Labels:
ASUS Motherboard
Network Throughput & System Power Consumption
The Gigabit PCIe Ethernet controllers vary on each board, so it will be interesting to see what the performance is on all the boards. Each of the boards uses a different Gigabit Controller. Let's take a look at what controller is on each of the boards.
· ASUS Sabertooth X58 - Realtek RTL8110SC
· MSI Big Bang XPower - Realtek 8111DL
· ASRock 890GX Extreme4 - Realtek RTL8111C
· ASRock 890GX Extreme4 - Realtek RTL8111E
· ASRock 890FX Deluxe4 - Realtek RTL8111E
· GIGABYTE 890FXA-UD5 - Realtek 8111D chip
· GIGABYTE 880GMA-UD2H - Realtek 8111D chip
· MSI 890FXA-GD70 - Realtek 8111DL
· ASUS Crosshair IV Formula - Marvell 8059 Gigabit LAN controller
· ASRock 890GX Extreme3 - Realtek RTL8111E
· Biostar TA890GXB HD - Realtek RTL8111E
To test the Gigabit PCIe Ethernet controllers we used the NTttcp tool to test Ethernet Throughput and the CPU utilization of the various Ethernet controllers used on the AMD Motherboards. To do this properly we set up one machine as the server; in this test an Intel Core I7 930 system with an EVGA X58 Tri-SLI motherboard acted as our Gigabit LAN server.
On the server side, we used the following Command Line for Gigabit PCIe Ethernet Testing:
Ntttcpr -m 4,0,<Client IP> -a 4 -l 256000 -n 30000
On the client side (the motherboard being tested), we used the following Command Line:
Ntttcps -m 4,0,<Server IP> -a 4 -l 256000 -n 30000
At the conclusion of each test we recorded the throughput and CPU utilization figures from the client screen as that is the system being tested.
Benchmark Results: This is the first time in ten motherboards that there was a dramatic difference in network performance. Nine of the ten boards that we have tested have had a network throughput of 942-945+Mbits/s. The ASUS Sabertooth X58 was only 752.336 Mbits/s. After a bit of trouble shooting to make sure my server wasn't faltering I got in touch with ASUS and found out my results were spot on. This is what they had to say:
"It is due to the use of a PCI based network controller as opposed to a PCI-E based network controller. Your number is approximate to what we have from internal testing. Due to the considerably higher quality components as well validation and other specific elements on the SABERTOOTH X58 this compromise was put into effect by the product manager. As most users realistically do not put heavy throughput loads on their network it was felt while not perfect it was practical implementation. In real world reality the performance offered of approximate 85 to 90MB is sufficient for even high performance data transfers including backups and HD streams."
Since power consumption is a big deal these days, we ran some simple power consumption tests on our test beds. The systems ran with the power supplies, case fans, video cards and hard drives. To measure idle usage, we ran the system at idle for one hour on the desktop with no screen saver and took the measurement. For load measurements, Prime95's in-place large FFT's was run on all cores to make sure each and every processor was at 100% load for maximum power consumption and heat. We also ran Furmark 1.8 at 1280x1024 to put a full load on the graphics card. All systems were tested using identical hardware.
Power consumption of the ASUS Sabertooth X58 motherboard was a little higher than I would have liked to have seen. At idle it was pulling 141 watts at the wall which is nearly 22% higher than the MSI Big Bang XPower. Once we start putting a load on the CPU alone the difference drops to 14.5%, firing up the CPU as well as the GPU there is only a 9.8% difference.