Will this work, is BIOS update needed?

Recommended by Wim's BIOS...

by **greenknight** » Fri Oct 30, 2009 3:41 pm

I have an old computer built from an iSystems bare-bones kit in 2000. The sticker on the motherboard says it's a Procomp BVK1M V.1.0; apparently it was made by Biostar - the ID number matches the Biostar M7VKA except for the date, which is 10/07/2000 instead of 10/06/2000 like the M7VKA. Here's the info I get from Everest:

Device Description
North Bridge: VIA VT8363(A) Apollo KT133(A)

Motherboard ID 10/07/2000-8363-686A-6A6LMB09C-00

Motherboard Name Biostar M7VKA/M7VKB/M7VKD / Procomp BVK2A

Award BIOS Type Award Modular BIOS v6.00PG

Award BIOS Message K1M1006

Ran BiosAgent Plus for confirmation, it gave the same ID number.

Currently, it's running an AMD Duron 700 Mhz CPU. I've got a Duron 1200 Mhz I want to put in instead. From the research I've done, I think this will work, and without a BIOS update - is this correct? (This is the main thing I want to know).

The only BIOS update Biostar has for the M7VKA doesn't appear to contain anything that applies.

Also, I'm wondering about this chipset. Everest says it's a KT133A, instead of the KT133 that is normally in the M7VKA; CPU-Z and Fresh Diagnose say the same. Could that be right? Could it run faster CPUs that require a 266 Mhz FSB? Are there other BIOS updates it would accept?

by cp » Fri Oct 30, 2009 8:36 pm

I've got a Duron 1200 Mhz I want to put in instead. From the research I've done, I think this will work, and without a BIOS update - is this correct? (This is the main thing I want to know).

yes, it should work without an update. as long as the bios doesn't stop the machine from booting everything should be fine.

Everest says it's a KT133A, instead of the KT133 that is normally in the M7VKA

the KT133 and KT133A have the same pci id. they could be distinguished by their (silicon) revision however. afaik there are no offcial documents that say from which revision the chipset is/should be a KT133A. even the Linux kernel doesn't know. but apart from the 133MHz FSB there are no real differences between them.
if you want to know for sure you have to take a look at the chip printing of the northbridge. 8363 is a KT133 and 8363A is a KT133A.

Could it run faster CPUs that require a 266 Mhz FSB? Are there other BIOS updates it would accept?

The KT133 just supports 100MHz FSB for the CPU and will most probably fail if set to 133MHz. some motherboards have PLLs that allow 133MHz FSB with the KT133 but i haven't seen one running flawlessly.

by **greenknight** » Sat Oct 31, 2009 9:44 am

Ok, thanks. I guess if it doesn't boot, I'll just put the old CPU back in.

I realize that it won't support more than 100 MHz clock if it's actually a KT133. Not planning to do any more with it right now, anyway, except add more RAM. Sometime in the future I may try playing around with it, if so I'll pull it out and see what it really is.

by **greenknight** » Wed Nov 04, 2009 10:13 am

Got the Duron 1200 installed and it's working fine, running at 1204 Mhz. Now I just need to add more RAM, and this will be a halfway decent computer.

by **greenknight** » Tue Mar 02, 2010 3:55 pm

I realized I forgot to come back and post what's not working. In POST, the CPU is identified as an Athlon 1200. Cache size is incorrectly detected, and I don't suppose SSE is enabled, either. While a BIOS update might fix this, I don't consider it worth risking - it's much faster than it was before, and that's all I hoped to achieve.

Haven't had any problems with it.

By the way, I upgraded the cooler at the same time - nobody should try this with the stock cooler.

by cp » Wed Mar 03, 2010 9:13 pm

SSE-capabilities are available through the cpuid command. they should be on by default but you can - as always - check with some tool like cpu-z

by **greenknight** » Thu Mar 04, 2010 3:55 pm

Other info I found while researching this upgrade, such as this page: http://www.duxcw.com/dcforum/DCForumID5/534.html, says that cpuid only tells you if the processor supports SSE - not if it's active. These sources indicate that I would have to both flash the BIOS and do a clean reinstall of the OS to get it working, as SSE support is detected only at installation (takes a lot of digging to find out about stuff this old, let me tell you).

Not worried about it, I'm satisfied that my cheap upgrade worked. I'm not going to risk bricking the mobo in search of a little better performance right now. Maybe when I'm a little less broke...

by **Denniss** » Thu Mar 04, 2010 7:58 pm

That SSE statement you found is BS. If the CPU supports it and it's activated by Bios then you have it usable if programs request it.

by **greenknight** » Fri Mar 05, 2010 12:48 pm

That's good, but the BIOS apparently doesn't do it; CPU-Z shows MMX+ and 3-D Now+, but not SSE. The only BIOS update Biostar has for the M7VKA would fix the cache detection problem, but not SSE. For that, I'd have to flash the BIOS with something not officially recommended for this board, not going to try that.

The cache detection thing isn't worth the trouble to fix, so I'm leaving the BIOS alone for now.

by **Denniss** » Sat Mar 06, 2010 11:26 am

My Elitegroup K7S5A had a similar problem, the Bios did not activate SSE for Athlon XP based on Thoroughbred-cores (and newer/related CPU) although these CPU were officially supported. They fixed in in a later Bios but without mentioning it as fixed.
It may be worth testing the latest Bios.

by **greenknight** » Sat Mar 06, 2010 4:40 pm

Not quite the same situation; in my case, the CPU isn't officially supported - Morgan-core Durons didn't exist yet when the board was manufactured. I've seen claims that BIOS updates for later releases in this board series would work, but I don't want to chance it at this time. Just trying to get by with this old computer for a while longer; once I get a replacement, I'll be more free to tinker with it.

Probably make it a Linux box.

by cp » Sat Mar 06, 2010 7:44 pm

since you use Windows anyway: get WCPUID and simply enable SSE.
tutorial is here : http://www.myplc.com/sony/sse_enabling_and_wcpuid.htm

well, yes. and this is what the kernel sources say:

/*
* Bit 15 of Athlon specific MSR 15, needs to be 0
* to enable SSE on Palomino/Morgan/Barton CPU's.
* If the BIOS didn't enable it already, enable it here.
*/

so you can use CrystalCPUID, too. it has a nice build-in MSR browser and editor (evil stuff btw!)

by **greenknight** » Sun Mar 07, 2010 2:53 pm

Interesting, but no go - I downloaded WCPUID, they only thing in the "Tweaks" menu was Real Time Clock Checker. The other solution mentioned on that page with the tutorial, a little file you put in the startup folder, didn't work either; when I rebooted, I got an error message that the driver failed to start , and that's all. Oh, well; WCPUID is a nice utility, anyway, so it wasn't a total waste of time.

I'm going to check out CrystalCPUID, too, though the editing you describe is a bit over my head.

<edit> Got CrystalCPUID, no way I'm going to try that. Wayyy over my head. <end edit>