numbers. Not the best performing Gigabit NIC, but acceptable nonetheless.

WiFi Testing

Testing was also done with the onboard MSI WiFi-g solution, with file transfers being done between it and a 54G capable Netgear WiFi router. In both upload and download testing, the average transfer rate averaging between 1.7 - 2.2 MB/s, which is more than acceptable considering there were no wires involved. Note that in testing, it took approximately 6.5 minutes to transfer a 750 MB file between the local and remote systems during both upload and download scenarios and CPU utilization averaged 20% while the transfer occurred. For sustained transfers, the wireless solution would seem slow, but for web browsing and other low bandwidth tasks, you would not even notice the speed difference.

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 K8T Neo2-FIR (VIA K8T800Pro) - Athlon64 3800+, Athlon64 3500+, 1GB (2x512MB) Corsair XMSDDR3200LLPro DDR400 (2,3,2,5), NVIDIA NV40 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

ABIT AV8 (VIA K8T800Pro) - Athlon64 3800+, Athlon64 3500+, 1GB (2x512MB) Corsair XMSDDR3200LLPro DDR400 (2,3,2,5), NVIDIA NV40 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

Asus A8V Deluxe (VIA K8T800Pro) - AthlonFX-53, Athlon64 3800+, Athlon64 3500+, 1GB (2x512MB) Corsair XMSDDR3200LLPro DDR400 (2,3,2,5), NVIDIA NV40 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

ABIT AA8 (i925X) - Intel 3.6GHz (model # 560), Intel 3.4GHz Extreme Edition, 1GB (2x512MB) Micron Engineering Sample DDR2-533 (4,4,4,12), NVIDIA NV45 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

ABIT AG8 (i915P) - Intel 3.6GHz (model # 560), Intel 3.4GHz Extreme Edition Intel, 1GB (2x512MB) Corsair XMSDDR3200LLPro DDR400 (2,3,2,5), NVIDIA NV45 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

Intel D875PBZ (i875P) - Intel 3.4GHz Prescott, Intel 3.4GHz Extreme Edition, 1GB (2x512MB) Corsair XMSDDR3200LLPro DDR400 (2,3,2,5), NVIDIA NV40 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

For Doom3 benchmark tests, the following system configuration was used (in addition to the MSI K8T Neo2 system):

DFI 925X-T2 (i925X) - Intel 3.6GHz (model # 560), 1GB (2x512MB) Micron Engineering Sample DDR2-533 (4,4,4,12), NVIDIA NV45 (400/550) (v61.45 driver), 2 x 36GB Western Digital Raptors RAID 0, Windows XP w/SP1 and DX9B.

Our two legacy systems will of course not be kept up to date with the latest technology, but they do both use DDR400 and ATA133 IDE hard drives and ATI 9800XT video cards.

ABIT NF7 (NVIDIA NF2) - AthlonXP 2400+ 133MHz System bus, 1GB (2x512MB) Corsair XMSDDR400Pro DDR400 (2,3,2,5), ATI 9800XT (ATI Catalyst 3.8 drivers), 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

Intel D875PBZ (i875P) - Pentium 4 2.4C - Intel, 1GB (2x512MB) Corsair XMSDDR400Pro DDR400 (2,3,2,5), ATI 9800XT (ATI Catalyst 3.8 drivers), 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

Graphs are labeled as follows: Motherboard - CPU Clock - FSB Clock - Memory Clock

SiSoft Sandra Memory Bandwidth Benchmark

Article Image

Note that all results above were obtained running the installed memory in Dual Channel mode.

The K8T Neo2 makes a strong showing in keeping pace with the other Athlon64 boards, but does little to differentiate itself from them performance-wise. Certainly from looking at the numbers, it is doing "what it is supposed to be doing" and looks no different from our ASUS or ABIT board.