cliu@rnd.GBA.NYU.EDU (Crocker H. Liu) (11/16/90)
Alas! My 3b1 (long abandoned by ATT) is too puny to run X-windows. But the wealth of choice in 386's is overwhelming! I also have an AT clone, which is basically used by my wife as a word-processor. It only has Hercules graphics and a 44MB harddisk. Does it make sense to swap in a 386 motherboard? What about the Unix? Does the new Dell4.0 sound good to you? The accompanying software seems to be everything anyone would ever desire. Well, it has X-windows anyway, which is what I'm worried about now. (I'm also hooked on emacs and TeX, but I presume they will build on just about any hardware I might pick up.) The system will be used for program development, so it can't be too non-standard. Software portability is an issue. Help! --- Crocker Liu cliu@rnd.gba.nyu.edu
fangchin@elaine41.stanford.edu (Chin Fang) (11/16/90)
liu>I also have an AT clone, which is basically used by my wife as a
liu>word-processor. It only has Hercules graphics and a 44MB harddisk.
liu>Does it make sense to swap in a 386 motherboard?
I am afraid the answer is NO! Just a motherboard is not adqueate. You also
need a large and fast hard disk. Personally I quite like Cornor's 210Megs
3.5" SCSI(or IDE). It's quiet, much quiter than the internal hd in my
SUN sparc 1+, it is inexpensive too. about $750 (SCSI adapter not included)
and I have several times success using it with ESIX Rev. D R3.2 for both
IDE and SCSI types.
As to vedio subsystem, I feel your current one also won't let you take
advantage of most Unices can offer, especially X. A good combo for code
developement would be a Tseng ET4000 chip based SVGA ($300 or less) and
a NEC 4D. Total price ~= $1300.
386 33Mhz motherboards are quite stable now. So get one that allows you to
install at least 8 megs Simm on the motherboard and 8 Megs more on 32 bit
memory expension card (Make sure the slot is for 32bit!!!!)
Total price motherboard + 16 Megs memory(with expension card)
~= $850 + $37x16 + $95 = $1537
If you like to put on a few more sram cache memory on, get one that allows
you to equip up to 128K. Once you have more than 32K sram cache, you would
have about 8 MIPs on your desk, which is about 3/4th of Sun's Sparc 1's.
Not bad! $300 more you get a 486 isa motherboard, then using the memory
configuration above, your potential system beats my sparc 1+!!!!
Get a new minitower case with a healthy power supply (over 220 W!) $100 or so.
I don't think you need math-coprossor. That saves you $500 for a 33Mhz i387.
A good sized, non-interlaced monitor like NEC 4D is EXTREMElY important to
your eyes if you like to do program developement. I confess being spoiled by
workstations' display. But don't you like to protect your eyes? Besides,
running X even using 800x600 is not really that helpful for looking at codes.
Sixty some lines would be good for looking at an entire function body assuming
you programm in C(?)
Liu>What about the Unix? Does the new Dell4.0 sound good to you? The
Liu>accompanying software seems to be everything anyone would ever desire.
Liu>Well, it has X-windows anyway, which is what I'm worried about now.
Liu>(I'm also hooked on emacs and TeX, but I presume they will build on
Liu>just about any hardware I might pick up.)
As mentioned in FAQ, it's up to you to decide. I can't help you on that. I
don't have R4 yet. I am sure someone else will give you help on this one.
Liu>The system will be used for program development, so it can't be too
Liu>non-standard. Software portability is an issue.
Liu>
Liu>Crocker Liu cliu@rnd.gba.nyu.edu
I don't think you will go too wrong with today's 386 hardware assuming you won't
buy some really untested/unheard el'chippo and then put in lot's networking
hardware, then most likely you should be fine. A few lession I learned so far
(1) go with AMI BIOS, the later, the better.
(2) ESDI adapters typically give you less trouble than SCSI. (Cornor's IDE is
fine in all cases I tried, so you might want to give it a look)
(3) get Toshiba SIMMs.
That's all I can think of now and it's late. As always, I gave you my opinions,
but you will be the judge.
Good Luck
Chin Fang
Mechanical Engineering Department
Stanford University
fangchin@portia.stanford.edu
fang@rock.cadcam.rok.comjca@pnet01.cts.com (John C. Archambeau) (11/17/90)
fangchin@elaine41.stanford.edu (Chin Fang) writes: >liu>I also have an AT clone, which is basically used by my wife as a >liu>word-processor. It only has Hercules graphics and a 44MB harddisk. >liu>Does it make sense to swap in a 386 motherboard? > >I am afraid the answer is NO! Just a motherboard is not adqueate. You also >need a large and fast hard disk. Personally I quite like Cornor's 210Megs >3.5" SCSI(or IDE). It's quiet, much quiter than the internal hd in my >SUN sparc 1+, it is inexpensive too. about $750 (SCSI adapter not included) >and I have several times success using it with ESIX Rev. D R3.2 for both >IDE and SCSI types. > >As to vedio subsystem, I feel your current one also won't let you take >advantage of most Unices can offer, especially X. A good combo for code >developement would be a Tseng ET4000 chip based SVGA ($300 or less) and >a NEC 4D. Total price ~= $1300. > >386 33Mhz motherboards are quite stable now. So get one that allows you to >install at least 8 megs Simm on the motherboard and 8 Megs more on 32 bit >memory expension card (Make sure the slot is for 32bit!!!!) > >Total price motherboard + 16 Megs memory(with expension card) >~= $850 + $37x16 + $95 = $1537 > >If you like to put on a few more sram cache memory on, get one that allows >you to equip up to 128K. Once you have more than 32K sram cache, you would >have about 8 MIPs on your desk, which is about 3/4th of Sun's Sparc 1's. >Not bad! $300 more you get a 486 isa motherboard, then using the memory >configuration above, your potential system beats my sparc 1+!!!! A 486/33 motherboard will yield about 14 to 15 MIPS. But keep in mind the bottleneck going across the ISA bus. A 64-bit processor running on a 16-bit bus. Sort of reminds me of a traffic jam when a highway suddenly goes from 4 lanes to 1 (which is the correct ratio). Do NOT get a 486 unless you're going to go EISA or MCA. It's a waste of CPU bus bandwidth if you don't. >Get a new minitower case with a healthy power supply (over 220 W!) $100 or so. >I don't think you need math-coprossor. That saves you $500 for a 33Mhz i387. >A good sized, non-interlaced monitor like NEC 4D is EXTREMElY important to >your eyes if you like to do program developement. I confess being spoiled by >workstations' display. But don't you like to protect your eyes? Besides, >running X even using 800x600 is not really that helpful for looking at codes. >Sixty some lines would be good for looking at an entire function body assuming >you programm in C(?) > >Liu>What about the Unix? Does the new Dell4.0 sound good to you? The >Liu>accompanying software seems to be everything anyone would ever desire. >Liu>Well, it has X-windows anyway, which is what I'm worried about now. >Liu>(I'm also hooked on emacs and TeX, but I presume they will build on >Liu>just about any hardware I might pick up.) > >As mentioned in FAQ, it's up to you to decide. I can't help you on that. I >don't have R4 yet. I am sure someone else will give you help on this one. > >Liu>The system will be used for program development, so it can't be too >Liu>non-standard. Software portability is an issue. >Liu> >Liu>Crocker Liu cliu@rnd.gba.nyu.edu > >I don't think you will go too wrong with today's 386 hardware assuming you won't >buy some really untested/unheard el'chippo and then put in lot's networking >hardware, then most likely you should be fine. A few lession I learned so far >(1) go with AMI BIOS, the later, the better. What the (*bleep*) does the BIOS have to do with Unix other than booting the machine and getting system configuration information? When you can't do a CTRL-ALT-ESC or CTRL-ALT-S to get into your CMOS setup during your Unix boot up is when the services of the BIOS are no longer needed. Phoenix will work just as well. ALR's are Novell and SCO certified and they use a Phoenix BIOS. >(2) ESDI adapters typically give you less trouble than SCSI. (Cornor's IDE is > fine in all cases I tried, so you might want to give it a look) ESDI gets a better thoroughput over SCSI on the ISA bus by about a third. I have yet to see a SCSI host adaptor yield greater than 10 Mbit/second on the ISA bus. ESDI gives a 15 Mbit/second. A lot of your high performance IDE's are the same throughput as ESDI (Conner, Maxtor, and Imprimis). >(3) get Toshiba SIMMs. Why Toshiba? Memory is memory. NEC, TI, et. al. Only memory I stay away from is Oki since their chips have a high failure rate (at least that's my experience with their DRAM chips). You don't need all of that to run a 386 Unix. I've been doing quite well fending with a 386SX and an ST412/506 MFM 2:1 interleave controller. Now this is what's going to get you. I'm only running 50 Mb total on my system. I have one Seagate ST151 (42.5 Mb) and a Rodime RO352 (10.6 Mb) on my system. Af After you lay down SCO Xenix 386 2.3.3 complete with VP/ix, you have about 10 have gulped down about 32 Mb of space on your drive roughly. Your mileage may very depending on how much swap you allocate. I have everything for Xenix 386 save Xsight, Streams, TCP/IP, and Xenix-Net. As for speed. Well, I have a DigiBoard COM/4 (dumb board, now called a PC/4) which is just four NS16450 UART chips with a modem and a Televideo 910 terminal hooked to it now. Speed isn't affected dramatically (if at all) while the DigiBoard ports are being used. Sure, all of that glitter is nice, but one could easily get by with a pair of 40 Mb drives or a single 80 Mb. Only problem is if you want to allocate a DOS partition, then you have to add the space of the DOS partition to your drive. If you want a solid product, SCO Xenix is still a good route to go inspite of SCO has planned for it in the next year or two. I suspect that one could reasonably expect just as pleasing results with respect to bare necessities to system configuration with respect to uPort, ISC, Dell, Intel, ESIX, and UHC (I know I missed some). // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | Small memory model only for ** ARPANET : crash!pnet01!jca@nosc.mil | Unix? Get the (*bleep*) out ** INTERNET: jca@pnet01.cts.com | of here! ** UUCP : {nosc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
larry@nstar.uucp (Larry Snyder) (11/18/90)
jca@pnet01.cts.com (John C. Archambeau) writes: >If you want a solid product, SCO Xenix is still a good route to go inspite of >SCO has planned for it in the next year or two. Interactive Unix release 2.20 is very solid and disk throughput will blow SCO Xenix away.. -- Larry Snyder, Northern Star Communications, Notre Dame, IN USA {larry@nstar, {uunet|backbone}!nstar!larry, larry%nstar@iuvax.cs.indiana.edu} backbone usenet newsfeeds available Public Access Unix Site (219) 289-0282 (5 high speed lines)
rajs@hpindda.cup.hp.com (Rajeev Seth) (11/19/90)
> / hpindda:comp.unix.sysv386 / larry@nstar.uucp (Larry Snyder) / 3:08 pm Nov 17, 1990 / > jca@pnet01.cts.com (John C. Archambeau) writes: > > >If you want a solid product, SCO Xenix is still a good route to go inspite of > >SCO has planned for it in the next year or two. > > Interactive Unix release 2.20 is very solid and disk throughput will > blow SCO Xenix away.. > > > -- > Larry Snyder, Northern Star Communications, Notre Dame, IN USA I agree absolutely. There could be some improvements in that there still a few bugs, say, in the sysmgmt program, but their tech support is pretty good, so it doesn't hurt much.
james@bigtex.cactus.org (James Van Artsdalen) (11/20/90)
In <5682@crash.cts.com>, jca@pnet01.cts.com (John C. Archambeau) wrote: > A 486/33 motherboard will yield about 14 to 15 MIPS. But keep in > mind the bottleneck going across the ISA bus. A 64-bit processor > running on a 16-bit bus. There's nothing 64-bit about a 486. It does have 128-bit cache lines. > Sort of reminds me of a traffic jam when a highway suddenly goes > from 4 lanes to 1 (which is the correct ratio). Do NOT get a 486 > unless you're going to go EISA or MCA. It's a waste of CPU bus > bandwidth if you don't. I disagree here. The data coming off of the hard disk is much less than the bandwidth of the AT bus. Therefore you can't win with just an EISA hard disk controller. What you CAN win with is a caching controller, or a controller that can do DMA & has a unix driver that can use it, or some other optimization not related to the bus bandwidth. Caching helps ISA too. It's no accident that the EISA hard disk controllers that are appearing have better hardware support (DMA, cache) than their ISA counterparts: they wouldn't be much faster if they didn't (note: DMA is slower on ISA anyway). Also note that the fastest EISA Ethernet cards are maybe 5% faster than the 8-bit WD-8003. 20Mbytes/sec bus bandwidth isn't needed to communicate over a 10Mbit wire. I will be buying EISA stuff for bigtex, but for the caching and DMA features. | (3) get Toshiba SIMMs. > Why Toshiba? Memory is memory. NEC, TI, et. al. No. Completely wrong. Look at a DRAM specification sheet some day. There are at least twenty different parameters that must be met. They are are a little different for each SIMM that plugs into the same socket. Part of the design effort in a system is to allow as many different SIMMs to be used as possible, and then make sure you *don't* ship the ones that won't work. RAS precharge time & friends: all that fun stuff. -- James R. Van Artsdalen james@bigtex.cactus.org "Live Free or Die" Dell Computer Co 9505 Arboretum Blvd Austin TX 78759 512-338-8789
flint@gistdev.gist.com (Flint Pellett) (11/21/90)
jca@pnet01.cts.com (John C. Archambeau) writes: >A 486/33 motherboard will yield about 14 to 15 MIPS. But keep in mind the >bottleneck going across the ISA bus. A 64-bit processor running on a 16-bit >bus. Sort of reminds me of a traffic jam when a highway suddenly goes from 4 >lanes to 1 (which is the correct ratio). Do NOT get a 486 unless you're going >to go EISA or MCA. It's a waste of CPU bus bandwidth if you don't. You don't need EISA in order to justify a 486: If you need floating point, and you are looking at a 33 MHz 386 plus a 33 MHz 387, you ought to look at a 25 MHz 486 instead: it will run a lot faster and probably cost less, because you don't have to spend the money for the 387. On your "traffic jam" analogy: on what type of highway do you think you'll be able to drive faster: one that is one-lane wide it's entire length, or one that is 2 lanes wide half the time (when going through cities) and 1 lane wide the rest of the time? Just because you can't afford to pay for 2 lanes for the entire distance (486 & EISA) doesn't mean that you can't get a lot of benefit from a partial, lower cost solution. -- Flint Pellett, Global Information Systems Technology, Inc. 1800 Woodfield Drive, Savoy, IL 61874 (217) 352-1165 uunet!gistdev!flint or flint@gistdev.gist.com
jca@pnet01.cts.com (John C. Archambeau) (11/21/90)
james@bigtex.cactus.org (James Van Artsdalen) writes: >In <5682@crash.cts.com>, jca@pnet01.cts.com (John C. Archambeau) wrote: >I disagree here. The data coming off of the hard disk is much less >than the bandwidth of the AT bus. Therefore you can't win with just >an EISA hard disk controller. What you CAN win with is a caching >controller, or a controller that can do DMA & has a unix driver that >can use it, or some other optimization not related to the bus >bandwidth. Caching helps ISA too. > >It's no accident that the EISA hard disk controllers that are >appearing have better hardware support (DMA, cache) than their ISA >counterparts: they wouldn't be much faster if they didn't (note: DMA >is slower on ISA anyway). > >Also note that the fastest EISA Ethernet cards are maybe 5% faster >than the 8-bit WD-8003. 20Mbytes/sec bus bandwidth isn't needed to >communicate over a 10Mbit wire. The reason for wanting EISA specific cards is because of the bus mastering capabilities. Admittedly the throughput isn't much, but I'd rather have all bus mastering cards on a bus that can handle bus mastering cards. Makes sense. If I pay for a bus mastering bus, I certainly want bus mastering cards. This goes for EISA or MCA. >I will be buying EISA stuff for bigtex, but for the caching and DMA >features. > >| (3) get Toshiba SIMMs. > >> Why Toshiba? Memory is memory. NEC, TI, et. al. > >No. Completely wrong. Look at a DRAM specification sheet some day. >There are at least twenty different parameters that must be met. They >are are a little different for each SIMM that plugs into the same >socket. Part of the design effort in a system is to allow as many >different SIMMs to be used as possible, and then make sure you *don't* >ship the ones that won't work. RAS precharge time & friends: all that >fun stuff. I have yet to find a memory module that you can't plug in. Whether it be a Mac, Sun, 286, or 386 box. If you buy chips from places such as the Chip Merchant, you are subject to what they have in stock. Admittedly, some memory modules work better than others. My 386SX motherboard manual lists a long list of memory that has been tested to work by manufacturer and their part number. But the chances of pulling a particular DRAM chip off the shelf and having it work are VERY high if it's a well designed board. Remember the memory chip shortage not to long ago? A lot of chip vendors just plain weren't selling in the USA, so you took what you could get. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | Small memory model only for ** ARPANET : crash!pnet01!jca@nosc.mil | Unix? Get the (*bleep*) out ** INTERNET: jca@pnet01.cts.com | of here! ** UUCP : {nosc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
richard@pegasus.com (Richard Foulk) (11/22/90)
>I have yet to find a memory module that you can't plug in. Whether it be a >Mac, Sun, 286, or 386 box. If you buy chips from places such as the Chip >Merchant, you are subject to what they have in stock. That doesn't follow my experiences. I've found that a very popular brand of SIMMs doesn't work reliably in a couple different brands of motherboards I've seen. Samsung; doesn't work with Mylex and a no-name I've used. Sun machines also seem to be brand specific. >Admittedly, some memory modules work better than others. My 386SX motherboard >manual lists a long list of memory that has been tested to work by >manufacturer and their part number. In the universe I live in, either your RAM works or it doesn't. -- Richard Foulk richard@pegasus.com
tmh@bigfoot.FOKUS.GMD.DBP.DE (Thomas Hoberg) (11/23/90)
In article <49991@bigtex.cactus.org>, james@bigtex.cactus.org (James Van Artsdalen) writes: |> In <5682@crash.cts.com>, jca@pnet01.cts.com (John C. Archambeau) wrote: |> |> > A 486/33 motherboard will yield about 14 to 15 MIPS. But keep in |> > mind the bottleneck going across the ISA bus. A 64-bit processor |> > running on a 16-bit bus. |> |> There's nothing 64-bit about a 486. It does have 128-bit cache lines. |> |> > Sort of reminds me of a traffic jam when a highway suddenly goes |> > from 4 lanes to 1 (which is the correct ratio). Do NOT get a 486 |> > unless you're going to go EISA or MCA. It's a waste of CPU bus |> > bandwidth if you don't. |> |> I disagree here. The data coming off of the hard disk is much less |> than the bandwidth of the AT bus. Therefore you can't win with just |> an EISA hard disk controller. What you CAN win with is a caching |> controller, or a controller that can do DMA & has a unix driver that |> can use it, or some other optimization not related to the bus |> bandwidth. Caching helps ISA too. If EISA and ISA were to cost the same or perhaps 10-20% difference, I'd certainly recommend EISA, even though right now, there very little to be gained by it in terms of performance. I agree that the current bandwidth provided by the ISA bus is sufficient for disks and frame buffers. The combination of an Adaptec 154x bus mastering SCSI controller and about five 3 1/2" 200MB SCSI drives and a disk striping driver for ISC's HPDD should be difficult to beat (not match) for any EISA board right now. I got a Chips and Technologies based Taiwan i486 clone for about a third of the price I'd have to pay for any EISA board right now. Sure, it doesnt support burst mode DRAM accesses and second stage caches, but I tend to think that those bells and whistles hurt the bang-to-the-buck ratio more than the're really worth. If what everybody keeps telling us is true, Multi Media is comming fast. That's when the EISA bus will become a must: With FDDI networks carrying real-time compressed HDTV video data, the ISA bus will become obsolete. That is probably still one or two years away yet. By then you will probably junk your present system anyway--computers are like kleenex:before you had them, you wouldn't have thought you'd need them and you wouldn't want to keep the same one around for any extendet period of time... ---- Thomas M. Hoberg | UUCP: tmh@prosun.first.gmd.de or tmh%gmdtub@tub.UUCP c/o GMD Berlin | ...!unido!tub!gmdtub!tmh (Europe) or D-1000 Berlin 12 | ...!unido!tub!tmh Hardenbergplatz 2 | ...!pyramid!tub!tmh (World) Germany | BITNET: tmh%DB0TUI6.BITNET@DB0TUI11 or +49-30-254 99 160 | tmh@tub.BITNET
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (11/23/90)
In article <1026@gistdev.gist.com> flint@gistdev.gist.com (Flint Pellett) writes: jca@pnet01.cts.com (John C. Archambeau) writes: >A 486/33 motherboard will yield about 14 to 15 MIPS. But keep in mind the >bottleneck going across the ISA bus. A 64-bit processor running on a 16-bit >bus. Sort of reminds me of a traffic jam when a highway suddenly goes from 4 >lanes to 1 (which is the correct ratio). Do NOT get a 486 unless you're going >to go EISA or MCA. It's a waste of CPU bus bandwidth if you don't. I'm sure somewhere there's someone who is running memory on the AT bus instead of the motherboard or a memory bus, but in general that's not the case. What is going to be backlogged by not using EISA? Not the 8 bit serial ports, or the eight bit parallel ports, or the 10-20Mbit disk drives, or the 16 bit video cards... The bandwidth is about 8MHz clock / 4 clocks per xfer, times 16 bits per xfer, or 32Mbit. Any device using less than 16 bit xfer is a bottleneck on an EISA bus, too. You are technically correct that if you have an EISA bus and EISA peripherals running in 32 bit mode, you can save a few bus cycles, but the AT bus is easily fast enough to keep up with normal peripherals. To imply that you will notice (or even be able to measure) the performance loss due to the slow bus is probably incorrect. The only thing you might be able to actually see would be a 32 bit video card, and then only with a custom driver, since current unix drivers usually work on a single byte most of the time (yes they might use string copy for bitblt). Having experience with many AT bus machines at work, and a small number (several dozen?) of EISA machines, I can tell you that several manufacturers do not present the same bus timing to AT cards on the EISA bus as the accepted AT bus timing for ISA, and that some AT cards won't work. If you are doing industrial control and use a *lot* of special cards which are finicky, you might well find that the ISA bus is still desirable. If you just want to buy commodity cards off the shelf and be sure they will work, you still might make a trade-off of incremental performance for reliability. I am *not* disagreeing that there is some performance loss, just saying that running benchmarks on the same manufacturer's ISA and EISA 486, that I didn't see it using typical 16 bit video and cached disk controller w/ 15Mbit disk. I don't think this is a serious issue, and I believe that people should choose the bus based on intended peripherals, and stay with ISA if the EISA is not going to be used. -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc and 80386 mailing list "Stupidity, like virtue, is its own reward" -me
james@bigtex.cactus.org (James Van Artsdalen) (11/25/90)
Quick preface: I don't argue that you can't buy SIMMs third party. As long as you stick with memory the manufacturer specifies you're OK (assuming the SIMMs weren't sorted - an unusual thing for SIMMs I hope). But buying Joe's Discount 80ns DRAM may or may not work - there may be good reason that the manufacturer doesn't use it. In <5763@crash.cts.com>, jca@pnet01.cts.com (John C. Archambeau) wrote: > Why Toshiba? Memory is memory. NEC, TI, et. al. | No. [...] Look at a DRAM specification sheet some day. There are | at least twenty different parameters that must be met. > I have yet to find a memory module that you can't plug in. These chips have nanosecond tolerances of many different types. A specification with a 30% tolerance may mean 13ns instead of 10ns. Just because it "plugs in" and passes POST tests while cold doesn't mean that you might not be eating deep into margin in a few minutes as the box heats up. > My 386SX motherboard manual lists a long list of memory that has > been tested to work by manufacturer and their part number. YES!!! That's because *those* SIMMs have been tested and work. If the vendor knew that every SIMM would work, why bother printing this list? > But the chances of pulling a particular DRAM chip off the shelf and > having it work are VERY high if it's a well designed board. I claim the opposite result: a well designed board will accept fewer types of SIMMs than a poorly designed board. If you consider speed or wait states a design goal, then you will design for tighter tolerances to eliminate wait states. Those tighter tolerances mean that fewer SIMMs meet your specification. A design that trades away performance for cost will have more wait states. This is not only easier & cheaper to design, but it means that more SIMMs will meet specification, and hence the memory cost is less. Carried to an extreme, this is a bad idea of course: a design with only one usable SIMM is no good from a manufacturing standpoint. > Remember the memory chip shortage not to long ago? A lot of chip > vendors just plain weren't selling in the USA, so you took what you > could get. After it passed engineering design check, systems validation, and environmental testing... At that time, a lot of people went back and tried to loosen design constraints. Maybe the specifications on the memory controller had tightened up since the design was shipped, maybe you could use a faster PAL: lots of things could let you loosen the SIMM requirements. -- James R. Van Artsdalen james@bigtex.cactus.org "Live Free or Die" Dell Computer Co 9505 Arboretum Blvd Austin TX 78759 512-338-8789