Radified User's Guide for the Asus CUSL2

Intro Mushkin CuWalker Intel ATA drivers Page 3
Page 4 Page 5 Benchmarks asusboards.com Home

Page 5

* Current BIOS is 1006_Final (released 02.24.01). I am using this BIOS with no problems. Each bios seems to be a little better. Can download latest release version of the BIOS here (have to scroll down the page). Somebody told me 1007 was the latest final, but at these sites, it says 1006.

* Asus has a new bios flash that lets you flash from Windows. I have not used this utility myself, but those who have, have reported no problems. I am from the old school on flashing your bios - which says to do it from (true) DOS. Doctor Mike from Greece is asking about the Live Update thing - says he gets the error:
"File BS_LOAD.DLL missing" and the program exited. I have not used. If anyone has an answer/solution ...

* You cannot use a Slot 1 CPU with this mobo (should be obvious). 

* Many people have been having problems with their IDE burners, especially with the new Intel ATA drivers. I have a SCSI burner (Plextor) and have had no problems. This is one of the biggies. It's my understanding that people who don't install the Intel ATA drivers are not having probs with their burners. 

But I don't have IDE burner, so if you do have - you should check more carefully - post a Q at forum and ask about best config of your particular burner (brand/model) & primary/secondary, master/slave config. My understanding is that, if you don't install the Intel ATA100 drivers, you won't have probs with your IDE burner - not absolutely positive, but pretty sure.

* Some people have complained of shutdown problems. I've had no probs w/ shutting down. An interesting note from Paul (Connecticut), who reports: I am running a Plextor 12/10/32 EIDE (W98SE). I have observed (reproducible) that if I leave a CD in the Plextor drive, it will hang on shutdown.  Remove CD, and it shuts down fine.

* updated 24oct2000 - There is also the infamous cold boot problem (CBP) that resets your CPU voltage back to default (1.65) during cold boots - or at least that's how I understand it. I don't have this problem. I set my CPU core voltage to 1.80v, and that's where it stays. Cicada says we should refer to this as the default core voltage on cold boot problem, as there is much confusion + misinformation on the subject. 

Not sure if this in some way stems from having all 3 RAM slots filled with PC133 ram. Since I don't have this prob myself, I haven't researched it much. I did read one post, however, that said someone fixed the prob by cranking up his voltage. If you know more, lemme know. I will copy-n-paste what cicada has said about it:

On a cold boot, the CUSL2 does NOT apply the manual voltage setting for the processor until after a successful POST (after the BEEP). This is only a problem when over-clocking your CPU to it's upper limits, as the default voltage may not be enough to make a successful POST. There is currently NO FIX for this (apart from modifying the processor pins) and EVERY CUSL2 does this on cool boot. It's as simple as that.

Props & thanks. So it seems like you have to crank your CPU back (fate worse than death for overclockers), or get new one.

* Guillermo writes: I'd like to add a comment on the CBP .. and I must point out that I do not know the exact thread, but I've read this on the cusl2.com site, but it seems to be a BIOS problem. This usually happens when you change the FSB and voltage in the same session, then reboot. 

The solution is to first change voltage, and reboot, let the system boot normally to windows. Then change the FSB (in a separate session). This workaround also seems to work for people having probs changing memory settings. Some say bios version 1003 beta 4 solves the cold boot problem, while some say it won't. I hope this helps overclockers.

Update 03mar2001: Eric writes: I have the infamous core voltage problem on cold boot, related to trying to overclock a Cel_566. As you may be aware, some sites encourage the use of a thin copper wire to fool the motherboard into cold booting at a higher voltage. But this risks blowing up the CPU.

I've found a way to bypass this problem, using the Suspend-to-RAM feature. It is not optimum because it involves spending a bit more in electricity costs, but I find it easier and less frustrating than having to shut down the power supply at the back of my computer every time.

The CUSL2 includes a feature called Suspend-to-Ram (enabled in BIOS). You need a compatible OS (W2K or WinME) & RAM. In fact, with the new BIOS version (1006), you can go straight to the BIOS, and enter the settings (both Vcore & bus speed) at the same time. Then, from the OS (W2K or WinME), when you are finished with your system, simply select Stand-by

Then the computer will turn off almost everything, except power to the ram (hence the small extra electricity cost mentioned earlier). When you wake up the computer, it won't POST in the BIOS, and the (over-clocked) CPU settings will remain intact. Off course, I'll reboot once in a while .. but at least I don't have to crawl to the back of the power supply, or make change to the BIOS. Surely some people are doing the same, but I've never had a board with Suspend-to-Ram before, and I find it a much more secure way to overclock my system. [Thx for the note, Eric]

* The CUSL2 will let you run your CPU at FSB 133, with your (PC100) RAM at 100MHz. Pretty cool if you have PC100 RAM and don't want to buy new (PC133). Most heavy hitters are using Mushkin's rev 2 stuff, for getting the highest RAM perf. Highest RAM perf comes from setting RAM timing to 2-2-2, 5/7. I run 2-2-2, 7/9. Heard the 1st number (X-X-X) makes biggest diff. In other words 2-2-2 will run much faster than 3-2-2 (all number don't have to be the same) than going to 3-2-2 from 3-3-3 (which doesn't do jack for performance). Biggest diff comes from going from 100MHz to 133MHz. 

* Heard some say that same PC100 ram that will run at 133 on BX, won't run at 133 on CUSL2. This would not surprise me.  

* Received a few notes asking about NT with legacy devices. Since I have neither, I can't comment. But if you know of any probs/solutions, I'd be glad to post. 

* Word on the street is that Windows Media Player 7 is the worst part of WinME. You can get around using Media Player 7 cuz most don't know that WinME also comes with MP6.4, and can be used over the sucky MP7 by:

Open mplayer2.exe from C:\Program Files\Windows Media Player. Go to view > Options > Formats & select all file formats that you want to associate wmp6.4. From then on, all files that you select will open with wmp6.4. (Thx to SkyPilot for that one)

* Daniele wrote to say that AGP4X kicks back to 2X at FSBs >140. True for anyone else? 

* AFAIK Redhat 7 is the only Linux distro with native sppt for the 815e chipset. Arvidjaar recommends waiting for Mandrake 7.2, citing Redhat 7 bugginess. Sounds smart to me.

Before closing, and finishing with the benchmarks page, I want to take a minute to mention <pimp> some other Radified Guides that you might find helpful. For example, there's the Guide to Norton Ghost, which is currently the most popular destination for Radified readers. There's also the ASPI Drivers Guide, the Guide to Ripping CD Audio & MP3 Encoding, the SCSI guide, the Guide to Partitioning a Hard Drive with FDISK.  

an't forget the Radified Hard Drive Partitioning Strategies Guide, a Benchmarks page, the Guide to Software Programs & Applications, and several others. </pimp>.

Next -> Asus CUSL2 Benchmarks