Asus A8N-VM CSM: NVIDIA GeForce 6150 Finally Arrives
by Wesley Fink on December 1, 2005 12:04 AM EST- Posted in
- Motherboards
Audio Performance
Since this is the first NVIDIA implementation of HD audio, we tested to see how it compared with HD audio implementations from ATI. For audio testing, we used Rightmark 3D Sound CPU utilization test Version 2.1. This benchmark measures the overhead or CPU utilization required by a codec or hardware audio chip. Since we found that Version 2.1 yields different results from the earlier 1.24, we have only included test results with version 2.1.
It is very important to point out that NVIDIA only provides HD audio for high-end AMD Integrated Graphics solution like this A8N-VM CSM board. There is no current NVIDIA HD option for nForce4 or nForce Dual x16 discrete chipsets for AMD processors. We are happy to see NVIDIA offering HD audio on the GeForce chipsets, and we hope that they will soon offer similar or better solutions for their discrete chipsets.
Ethernet Performance
Asus used the hooks in the 430 Southbridge to provide Gigabit LAN on the A8N-VM. Since this was the first time we had seen the Marvell 88E1111 PHY, we decided to test the performance of the on-board Ethernet.
The Windows 2000 Driver Development Kit (DDK) includes a useful LAN testing utility called NTttcp. We used the NTttcp tool to test Ethernet throughput and the CPU utilization of the various Ethernet Controllers used on the AMD motherboards.
We set up one machine as the server; in this case, an Intel box with an Intel CSA Gigabit LAN connection. Intel CSA has a reputation for providing fast throughput and this seemed a reasonable choice to serve our Gigabit LAN clients. At the server side, we used the following Command Line as suggested by the VIA whitepaper on LAN testing:
Since this is the first NVIDIA implementation of HD audio, we tested to see how it compared with HD audio implementations from ATI. For audio testing, we used Rightmark 3D Sound CPU utilization test Version 2.1. This benchmark measures the overhead or CPU utilization required by a codec or hardware audio chip. Since we found that Version 2.1 yields different results from the earlier 1.24, we have only included test results with version 2.1.
None of the onboard audio solutions were quite as low in CPU utilization as the hardware Creative SoundBlaster Live! chip, which we have tested on both AMD and Intel top-end boards from MSI. However, ATI's High Definition Audio on the ADI Soundmax used on the Asus turned in excellent performance in all the utilization tests. Results were in every case almost the same as results from the ATI chipset Asus A85-MVP, which uses the same ADI codec with the ATI Crossfire chipset. Azalia HD is demanding of CPU power, but CPU utilization remained below 5%, even in 3D audio rendering as tested with RightMark 2.1.
It is very important to point out that NVIDIA only provides HD audio for high-end AMD Integrated Graphics solution like this A8N-VM CSM board. There is no current NVIDIA HD option for nForce4 or nForce Dual x16 discrete chipsets for AMD processors. We are happy to see NVIDIA offering HD audio on the GeForce chipsets, and we hope that they will soon offer similar or better solutions for their discrete chipsets.
Ethernet Performance
Asus used the hooks in the 430 Southbridge to provide Gigabit LAN on the A8N-VM. Since this was the first time we had seen the Marvell 88E1111 PHY, we decided to test the performance of the on-board Ethernet.
The Windows 2000 Driver Development Kit (DDK) includes a useful LAN testing utility called NTttcp. We used the NTttcp tool to test Ethernet throughput and the CPU utilization of the various Ethernet Controllers used on the AMD motherboards.
We set up one machine as the server; in this case, an Intel box with an Intel CSA Gigabit LAN connection. Intel CSA has a reputation for providing fast throughput and this seemed a reasonable choice to serve our Gigabit LAN clients. At the server side, we used the following Command Line as suggested by the VIA whitepaper on LAN testing:
Ntttcps -m 4 ,0,‹client IP› -a 4 -l 256000 -n 30000On the client side (the motherboard under test), we used the following Command Line:
Ntttcpr -m 4 ,0,‹server IP› -a 4 -l 256000 -n 30000At the conclusion of the test, we captured the throughput and CPU utilization figures from the client screen.
As you can clearly see, the Marvel 88E1111 PHY performs about the same as the top PCIe Ethernet Gigabit LAN solutions. True PCIe Gigabit LAN is capable of about 35% faster speed than the PCI Gigabit LAN used on some boards. In practical terms, this won't matter to most users, since high-speed internet barely taxes a 10Mb/s connection. The speed difference may be important if you routinely transfer many large files on a full 1 Gigabit network.
56 Comments
View All Comments
highlandsun - Thursday, December 1, 2005 - link
Yes, Googer is missing the point that this is supposed to be a complete *Media Center*, that means audio too. I need SPDIF in/out for my minidiscs and other audio devices. I have a multichannel amplifier but I'd prefer to feed it with a pure digital signal. And yes, the PCI slots are already spoken for (Fusion HDTV tuner).BigLan - Thursday, December 1, 2005 - link
Actually, spdif and HD audio are pretty much mutually exclusive. The main benefit of spdif is to pass a dolby digital or dts audio stream untouched to a receiver from the dvd disk (or .avi file.) HD audio allows the motherboard to do 7.1 sound on the motherboard, which is then sent out of the analogue outputs - the 7.1 sound in games etc doesn't use spdif unless you have a soundstorm2 board or certain soundcards which do dolby digital encoding.For most HTPCs, HD audio doesn't actually do anything.
Googer - Thursday, December 1, 2005 - link
If i am not mistaken, ATI's chipset allows for 32 or 64MB of dedicated video RAM to be soldered in to the motherboard. This prevents it from having to resort to system memory.USAF1 - Thursday, December 1, 2005 - link
Hmm... I could be mistaken, but I'm pretty sure that the Marvell 88E1111 "Alaska" chip is just functioning as the PHY for the GbE resident in the nForce 430 southbridge chip. I don't see anything in the 88E1111 tech docs that indicate that it's a fully functional PCIe GbE controller - for that you'd need something like the Marvell's 88E8050 "Yukon" chip. Here are some links:http://www.marvell.com/products/transceivers/singl...">http://www.marvell.com/products/transceivers/singl...
http://www.marvell.com/products/transceivers/singl...">http://www.marvell.com/products/transce.../Alaska_...'marvell%2088e1111'
http://www.marvell.com/products/pcconn/yukon/index...">http://www.marvell.com/products/pcconn/yukon/index...
Wesley Fink - Thursday, December 1, 2005 - link
The Marvell is a Gigabit PHY, just as in other nForce4 chipsets. As our benchmarks show, it is definitely operating at PCIe speeds.USAF1 - Thursday, December 1, 2005 - link
The fact that a certain PHY (Marvell 88E1111, Vitesse VSC8201RX, etc) and nForce4 MAC combo runs at speeds similar to a dedicated PCIe controller, doesn't make it a PCIe-based solution. Your article would lead one to believe that the Marvell 88E1111 is a PCIe-enabled GbE controller, which in fact it is not.Wesley Fink - Thursday, December 1, 2005 - link
We stated clearly in the Features Chart that the Marvel is a PHY (Physical Layer) chip. This is what we have been seeing in nF4 chipsets for quite a while.USAF1 - Thursday, December 1, 2005 - link
Sorry Wesley, but I still think that your article is misleading. The quote "Asus used the PCIe Gigabit hooks in the 430 Southbridge to provide PCIe Gigabit LAN on the A8N-VM." is just not factual. Neither are the "PCIe" labels next to "Marvell 88E1111" on your graphs. The fact of the matter is that there is no PCIe GbE on this board, yet it's plastered all over the "ethernet performance" section of your article. Why don't you just fix the mistake? Where is Anandtech's vaunted journalistic integrity???Wesley Fink - Friday, December 2, 2005 - link
Upon further research, it does look like the 88E1111 does use a different approach to Gigabit LAN than the PCIe solutions. Marvell refers to the chip as a "Single Port Transceiver". From a users perspective, the performance was found to be the same as PCIe Ethernet, but we have made a few changes in the wording and graphs to more accurately describe the Gigabit LAN used on this board.USAF1 - Friday, December 2, 2005 - link
Thank you very much, Wesley. Now I can stop talking bad about you and your extended family. ;)