bear@bgsuvax.UUCP (Michael D. Bear) (08/16/89)
I recently took a close look at a 386sx machine based on a Mylex motherboard. I was fairly impressed with it, compared to what I've read on the net in the past. My question is, why would you reccommend against a Mylex motherboard. I have read bad reccommendations in the past, but no real reasons why. Please respond with any experiences, good or bad. The machine I looked at, had Phoenix BIOS, Phoenix/VLSI chipset, 7 16 bit slots, and 1 8 bit slot, and appeared to be of the highest quality.
allred@ut-emx.UUCP (Kevin L. Allred) (08/16/89)
In article <4843@bgsuvax.UUCP>, bear@bgsuvax.UUCP (Michael D. Bear) writes: > > I recently took a close look at a 386sx machine based on a Mylex > motherboard. I was fairly impressed with it, compared to what I've read > on the net in the past. My question is, why would you reccommend against > a Mylex motherboard. I have read bad reccommendations in the past, but no > real reasons why. Please respond with any experiences, good or bad. > The machine I looked at, had Phoenix BIOS, Phoenix/VLSI chipset, 7 16 bit > slots, and 1 8 bit slot, and appeared to be of the highest quality. ^^ don't forget 1s and 1p too I have one of these boards running in my system now. It works almost perfectly, so far. I ran into shadow BIOS incompatibility, and it is only a minor inconvenience, since I can map any memory not used for shadowing out into the extended address space. I have a rather unusual hardware configuration, but judging by past posting the shadowing incompatibility is probably the fault of the Pheonix BIOS rather than the cards. For your info, my configuration is: Mylex MXS-16 motherboard w/ 4MB (ie. 4 1MB 100ns SIMMS) Video-7 Fastwrite VGA card w/ 256KB and Magnovox Mono VGA monitor. Seagate ST02 SCSI host addaptor and ST296N 84 MB SCSI HD. When I enable Video BIOS shadowing the ST02 can't find the HD. I tried changing the ST02 BIOS location to different places, but the problem wouldn't go away. My solution was to dissable video BIOS shadowing (System BIOS shadowing works fine); remap the RAM out into the extended memory space, and install the RAMBIOS driver shipped with the Fastwrite in my config.sys file. I wish I had a RAM installable driver for the SCSI BIOS as maybe then I could get transfers to take place fast enough to format the drive with a 1:1 interleave -- Oh well 450+ KBPS with a 2:1 interleve isn't slow, and a big RAM cache makes the actual drive speed almost unimportant. I don't know that not having the SCSI bios in RAM is what is preventing the transfers from going fast enough, but I do know that other people with 16MHz machines have claimed to be able to get the 900+ KBPS transfer rates. The system has a Norton 4.5 Si rating of 16.(5-6) at 16MHz and about 11 at 8Mhz (you figure why it is not half). I haven't had any normal DOS software that wouldn't run. I have run VMOS/3 which runs in protected mode. VMOS seems to work fairly well, but is not very robust -- It is after all a young product. I have been able to get 4 programs (the machine wasn't anywhere near out of resources so I probably could have started more) running simultaneously under VMOS -- some of them were even games. I wish I had UNIX to try out, but I'm waiting for GNU or UNIX prices to drop. I do know that Intel says their UNIX won't run on a 386sx ( you figure the company that makes the processor would know what to change :-). I don't know about the other UNIX vendors. I am happy with the Mylex board. It is a good deal for a entry level 386 board. I bought it from: MicroSource Distributors (800)-326-4276 $483 0KB shipped 2nd day UPS. PS. If anybody has any suggestions for how to solve the shadowing conflict or speed up the SCSI transfers, I would love to hear them. Buying new cards isn't the solution I want to hear :-) -- Kevin Allred allred@emx.cc.utexas.edu allred@ut-emx.UUCP
davidsen@sungod.crd.ge.com (ody) (08/17/89)
In article <17144@ut-emx.UUCP> allred@ut-emx.UUCP (Kevin L. Allred) writes: | UNIX prices to drop. I do know that Intel says their UNIX won't run | on a 386sx ( you figure the company that makes the processor would | know what to change :-). I don't know about the other UNIX vendors. | I am happy with the Mylex board. It is a good deal for a entry level | 386 board. I bought it from: Has anyone else verified this? Several 386SX vendors have sworn to me that their systems would run Xenix and UNIX. Is this something particular to Intel? I assume that you can tell if the system is an SX or not, but why do it? bill davidsen (davidsen@crdos1.crd.GE.COM) {uunet | philabs}!crdgw1!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me
allred@ut-emx.UUCP (Kevin L. Allred) (08/18/89)
In article <1711@crdgw1.crd.ge.com>, davidsen@sungod.crd.ge.com (ody) writes: > In article <17144@ut-emx.UUCP> allred@ut-emx.UUCP (Kevin L. Allred) writes: > | UNIX prices to drop. I do know that Intel says their UNIX won't run > | on a 386sx ( you figure the company that makes the processor would > | know what to change :-). I don't know about the other UNIX vendors. > Has anyone else verified this? Several 386SX vendors have sworn to me > that their systems would run Xenix and UNIX. Is this something > particular to Intel? I assume that you can tell if the system is an SX > or not, but why do it? Since my earlier posting, I talked to a sales rep. for SCO and he said that their 386 products would run on a 386sx no problem, but remember you can believe everything a sales rep tells you ;-). He did ask if I had an AT bus, and it was after I said yes, that he said no problem. Maybe there is problem in microchannel based 386sx systems. Intel wasn't very clear as to why their current unix product wouldn't run on a 386sx. It had something to do with not being able to do full 32 bit memory accesses in the first megabyte of memory (huh?). I have know idea why this would be a problem, as it should just take longer to do 32 bit accesses. As to being able to tell the difference between a 386 and 386sx, all of the 386 specific software I have tried so far has identified the processor as a 386 and run as advertised. To tell the difference, I think software would have to count the nember of clock cycles to do 32 bit memory accesses. I guess the time would be around twice as long for the sx. -- Kevin Allred allred@emx.cc.utexas.edu allred@ut-emx.UUCP
aland@infmx.UUCP (Dr. Scump) (08/22/89)
In article <17241@ut-emx.UUCP> allred@ut-emx.UUCP (Kevin L. Allred) writes: >In article <1711@crdgw1.crd.ge.com>, davidsen@sungod.crd.ge.com (ody) writes: >Since my earlier posting, I talked to a sales rep. for SCO and he said >that their 386 products would run on a 386sx no problem, but remember >you can believe everything a sales rep tells you ;-). He did ask if I >had an AT bus, and it was after I said yes, that he said no problem. >Maybe there is problem in microchannel based 386sx systems. Intel >... > Kevin Allred You might ask said salesperson why, if they will run on a 386SX without problem, they do not list a single 386SX-based machine on their current compatibility list ("Configuration Guide") -- not even the Compaq or the IBM (PS/2 Model 55Z or whatever). I just got the list from them last week; it's dated June 1989. -- Alan S. Denney @ Informix Software, Inc. {pyramid|uunet}!infmx!aland "I want to live! -------------------------------------------- as an honest man, Disclaimer: These opinions are mine alone. to get all I deserve If I am caught or killed, the secretary and to give all I can." will disavow any knowledge of my actions. - S. Vega