Ethernet Performance


The current motherboard test suite includes LAN performance measurements. All of these boards utilize PCI Express controllers with the only difference being the supplier of the core logic.

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 Intel 975x motherboards.

We set up one machine as the server; in this test, an Intel system 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:
Ntttcpr -m 4,0,‹server IP› -a 4 -l 256000 -n 30000
On the client side (the motherboard under test), we used the following Command Line:
Ntttcps -m 4,0,‹client 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 Throughput

The Marvell 88E8062 PCI Express Dual LAN solution exhibits excellent throughput along with very good CPU utilization rates. The Broadcom 5789KFB option on the Gigabyte board offers excellent throughput, but at a slightly higher CPU utilization than the Broadcom 5781 chipset.

All standard Ethernet tests were performed with standard frames and the NVIDIA Active Armor suite disabled unless otherwise noted. Gigabit Ethernet supports Jumbo frames as well and provides a further reduction in CPU overhead. We added a further test scenario in which ActiveArmor and Jumbo frames were enabled on the Asus P5N32-SLI Deluxe board via the new 6.82 WHQL platform driver set. This is shown for illustrative purposes and shows the favorable impact of this technology.

We witnessed one anomaly that has been reported to Asus. The Ethernet ports are only available if the Intel PCI-X chipset is enabled in the BIOS.

Firewire and USB Performance Audio Performance
Comments Locked

31 Comments

View All Comments

  • Cygni - Wednesday, December 7, 2005 - link

    Workstation /= server...

    Overclocking isnt wacky in the worstation world at all. Picture a guy using 3dsmax everyday in his personal studio, or rendering TV streams in Maya, or doing texture work, or compiling source codes at a game studio...

    Overclocking in the Workstation market is probably ALOT more common than people would think. Take overclocking that Pressler. That big of an overclock could mean whole DAYS of extra time per year that the CPU would have otherwise been spending cranking away on rendering. Thats a serious gain.

    I dont think many serious servers will find a board like this one in em, to boot. Its lacking alot of features that a serious server hosting lots of users would deffinitly feel the need for... like built in SCSI, multi cpu support, RAM support over 8gig/4slot, etc.

    On the other hand, this is a near perfect workstation board. High clocked RAM support for both ECC and non ECC. Non buffered memory. Both SLI and Crossfire support at 8x/8x. Plenty of SATA II plugs for cheap storage.

    I took notice of the OC results and settings, and i like what i saw. :)
  • Cygni - Wednesday, December 7, 2005 - link

    That was a reply to the poster one level up, sorry. :D

    Thanks for taking the time to do the OC tests, however. Some people out there did like to see em.
  • Zebo - Tuesday, December 6, 2005 - link

    If those 4.8ghz air overclocks are normal intel won't have to wait for conroe to take back enthusiasts crown. That's amazing.
  • Leper Messiah - Tuesday, December 6, 2005 - link

    Hell yeah. 4.8GHz on air with Dual core is a 20% over what an 840EE will do. And I bet that with nF4 and a DFI 5.0+ is possible. oooh. Sexay.
  • stephenbrooks - Tuesday, December 6, 2005 - link

    Yeah, I'm wondering if the Pentium D 920 might become a new "favourite" chip for them to play with :)
  • Niv KA - Tuesday, December 6, 2005 - link

    Talking about Conroe, isn't the 975X supposed to support upgradability to it
  • xtremejack - Tuesday, December 6, 2005 - link

    If you look at the chipset price guides for this month by Anand, 975x will not support Conroe
  • Gary Key - Tuesday, December 6, 2005 - link

    Intel has not officially confirmed either way on the 975X to Conroe path. We have heard both a positive and negative on this but from different sources. As soon as we have a clear path it will be posted.
  • Niv KA - Tuesday, December 6, 2005 - link

    That would mean I could buy a 975X based MB with a Celeron and buy a Conroe in the summer without having to buy a new motherbord and all
  • Calin - Tuesday, December 6, 2005 - link

    When the graphs shows both min framerate and max framerate, I suggest that ordering should be done by the min framerate instead of the max framerate. Only on F.E.A.R. Performance test on the 9th page the results will change, and only for the top two boards - however, I prefer to have good consistent high framerate in the worst cases than extremely high framerate in the best conditions.
    Thanks

Log in

Don't have an account? Sign up now