The BIOS
The BIOS is pretty much unchanged from the i845E, and is still the AMI BIOS. I used to find it lacking when compared to the Award BIOS others use, but after working with a Pheonix/AwardBIOS, I've learned to appreciate the AMI BIOS. There are a fair amount of tweaking options available, and should be sufficient for most power users.
The basics are in place, such as enabling or disabling features such as the Promise controller, USB, sound, and various other integrated devices. The fun begins with the tweaking. The CPU will likely be locked, so you're forced to do your overclocking via the multiplier. Adjustments in 1MHz increments are available, and simply keying in what FSB you want is a snap. In this menu, you can also adjust the voltage of your CPU, AGP and DDR in case your system isn't getting enough juice.

Memory options are a little dissapointing as you're only able to select between PC1600, PC2100 and PC2700 speeds. There aren't any dividers, or any HCLK+33 option. This will limit the memory overclocking options, and tweakers will find this to be a bit of a drag.
Overclocking
The processor used for this review is the Intel 2.4GHz, 533FSB. The default setting is 18x133, so as you can see, the multiplier is already pretty high. As we've mentioned ealier, the multiplier is locked, so we had to go about it the old fashion way, upping the FSB.
Cooling was done with the Vantec Aeroflow 478 (pretty darn good cooler I may add). vCore was boosted to 1.8v, and the ram timings were set to the most conservative settings, and at 133FSB.
Right off the bat, we went for a 166FSB (we were using PC2700 ram), which proved to be a mistake as the system never got to POST. I dropped down to 160FSB with similar results. In fact, the PC did POST, but the screen went blank at the Promise boot up.
At 159FSB, we managed to get to the Windows splash screen, but were immediately greeted with a blue screen, critical failure. At 157FSB, we got into Windows, and were even able to run the SiSoft CPU benchmark. I tried opening Photoshop to capture the screenshot, but it wouldn't start. In fact, all other applications were unresponsive when the executible was double-clicked. I rebooted, and the system wouldn't POST, forcing a CMOS reset. Although we were having problems, you may notice that the actual OC speeds are a little better than the 845E reviewed earlier.
Like the 845E, our conclusions up until this point are that the motherboard is capable of overclocks of 155+, if your cooling is up to speed. Temperatures during this time were about 55C, so I do believe we've hit the limits of our CPU.
We finally managed a 156FSB overclock, which was a 3MHz improvement over the Max2-BLR. I thought 158 (followed by 157) was a success, but after 15 minutes, the PC just shut itself down. We got one of the first batches of Northwood "B" CPUs, so perhaps a more recent revision would have better luck. We were able to do everything we typically do, such as play a few games, run benchmarks, and working on this article. I've seen higher overclocks from this board, so your milage may vary.
Test Setup
Intel Pentium 4 "B" 2.4GHz @ 533FSB
MSI 845PE Max2
2 x 256MB Crucial PC2700 DDR
2 x 80GB Maxtor ATA100 RAID-0
PNY Verto Ti4600
We will be testing the above against the MSI 845E Max2-BLR. This MSI board happens to be the one we're giving away, and it is a 533FSB, DDR part, already patched with the latest BIOS update, but supports only DDR266. This is the very same board that the 845PE more or less replaces, so we feel this comparison is warranted.
We will be certain to perform tests where the video card should not be a limiting factor. What that means is for the gaming benchmarks, tests will be run at 640x480, stressing the CPU and motherboard subsystem.
Previous Page - The Board Continued
Next Page - Benchmarks
|