Ethernet Performance

We have often talked about the throughput advantage of PCI Express compared to PCI and its impact on Gigabit LAN performance. To show you this, the new motherboard test suite needed to include LAN performance measurements.

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 nForce4 SLI 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 30000
On the client side (the motherboard under test), we used the following Command Line:
Ntttcpr -m 4,0,‹server IP› -a 4 -l 256000 -n 30000
At the conclusion of the test, we captured the throughput and CPU utilization figures from the client screen.

Ethernet Throughput

Ethernet Overhead

While Marvel PCI has a bit lower overhead, the throughput for Gigabit LAN on PCI is almost 30% lower than Gigabit on PCI Express. The nVidia on-chip PCIe LAN also exhibits lower CPU utilization than the full Gigabit LAN on PCIe. It has also been reported that enabling nVidia Active Armor lowers CPU overhead even further.

All Ethernet tests were performed with standard frames, but Gigabit Ethernet supports Jumbo frames as well. Jumbo frames will theoretically provide a further reduction in CPU overhead. We have seen test results that show the combination of Active Armor and Jumbo Frames, which can reduce CPU utilization below 10%, a very respectable performance for on-chip gigabit LAN.

Firewire and USB Performance Audio Performance
POST A COMMENT

108 Comments

View All Comments

  • TigerFlash - Monday, July 4, 2005 - link

    I thought this link would be rather important to see:

    http://forum.msi.com.tw/index.php?topic=82427.0
    Reply
  • TigerFlash - Monday, July 4, 2005 - link

    Reply
  • NightCrawler - Thursday, May 5, 2005 - link

    You make a big deal out of the fact that the DFI can hit 318 but they both do the same 2.8 ghz, users won't see much difference, if any.

    Asus: Maximum OC:
    (Standard Ratio) 234x12 (Auto HT, 2-3-3-7, 1T, 2.8V)
    2808MHz (+17%)
    Maximum FSB:
    (Lower Ratio) 255x11 (2805MHz) (4X HT, 2.5-3-3-7, 2.7V)
    (1:1 Memory, 1T, 2 DIMMs in DC mode)
    (+28% Bus Overclock)

    DFI: Maximum OC:
    (Standard Ratio) 238x12 (Auto HT, 2-3-2-7, 1T, 2.9V)
    2856MHz (+19%)
    Maximum FSB:
    (Lower Ratio) 318x9 (2862MHz) (Auto HT, 2.5-4-3-7, 2.9V)
    (1:1 Memory, 1T, 2 DIMMs in DC mode)
    (+59% Bus Overclock)
    Reply
  • DeanO - Monday, April 18, 2005 - link

    Don't know if anyone's noticed yet, but I just took a trip over to MSI's website, and guess what? Only the SLI mobo has the Creative chip. The Neo4 (i.e. nF4 Ultra chipset) mobo uses the Realtek ALC850. I for one was disappointed...
    That makes for an interesting decision: the SLI board is still cheaper than the Ultra board plus a Creative 24-bit sound card. Hmmm...
    Reply
  • phusg - Friday, March 4, 2005 - link

    New PCI card with C-Media DDL chip: http://www.avsforum.com/avs-vb/showthread.php?s=&a... ge=20&pagenumber=1

    Currently only available via ebay apparantly:
    http://search.ebay.com/HDA-Digital-X-Mystique-7-1-...

    If it has the same performance as Soundstorm remains to be seen. Reading the thread the EAX support is just as dodgy as it was on Soundstorm.
    Reply
  • ElFenix - Thursday, March 3, 2005 - link

    What chipsets did your USB and firewire drives have?

    thanks for the great review!
    Reply
  • bjorn44 - Thursday, March 3, 2005 - link

    Anyone know how they did the memory benchmark with memtest86 3.2? I can't find any option for testing bandwidth.

    Thanks,

    Bjorn
    Reply
  • giz02 - Wednesday, March 2, 2005 - link

    Well if it's any consolation, PCSTATS have updated thier site review of the MSI Neo4 Plat SLI (and will probably make two more updates to it)
    - now states 96Khz
    - will modify DICE statement
    - they are indicating that the sil3132 can do raid5, but I'm not sure that it can...

    Wow Roomraider, that's quite the system you have there.

    Reply
  • Roomraider - Wednesday, March 2, 2005 - link

    #82 u r absolutely correct sir. I have the top SB card available(Audigy 4 Pro)& the only way i get DTS or Dolby Digital of any form is SPDIF out Via Coax or Fiber optic cable with settings for (Passthrough) to my Yamaha 7.1 Amp.



    MOBO Gigabyte Ga-K8NXP-SLI
    CPU AMD Athlon 64 FX-55
    Cooler Gigabyte 3D CoolBlue Ultra Gt
    PSU Thermaltake Purepower 650 Watt
    MEMORY 4xCorsair 512Mb 3200XMS PRO Tracer Ram/Dual channel 2-2-2-5
    Video 2xBFG 6800GT OC PCIE W/Serials in order
    HDD 2xWD-74 GB Raptor HDD/Raid(0)configged
    2xMaxtor 300 GB SATA HDD
    OPTICAL 2xPlextor PX716SA-SATA 16xDual Layer+-DVDRW-48xCDR
    CASE Lian-Li P60 W/clear side panel
    MODS 4 Blu 80mm/1 Blu 92mm(roof/exh)& 4 Blu Cold Cathode Lite Strips
    MONITOR Sony SDM-P234 23" 1920x1200 native
    SOUND Creative Audigy-4 Pro,YamahaDSP-A3090 7.1ch amp/Boston Micro90 spks/Bose AM-5 W/Sub

    ADD-ON MSI TV@nywhere Personal Cinema FX5200 TV/FM tuner
    Reply
  • Tatunkhamon - Wednesday, March 2, 2005 - link

    I admit this is slightly OT, but as I first got excited about the possible DD-encoding feature on the MSI-mobo and then let down by the obvious lack of it, I was happy to find these news:

    http://news.designtechnica.com/article6709.html
    http://www.engadget.com/entry/1234000683034067/

    I know many of us don't like the DRM/HDCP-features of HDMI, but HDMI certainly is the way to transmit high-definition, multichannel audio *without the need* to compress ie. encode to DD. And live content, such as games, would not probably have the copyprotection flags on, anyway. Of course, getting enough coverage for HDMI in both h/w and s/w will take time, but I bet this is the way it's ment to be played in the near future.

    For example, think about combining this with s/w generated mc-audio and Intel HDA. No need for badly implemented codec/DAC in this model. Of this combined with discreete graphics card and the audio generated with the help of vector processing on the card.. I just hope Intel/Nvidia/ATI/whoever would start a strong enough, open standard to compete with EAX. Then Creative would either have to run, fast, or join their forces.

    Meanwhile, because there is not much HDMI-support (except for the earlier, non- multichannel-high-def-audio-supported HDMI-standard, for mainly graphics) some solution providing DD-encoding to be sent over standard S/PDIF would still be very, very desired for many of us.

    I end this thread-hijacking attempt here and apologize if being OT. Now back to our regular programming... :)

    Wbr, Tatu
    Reply

Log in

Don't have an account? Sign up now