ASUS P5N32-E SLI Plus: NVIDIA's 650i goes Dual x16
by Gary Key on April 2, 2007 3:30 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% over the 680i LT and 650i MCP in these particular benchmarks. Although our NVIDIA boards generally score lower in these "pure" throughput tests, we find their actual performance in disk intensive applications are generally equal to or 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 boards lead all in power consumption but the ASUS P5N32-E SLI Plus is right behind indicating that the 570 SLI MCP draws slightly more power than the 430 MCP on the true 650i boards. 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 getting long in the tooth.
FSB Overclocking
While the ASUS P5N32-E SLI Plus offers very good overclocking capabilities, it quickly runs out of headroom near the 500FSB level. We believe this is a limitation of the 650i SPP based upon our results with other 650i boards. ASUS has extracted just about every ounce of performance out of this particular chipset. While it may not satisfy the extreme overclocker, the overclocking levels should be fine for a majority of users.
We found during overclocking that we had to set the CPU voltages a little higher than on the other boards to attain the same or better overclocks. We experienced this with the MSI 650i and EVGA 680i LT SLI boards, and it was generally due to tighter memory/chipset timings in order to improve the performance of the board at stock speeds. Overall, our overclocking experiences were very good provided we had the memory sub-timings dialed in for each memory type we utilized.
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% over the 680i LT and 650i MCP in these particular benchmarks. Although our NVIDIA boards generally score lower in these "pure" throughput tests, we find their actual performance in disk intensive applications are generally equal to or 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 boards lead all in power consumption but the ASUS P5N32-E SLI Plus is right behind indicating that the 570 SLI MCP draws slightly more power than the 430 MCP on the true 650i boards. 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 getting long in the tooth.
FSB Overclocking
While the ASUS P5N32-E SLI Plus offers very good overclocking capabilities, it quickly runs out of headroom near the 500FSB level. We believe this is a limitation of the 650i SPP based upon our results with other 650i boards. ASUS has extracted just about every ounce of performance out of this particular chipset. While it may not satisfy the extreme overclocker, the overclocking levels should be fine for a majority of users.
We found during overclocking that we had to set the CPU voltages a little higher than on the other boards to attain the same or better overclocks. We experienced this with the MSI 650i and EVGA 680i LT SLI boards, and it was generally due to tighter memory/chipset timings in order to improve the performance of the board at stock speeds. Overall, our overclocking experiences were very good provided we had the memory sub-timings dialed in for each memory type we utilized.
Facebook
Twitter
RSS
37 Comments
View All Comments
Spanki - Monday, April 2, 2007 - link
Nice review, but I must say that at worse it's misleading and at best it's potentially confusing regarding the memory timings used (in particular, the command-rate). I really liked seeing both 1T and 2T benchmark numbers, but the images would have been a lot better if the command-rate was listed for _every_ board tested (every line-item).In the absense of that, I have to assume that the rate was 2T for all other boards, unless specifically stated, but even that pattern wasn't followed on the first graphic. I guess this means that I'm casting a vote to see the 1T numbers/comparisons on all the boards (whether or not it makes a 'significant' difference) - if you threw out all the 1T results in this review, I think you get an overall different 'picture' of how well this board compares with the others.
JarredWalton - Monday, April 2, 2007 - link
I believe Gary is in the process of retesting boards with 1T where it works, and that will be part of an upcoming roundup. The problem is that until recently, the 680i boards were unable to support 1T at DDR2-800, and no other chipset has managed it either. Now quite a few 680i (and 650i maybe?) boards have 1T support, with the appropriate DIMMs. Personally, I'm okay with using more typical (and cheaper) 2T RAM and overclocking to make up the less than 5% difference. Some people want the absolute best, though.yyrkoon - Monday, April 2, 2007 - link
Hmm, actually, now that I think about it, the CPU Bus speed *is* 250Mhz, and I did drop the memory down to the 667Mhz divider, but technically, the memory is still running above 800Mhz DDR2 (I think)/me checks while blushing in the process
JarredWalton - Monday, April 2, 2007 - link
But you're running an AM2 configuration, right? I'm specifically talking about Core 2 chipsets, where 1T support for DDR2-800 is a more recent addition. I should have made that clearer. I'm pretty sure P965, 975X, 680i, 650i, 590i Intel, and 570i Intel all initially didn't have 1T support at that speed.yyrkoon - Monday, April 2, 2007 - link
Uh, my ABIT NF-M2 nView has run nothing BUT 1T, even with the memory OC'd (it is 5-5-5-12 DDR2 6400 Promos, but running 4-4-4-12 1T right_now). Perhaps you meant 'current chipsets' ?Gary Key - Monday, April 2, 2007 - link
Jarred meant current Intel chipsets but 1T is still not working right at DDR2-800 on the P965, close but not there yet. The NVIDIA Intel chipsets have progressed rapidly with 1T operation up to DDR2-900 with decent PC2-6400 memory at fairly low latencies now. We have hit DDR2-1000 at 1T on the RD600 with relaxed timings, have the new board revision on the EVGA 680i and we are so close to DDR2-1000 1T with some "affordable memory" at this time.
BladeVenom - Monday, April 2, 2007 - link
The last 20 motherboard articles on Anandtech have all been for Intel. Doesn't that seem a little excessive? I know Intel is the preferred processor for midrange and up, but with the prices of the x2 3600 and the less expensive motherboards, lots of people are still buying AMD.fliguy84 - Monday, April 2, 2007 - link
Shouldn't it be 2MB on the 4th page?JarredWalton - Monday, April 2, 2007 - link
Yes - fixed.yacoub - Monday, April 2, 2007 - link
Dude, 2T? Eww, why? Is that mandated for quad-core systems or were you simply unable to get any stability with 1T at that OC level? :(