The first thing we'll check is the audio. We downloaded and installed Audio Winbench to test its CPU utilization.
CPU utilization was fairly high throughout the DirectSound3D tests. CPU utilization never got above 16.1%, but it did average in the 9% - 13% range, which is a lot higher than the <2% averages we've seen with the nForce2. It's not the end of the world, given that the slowest CPU you could be using is a 2.8GHz "E". Granted this is a new line of Audio Chipset and Drivers, I am sure the numbers will creep lower as the drivers mature. This is a synthetic benchmark though, and since I know all of you enjoy a game or two, let's see how the sound will affect UT 2004 performance.
We built a batch file to run a Timedemo on br-colossus, a fairly heavy duty map. The first run is with sound enabled at 640x480 minimum details to force the CPU to handle most of the work, we then run the same configuration with sound off.
UT2K4 is a real system killer, and can bring many systems to their knees. This is an excellent test that shows the amount of drain the audio chipset has on the system when the CPU is being heavily tasked. Since the systems themselves are very different from each other, comparing frame rates like for like is pointless, but notice the differences between sound on and sound off between the three.

The P5GDC performed well as we noticed only an 11% drop in frame rate, this was at stock speeds or overclocked. By contrast, the P5AD2 Premium scored a 20% reduction in frame rate, and the AI7 a whopping 22% decrease. Looks like the newer Audio chipsets are taking on more of the load. Note that you will not see a 22% or even 11% increase in performance if you disabled sound, typically you would run a game in such a manner that the GPU would be handling everything it that it can, so the audio subsystem would have minimal effect on game play.
Hard Drive Performance
In this test, the P5GDC is using the ICH6R along with the P5AD2, whereas the AI7 is using the ICH5R. For this test we use HD Tach Version 3.010 and 2.61.
Left to right, ASUS P5GDC, Abit AI7 and ASUS P5AD2
Although it is close, there is a definite advantage to the ICH6R over the ICH5R. The P5GDC comes in at 13.9ms response with an average read speed of 49.2MB/s, where as the AI7 is 14.3ms with an average read speed of 47.7MB/s. The P5AD2 was tested with a 120GB drive, and although was outperformed by both motherboards here, I would imagine using the same drive would have given similar results as the P5GDC.
Network Performance
We used DU Meter to test the networking speed, and Windows Task Manager for CPU usage. We copied a variety of install files, varying in size of 300kb to as much as 150MB per file totaling 758mb and an ISO image totaling 761mb from the P5GDC Deluxe machine to my Foxconn box (it is the only other GB NIC I have). In between the two I positioned an Enterasys Networks E1 with 1000BaseTX SFP modules. I used the switch because I wanted to make sure that one of the NIC’s was not limiting the other. The Enterasys Networks Switch has built in buffers, so therefore you would be hard pressed to overdrive even the 1000BaseTX port. Here are the results from both an upload and download (as seen from the P5GDC Deluxe) of the same files.
Small Files Test
|
Time to Copy
|
Avg Transfer mB/sec
|
CPU %
|
ASUS P5GDC Upload
|
21.9
|
35.89
|
9
|
ASUS P5AD2 Upload
|
22.2
|
35.17
|
9
|
Foxconn Upload
|
22.3
|
34.92
|
23
|
ASUS P5GDC Download
|
27.2
|
27.94
|
12
|
ASUS P5AD2 Download
|
27.3
|
27.92
|
13
|
Foxconn Download
|
27.5
|
26.27
|
26
|
Large File Test
|
Time to Copy
|
Avg Transfer mB/sec
|
CPU %
|
ASUS P5GDC Upload
|
20.3
|
38.43
|
9
|
ASUS P5AD2 Upload
|
21.4
|
37.87
|
8
|
Foxconn Upload
|
21.4
|
37.82
|
19
|
ASUS P5GDC Download
|
24.5
|
32.76
|
11
|
ASUS P5AD2 Download
|
24.6
|
32.58
|
11
|
Foxconn Download
|
24.6
|
32.33
|
23
|
The AI7 could not be compared in this testing as it only has a 10/100mb interface, therefore in this test and this test alone, I am comparing with the Foxconn 875P chipset motherboard with a 1000BaseTX interface. The P5GDC and P5AD2 are very close in performance and CPU utilization. Any variance I would attribute to Hubert not being able to use a switch instead of a crossover cable. Interesting to note on this test is the CPU utilization the 875P chipset endured. Here is where PCIe starts to shine over PCI, if it had been playing audio the results would have been even more drastic.
Final Words
has given us a great board to play with without making us take out a loan to pay for it, yet with all of the built in features and Overclockabiltiy, this motherboard is anything but ‘cheap’. The overall package is very good, with plenty of extras included, although the box itself comes across as a ‘no frills’ style. Considering the overall package and the performance of the board itself, this board is anything but no frills.
I enjoyed the layout and attention to detail to things as minute as positioning of the headers and memory slots to reduce interference, allowing someone as messy as I am to route cables neatly. ASUS also have a solution to keep the PWM cooler integrated on the board, and with the Prescott’s working the motherboards hard, this is a welcome sight.
The Intel ICH6R provides one regular IDE and 4 SATA connections, along with RAID options, which should be plenty for most folks. You also have the option of a further two IDE ports from the ITE 8212F IDE Controller which again provides RAID options. 8 USB 2.0 and 2 Firewire ports are present as well as the obvious legacy ports such as parallel and the PS/2 ports.
Onboard Audio has steadily been improving over the past few years and the HD Audio on this board is second to none. The Audio ports themselves are very flexible, with you being able to assign functions to the port via software, and also provide you with 7.1 channels.
Although a Socket 478 Springdale (one of the highest performing ones at least) was able to maintain pace with the , I am sure we are only a driver version or two away from that not being the case any longer. Overclocking went very well, producing a 250 + FSB, although I personally found that DDRII provided a higher overclock than DDR.
With the ability to upgrade yet maintain your current DDR, this motherboard provides enough flexibility to make it well worth a look. You will need a new graphics card since the 915 Grantsdale chipset doesn’t support AGP, replacing it with PCIe, but this statement rings true for the majority of 915 boards (discounting the AGPe equipped boards).

Pros: Great Performance at a reasonable price, Audio is 2nd to none, Layout is well designed, Great Overclockability (and hopefully more so with new drivers), 1000BaseTX on a PCIe bus with great performance
Cons: ASUS N.O.S was temperamental and in fact worked properly for only a short time, Have to replace CPU / Video and likely PSU to get it going
Bottom Line: As the PCIe drivers mature, I have a distinct feeling this motherboard will not only rock, but could be the one to beat on the 915P line, in fact I think the 925P doesn't have much on it other than the dual NIC's. If you must go LGA-775 with PCIe, then you MUST go with the !!
HOME