ASUS M2N32-SLI & Epox MF570SLI: AM2 Wunderkinder
by Wesley Fink on June 28, 2006 5:10 AM EST- Posted in
- Motherboards
Overclocking
Any board that reaches a 300 clock speed or beyond must be considered an excellent overclocker. The Epox reached 318 clock speed with a 9x multiplier and 250 at the stock 12x multiplier. The 570 is a simpler design than the 590 and it should be as good at overclocking, or better than, the 590. There is no logical reason both 570 boards we have tested top out around 315.
Both platforms use the most up-to-date 590 MCP, with a few items disabled or unimplemented in the 570 version. We are left to wonder if perhaps the maximum overclock of the 570 is being artificially limited so as to make the 590 more attractive? We have no hard evidence of this thought; we are just trying to understand why a simpler solution - which normally overclocks better - does worse with the MSI and EPoX 570 SLI boards.
Memory Stress Testing
Optimum tRAS
The Epox MF570SLI behaved like the Foxconn and ATI in tRAS memory testing. As a result all testing was performed at 3-3-3-13 timings. This provided the greatest bandwidth and the best stability.
Memory Stress Testing
The Epox board does not offer the fine granularity of memory voltage controls that the ASUS board provides, as you can only increase voltage in 0.1V increments. We would prefer at the very least 0.05 V increments, and even better are the 0.025V increments that ASUS, Foxconn and others offer.
With two DIMMs installed, testing was completely stable at 3-3-3-8 2T timings at DDR2-800. Even with further tuning, we were not able to run at 1T command rate at DDR2-800. The highest speed that could run at 1T was DDR2-667. As already stated in the ASUS memory testing, the AM2 does not officially support 2T Command Rate at DDR2-800, so it should come as no surprise that the Epox could not handle the 1T settings. The Gigabyte and ASUS are the only AM2 boards tested thus far that can run at DDR2800 2T, but extensive testing showed no advantage at all in real-world performance using the DDR2-800 1T timings.
Installing four DIMMs stresses the memory subsystem further. We had to drop the timings slightly to a 4 RAS-to-CAS setting at 2.3V to achieve stable performance with 4 DIMMs. In real-world testing this is practically the same as the fastest 3-3-3 timings we achieved on the fast ASUS M2N32-SLI Deluxe.
| Epox MF570SLI Overclocking Testbed |
|
| Processor: | AM2 4800+ (x2, 2.4GHz, 1MB Cache per core) |
| CPU Voltage: | 1.5V (default 1.4V) |
| Cooling: | AMD Stock Heatpipe FX62 Cooler |
| Power Supply: | OCZ Power Stream 520W |
| Memory: | Corsair Twin2x2048-PC2-8500C5 (2x1GB) (Micron Memory Chips) |
| Hard Drive | Hitachi 250GB 7200RPM SATA2 16MB Cache |
| Maximum OC: (Standard Ratio) |
250x12 (5x HT, 3-3-3-13) 3000MHz (+25%) |
| Maximum FSB: (Lower Ratio) |
318 x 9 (4x HT, 3-3-3-13) (2862MHz, 2 DIMMs in DC mode) (+59% Bus Overclock) |
Any board that reaches a 300 clock speed or beyond must be considered an excellent overclocker. The Epox reached 318 clock speed with a 9x multiplier and 250 at the stock 12x multiplier. The 570 is a simpler design than the 590 and it should be as good at overclocking, or better than, the 590. There is no logical reason both 570 boards we have tested top out around 315.
Both platforms use the most up-to-date 590 MCP, with a few items disabled or unimplemented in the 570 version. We are left to wonder if perhaps the maximum overclock of the 570 is being artificially limited so as to make the 590 more attractive? We have no hard evidence of this thought; we are just trying to understand why a simpler solution - which normally overclocks better - does worse with the MSI and EPoX 570 SLI boards.
Memory Stress Testing
Optimum tRAS
The Epox MF570SLI behaved like the Foxconn and ATI in tRAS memory testing. As a result all testing was performed at 3-3-3-13 timings. This provided the greatest bandwidth and the best stability.
Memory Stress Testing
The Epox board does not offer the fine granularity of memory voltage controls that the ASUS board provides, as you can only increase voltage in 0.1V increments. We would prefer at the very least 0.05 V increments, and even better are the 0.025V increments that ASUS, Foxconn and others offer.
| Epox MF570SLI DDR2-800 Timings - 2 DIMMs (2/4 slots populated - 1 Dual-Channel Bank) |
|
| Clock Speed: | 200MHz |
| Timing Mode: | 800MHz - Default |
| CAS Latency: | 3 |
| RAS to CAS Delay: | 3 |
| RAS Precharge: | 3 |
| RAS Cycle Time: | 8 |
| Command Rate: | 2T |
| Voltage: | 2.2V |
With two DIMMs installed, testing was completely stable at 3-3-3-8 2T timings at DDR2-800. Even with further tuning, we were not able to run at 1T command rate at DDR2-800. The highest speed that could run at 1T was DDR2-667. As already stated in the ASUS memory testing, the AM2 does not officially support 2T Command Rate at DDR2-800, so it should come as no surprise that the Epox could not handle the 1T settings. The Gigabyte and ASUS are the only AM2 boards tested thus far that can run at DDR2800 2T, but extensive testing showed no advantage at all in real-world performance using the DDR2-800 1T timings.
| Epox MF570SLI DDR2-800 Timings - 4 DIMMs (4/4 slots populated - 2 Dual-Channel Banks) |
|
| Clock Speed: | 200MHz |
| Timing Mode: | 800MHz - Default |
| CAS Latency: | 3 |
| RAS to CAS Delay: | 4 |
| RAS Precharge: | 3 |
| RAS Cycle Time: | 13 |
| Command Rate: | 2T |
| Voltage: | 2.3V |
Installing four DIMMs stresses the memory subsystem further. We had to drop the timings slightly to a 4 RAS-to-CAS setting at 2.3V to achieve stable performance with 4 DIMMs. In real-world testing this is practically the same as the fastest 3-3-3 timings we achieved on the fast ASUS M2N32-SLI Deluxe.
Facebook
Twitter
RSS
23 Comments
View All Comments
leexgx - Sunday, April 15, 2007 - link
i am trying to get an stable overclock from this m2n32-sli deluxe got an 3800+ X2 at 2.65 (10x265) when i set it to 2.70 270 it just BSODdo i need to up MB and SB volts up ? as well or lower the NB to SB as well
if any one could point me to an web site that has overclocked one of these motherboards be usefull
mss242 - Tuesday, December 5, 2006 - link
doesn't the asus also offer raid 5? Page 2 lists raid options as 0,1,0+1,10, and JBOD.darkswordsman17 - Tuesday, July 18, 2006 - link
Something I've noticed is a lot of boards are starting to get DDL/Dolby Master Studio and DTS Connect capabilities, but Anandtech doesn't even mention it half the time, and doesn't even test it at all.I think it would be worth looking into, as its really starting to become a viable alternative to Creative's surround solutions (EAX support wouldn't even matter if its being encoded in DD/DTS would it), which more than a few people do not like (although myself am fairly neutral as I've liked the Creative cards I've owned). Also, there are plenty of people sore over losing DDL support when they moved to a newer platform than nForce2 with SoundStorm. I often see people saying how they still miss it, and yet, its been here for almost 6 months already (Intel Bad Axe, possibly others). I see a lot of new boards from ASUS, Gigabyte, and Abit featuring support for these.
Also a lot of these new boards are using different chips to handle processing, so maybe that makes a difference as far as quality or performance.
I just think it would be beneficial at the very least to make a note of it, as its not always easy to find out what boards actually do support them.
Just a thought.
classy - Wednesday, June 28, 2006 - link
Why not show the results of the scores gained from oc?DrMrLordX - Friday, June 30, 2006 - link
Probably because it's a review of the motherboard and not the CPU itself. All a motherboard review needs is information about the maximum HTT speeds achieveably on the board at stock.saratoga - Wednesday, June 28, 2006 - link
The CPU use numbers for the on board audio are great, but it'd also be nice to know just how good the actual quality is. If theres massive THD or a resampling problem, benchmarks will look great, but the part may still suck.Using RMAA would allow people to see immediately if there were any serious issues with the sound quality such as poor resampling or noise.
Seeing as other tech sites have started using it, it'd be nice if you guys could too. See this review:
http://techreport.com/reviews/2005q4/soundblaster-...">http://techreport.com/reviews/2005q4/soundblaster-...
Obvious so much info isn't needed for a motherboard review, but posting the summary chart that RMAA spits out with the crosstalk, SNR, IMD and THD numbers would be great.
Gary Key - Thursday, June 29, 2006 - link
We will post RMAA results when a new audio chip is introduced. I will run the results on this chipset and have it available in our next article. We actually used RMAA 5.5 in a previous article and had more comments wanting subjective analysis. However, we will do the short version of the test results. :)Sifl - Wednesday, June 28, 2006 - link
With a new and interesting built-in WiFi option (as far as ATX MOBO's go), why not show us the antenna and where it goes?
For the Asus board layout, I can see all the IO ports (letter designations on the image could help identify which connector is which) but I'm not sure where the WiFi antenna would go. Is it the little gold colored thing off to the right in the picture for rear IO? Because I don't see that same thing in the top views. Maybe another view is better like a perspective view of the ports, rather than the straigh-on view.
And why does the Epox lack Firewire -- But has 10 USB's ?! Who uses 10 USB ports? Firewire is just basic for any digital video equipment. I chose my current MOBO (Epox 9NPA+) because it HAS Firewire. Maybe they will have another model with it included.
Myrandex - Wednesday, June 28, 2006 - link
"Who uses 10 USB ports?" I do...But then again I also use firewire, bure still it would have been stupid for Epox to have included less USB ports (arrggg EVGA's microATX nForce4 SLI mobo of mine only supporting 8). Back 6 have printer, Windows Media Center Remote sensor, mouse, cell phone data/charger for MPX220 (mine), and cell phone data/charger for Motorola V3 Razr (g/f's), and webcam. Then only 2 of my front 4 USb ports can be connected, and that means I can only use 2 USB devices in the front (ranging from hard drive cages for testing, flash sticks, USB controllers for emulators, etc.) For a manufacturer to have a chipset with 10 USBs yet only implement 8 is just kind of backwards and extremely pointless :-/
On a side note I did have to laugh at PCI/AGP Fixed at 33/66 as there seem sot be no AGP on these boards for it to be locked at 66 :)
Jason
Wesley Fink - Wednesday, June 28, 2006 - link
Yes, the antenna screws to the gold connector on the right of the rear IO port on the Asus.As for no firewire on the Epox, it is likely a cost savings to meet a target price point. Most of the digital cameras we have seen recently have emphasized high-speed USB2 instead of Firewire, although we agree Firewire is still widely used in digital cameras and video. Add-on Firewire cards are very reasonable, but they would be an added expense if you required Firewire on the Epox.
The MSI 570 reviewed at http://www.anandtech.com/mb/showdoc.aspx?i=2773&am...">http://www.anandtech.com/mb/showdoc.aspx?i=2773&am... is also based on the nForce 570 chipset and does feature Firewire.