MSI P6N SLI Platinum: NVIDIA's 650i Part Two
by Gary Key on March 13, 2007 1:00 AM EST- Posted in
- Motherboards
Disk Controller Performance
The AnandTech iPeak test is designed to measure "pure" hard disk controller performance, and in this case, we keep the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of each hard drive controller with the same hard drive.
We played back our raw files that are recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPeak utility was then used to play back the trace file of all I/O operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. The drive is formatted before each test run and a composite average of three tests on each controller interface is tabulated in order to ensure consistency in the benchmark.
iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each I/O operation. In order to make the data more understandable, we report the scores as an average number of I/O operations per second so that higher scores translate into better performance. This number is meaningless as far as hard disk performance is concerned, as it is just the number of I/O operations completed in a second. However, the scores are useful for comparing "pure" performance of the storage controllers in this case.
The performance patterns hold steady across both Multimedia Content I/O and Business I/O with the SB600 outperforming the Intel ICH7R, Intel ICH8R, and NVIDIA 680i/650i chipsets in our non-Raid tests. The Promise controller on the DFI board is the slowest of all solutions with the Marvell controller on the Intel 975X performing only slightly better. The 680i and 650i MCP chipsets are different but their disk performance in our tests is basically alike. We also ran RAID 0 and RAID 5 tests on each controller with the results being in favor of the 680i MCP by around 2%.
We did not witness any data corruption on our NVIDIA based boards as has been reported on the 680i reference boards, and we consider this problem to be fixed since our P23 BIOS update. We did have data corruption issues with the Gigabyte board when using 1T Command Rate at DDR2-800 with an early F2 beta BIOS but this was solved in the final F3 BIOS. Although our NVIDIA boards generally score lower in these "pure" throughput tests, we find their performance in disk intensive applications is generally better than the other solutions.
Power Consumption
Our power consumption numbers are based on our test setup for this article that consists of the MSI 8800GTX video card, and our boards have C1E/EIST turned off. The 680i leads all boards in power consumption and the 650i SLI is second though on average it still consumes about 7% less power than the 680i at idle and 3% at load. The P965 consumes the least amount of power at idle but around 10% more than the RD600 at load. The 975X still offers a decent level of power conservation compared to the other chipsets even though it is over two years old now.
Overclocking
As we mentioned earlier in the article, the overclocking capability of the MSI board has varied widely with each BIOS release we have tested. This will be disappointing to some users and we expected a little better overclocking capability based upon the quality of the board and its BIOS options. However, like most recent MSI boards we find they have generally implemented very aggressive memory and chipset timings at stock speeds at the expense of overclocking. This is generally true of the MSI P6N SLI Platinum but we have found it overclocks better than most of the MSI boards we have tested recently.
One area we found interesting is the MSI board experienced several FSB holes with the 1.22 BIOS that we witnessed on the 680i reference boards. We did experience the typical FSB hole around 416FSB and hit another one right around 460FSB and 500FSB in the latest BIOS. We also found that during overclocking we had to set our CPU voltages a little higher than on the other boards to attain the same overclocks. We have experienced this before with other MSI boards and it was due to tighter sub-timings in order to improve the performance of the board at stock speeds.
The AnandTech iPeak test is designed to measure "pure" hard disk controller performance, and in this case, we keep the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of each hard drive controller with the same hard drive.
We played back our raw files that are recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPeak utility was then used to play back the trace file of all I/O operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. The drive is formatted before each test run and a composite average of three tests on each controller interface is tabulated in order to ensure consistency in the benchmark.
iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each I/O operation. In order to make the data more understandable, we report the scores as an average number of I/O operations per second so that higher scores translate into better performance. This number is meaningless as far as hard disk performance is concerned, as it is just the number of I/O operations completed in a second. However, the scores are useful for comparing "pure" performance of the storage controllers in this case.
The performance patterns hold steady across both Multimedia Content I/O and Business I/O with the SB600 outperforming the Intel ICH7R, Intel ICH8R, and NVIDIA 680i/650i chipsets in our non-Raid tests. The Promise controller on the DFI board is the slowest of all solutions with the Marvell controller on the Intel 975X performing only slightly better. The 680i and 650i MCP chipsets are different but their disk performance in our tests is basically alike. We also ran RAID 0 and RAID 5 tests on each controller with the results being in favor of the 680i MCP by around 2%.
We did not witness any data corruption on our NVIDIA based boards as has been reported on the 680i reference boards, and we consider this problem to be fixed since our P23 BIOS update. We did have data corruption issues with the Gigabyte board when using 1T Command Rate at DDR2-800 with an early F2 beta BIOS but this was solved in the final F3 BIOS. Although our NVIDIA boards generally score lower in these "pure" throughput tests, we find their performance in disk intensive applications is generally better than the other solutions.
Power Consumption
Our power consumption numbers are based on our test setup for this article that consists of the MSI 8800GTX video card, and our boards have C1E/EIST turned off. The 680i leads all boards in power consumption and the 650i SLI is second though on average it still consumes about 7% less power than the 680i at idle and 3% at load. The P965 consumes the least amount of power at idle but around 10% more than the RD600 at load. The 975X still offers a decent level of power conservation compared to the other chipsets even though it is over two years old now.
Overclocking
As we mentioned earlier in the article, the overclocking capability of the MSI board has varied widely with each BIOS release we have tested. This will be disappointing to some users and we expected a little better overclocking capability based upon the quality of the board and its BIOS options. However, like most recent MSI boards we find they have generally implemented very aggressive memory and chipset timings at stock speeds at the expense of overclocking. This is generally true of the MSI P6N SLI Platinum but we have found it overclocks better than most of the MSI boards we have tested recently.
One area we found interesting is the MSI board experienced several FSB holes with the 1.22 BIOS that we witnessed on the 680i reference boards. We did experience the typical FSB hole around 416FSB and hit another one right around 460FSB and 500FSB in the latest BIOS. We also found that during overclocking we had to set our CPU voltages a little higher than on the other boards to attain the same overclocks. We have experienced this before with other MSI boards and it was due to tighter sub-timings in order to improve the performance of the board at stock speeds.
Facebook
Twitter
RSS
20 Comments
View All Comments
ranutso - Wednesday, March 21, 2007 - link
Great article. Thank you Gary.cosmotic - Wednesday, March 14, 2007 - link
How can you say that MSI software is decent? It's totally hideous. I think Anandtech owes it to the community to encourage motherboard manufactures to start writing native-feeling Windows applications instead of these crap piles all the manufactueres are shipping now. This includes AMD/ATI, nVidia, Realtek, and many others for their drivers as well.Gorgonzola - Thursday, June 21, 2007 - link
I could not agree more!anandtech02148 - Wednesday, March 14, 2007 - link
Here I go againcomplaining about the psu and power consumption, but 300watts load,200watts idle,
not to mention fancy subwoofer, a few electronics here there,
good gaming is in the summer time, and i'll be cranking up the AC too which is another 250wtts.
i wish newegg.com would sell me a n.korean light water nuclear reactor so i can run all my greatest hardwares.
Spanki - Wednesday, March 14, 2007 - link
Hey Gary,Since it looks like this mb outperforms pretty much every other board in the review in most tests (at stock speeds, where head-to-head comparisons usually take place) - including the much touted 'Extreme' board(s), do you plan to include it for comparisons in future reviews?
Olaf van der Spek - Tuesday, March 13, 2007 - link
The 650i uses 21% more power on idle compared to the ICFX3200. What is nV doing with all that power? This seems absurd.Gary Key - Tuesday, March 13, 2007 - link
It is being channeled into the on-board Flux Capacitor. ;-) I can tell you that we have hounded NVIDIA to no end about this issue with their chipsets. It should be addressed when they finally go to a single chip solution later this year (we are still hoping this occurs).
Frumious1 - Tuesday, March 13, 2007 - link
If I were to venture a guess, NVIDIA probably isn't doing any proper power savings work for the chipset. Just like with CPUs and GPUs, there's a lot of stuff on the chipset that is often not in use and can be put into a sleep/deep sleep mode. The 650i and 680i use 100-107W more power at full load than at idle. The 975X uses 141W more at load, P965 139W more, and RD600 105W more.IIRC, AMD is using a newer process technology for RD600, so that would help explain their lower overall power. Intel seems to benefit from power savings in idle mode, but at full load they are pretty close to NVIDIA. The extra "stuff" in 680i relative to 650i could easily account for the added ~10W that it requires. Seems to me like all companies involved could do more with chipset power savings. AMD is just ahead on the process tech (again, I think); Intel uses an older process but decent power saving circuitry; NVIDIA doesn't do anything to conserve chipset power.
When you consider that at idle the PC is doing nothing important, AMD and Intel should drop CPU clocks further (600 MHz ought to be enough), and they could drop FSB/bus speeds and chipset voltages as well. Why run 1066FSB when you're doing essentially nothing? Why run 1000MHz HyperTransport to transfer... nothing? I believe AMD does drop HT speeds at idle on their mobles chips, so why not on desktop offerings?
Just my two cents.
WT - Tuesday, March 13, 2007 - link
Patiently awaiting the Gigabyte version of this board, as I was most interested in upgrading to the 965DS3 board rev 3.3, but the 650 look like it is worth the wait. Also, since the C2D price drop isn't until late April, I have time to wait and make a decision once that board is available. Good read as usual guys !ghitz - Tuesday, March 20, 2007 - link
Exactly what I was thinking !!