Date: Mon, 14 Jul 1997 08:39:20 -0700 From: "Jin Guojun[ITG]" <jin@george.lbl.gov> To: dg@root.com, drussell@saturn-tech.com Cc: FreeBSD-hardware@FreeBSD.ORG, brett@lariat.org, jkh@time.cdrom.com, jwm@CSUA.Berkeley.EDU, root@meeko.eecs.Berkeley.EDU Subject: Re: AMD K6 Message-ID: <199707141539.IAA11025@george.lbl.gov>
next in thread | raw e-mail | index | archive | help
I cat two messages together to respond since I missed most time last weekend. I did not see any one give more clear spce. First of all, I have AMD K6-166, bought when it just born, and have overclocked to 200 MHz, and so far it has no problem at all. I have post the result of its performance on "Hardware performance guide for Pentium family" under http://www-itg.lbl.gov/ISS/hardware The Cyrix 6x86, some one has tested, is little bit faster and much cheaper, but less compatiability. It DOES not work on most moatherboard, I was told. The ADM K6, generally, works on most new motherboards. The key is that it has to be HX or TX PCI chipset based motherboard. If you use other PCI chipset, such as VX, it is not guaranted to work. The ASUS PCI-P/I-P55TVP4 is the bottum line, and ASUS TX97 is very reliable one. No big suprise. These motherboard is about $140-$155 in the past. Uauslly, the price should be dropped, but it seems goes up by $5 this time. I wonder it is the reason for K6 compatiability. I doubt those $80 motherboard can work for K6. -Jin }> I ma currently test driving an AMD K6/200 in a }> FIC PA-2005 motherboard and I am getting a lot of crashes. }> (actually it looks like a reset, no cores, no console messages }> just a spontaneous reboot). I am wondering if anyone has seen }> the K6 do similar things. The vendor says "The K6 does support }> UNIX", but I am inclined to believe that he doesn't have a clue. }> I am thinking I will return the K6 and get a Cyrix M2, any thoughts? } }We've been totally unable to get the K6 to work reliably, nor have any }of the folks we've been talking to had much luck - the symptoms are }basically the same in each case, signal 11s all over the place and }make world failures. What's weirder is that it will often work just }fine for days or weeks, the initial make world tests going fine, and }then it will just stop, no make world making it through from that }point on. I'd never seen a CPU fail due to heat death, but it sure }seems like the only explanation here. What's more worrisome is that }the other 4 testers had exactly the same thing happen. } ----- another message ------- } }Anyway, FreeBSD, Inc. now owns an expensive piece of chip jewelry from }AMD since our 30 day return period also elapsed during testing. :( } }>> Have you tried the obvious controlled experiment: putting a Pentium into }>> the same motherboard? }> }>The problem is that there are still too many variables. The K6 and a }>plain Pentium use different voltages, etc. My best guess would be to try }>running the thing at a different voltage. Several people have said they }>had to up the voltage a little to overclock the chip. We might try }>bumping the voltage by .1 or .2 volts and see what happens. Checking }>cooling never hurts, either. } } Did that, no difference. I also changed the motherboard to a different }kind...no difference.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199707141539.IAA11025>