How the IBM PC became the de facto standard for desktop computers

I saw a question on a vintage computing forum this week: How did the IBM PC become the de facto standard for PCs, and the only desktop computer architecture from the 1980s to survive until today?

It’s a very good question, and I think there were several reasons for it. I also think without all of the reasons, the IBM PC wouldn’t have necessarily won. In some regards, of course, it was a hollow victory. IBM has been out of the PC business for a decade now. Its partners Intel and Microsoft, however, reaped the benefits time and again.

Read more

Building a Win95 box

Building a Windows 95 box? Why? You nuts?
Why not? You’ve got old hardware, you’ve got a ton of licenses to run an obsolete operating system… It’s a good match. Remember, a Pentium-120 was a titan of a PC in 1995. You couldn’t get anything faster. Running Windows 95 on a Pentium-120 with 24 MB RAM, 1.2 GB HD, and 8X CD-ROM in 1995 seemed like running Windows 2000 on a decked-out 1.4 GHz Athlon today. Maybe it seemed even more extreme than that; I remember selling a good number of 486DX2/66s and DX4/100s in the summer of 1995. They were low-end, yes, but they were at that $1,000 sweet spot. You’d pick up a DX2/66 for $800 and a 14″ monitor for $200, and sometimes as a weekend special we’d bundle the two together with a printer for $1,099 or something.

We had a Pentium-120 to rebuild at work, and we had its Win95 license, so it made sense to just rebuild it with the stuff it had. I know Jerry Pournelle had a really hard time building a Win95 box a few months back. I didn’t have much trouble at all, so I might as well document the pitfalls.

First of all, I used vintage hardware. That helps. Win95 was designed for 1995-era hardware. This PC probably dates from 1996 or so; it has the strange pairing of an Intel 430HX chipset and a Pentium-120. The 120 was more frequently bundled with the earlier 430FX chipset; by the time of the HX, the 133 was considered low-end, the 200 high-end, and the 166 was mainstream. The video card was a plain old Cirrus Logic-based PCI card; no issues there. AGP sometimes threw Win95 for a loop. None of that here. While DMA drivers certainly improved the 430HX, they weren’t necessary for stable performance. In other words, a 430HX-based board with a Cirrus video card works acceptably straight out of the box, with no additional drivers.

Other hardware: A Mitsumi 8X CD-ROM. I don’t remember exactly when 8X came out, but for the most part an IDE CD-ROM is an IDE CD-ROM, from a driver standpoint. A Creative Labs Sound Blaster 16. That was a very common, very well-supported sound card. A DEC 450 network card. Those DEC cards can be a real pain to get working sometimes, but Win95 surprised me and detected it straight up.

But Setup wouldn’t run initially. It took some figuring, but I solved that problem. My colleague had booted with a Win98 boot disk I made over a year ago. He did an FDISK and format to wipe the drive, but he formatted the drive FAT32. The original Win95 didn’t know about FAT32, so Setup was throwing a hissy fit when it saw it. I did another FDISK and format, switched to plain old FAT16, and Setup installed very happily.

Once I got Setup to run, it installed, and quickly at that. And with absolutely no issues. Remember, Win95’s footprint was only about 35 megs. It doesn’t take long for an 8X drive to deliver 35 megs. And the system booted quickly. I didn’t sit down and time it, but I’m used to calling a minute a reasonably fast boot time, and this thing didn’t seem slow to me at all. A little optimization would help, of course. A little logo=0 in c:msdos.sys goes a long way.

Running Win95 on newer hardware is possible, but remember, it’s been nearly four years since it was the mainstream OS. And you can have a lot of headaches trying to do it. Windows 3.1 is in the same boat–it’s downright hard to find device drivers for modern video cards. Then again, I can think of circumstances under which I’d want to run Win95. I can’t think of any compelling reason whatsoever to run Win3.1 at this point in time. (And there wasn’t any compelling reason to run it in 1994 either.)

If I had to build up a Win95 box today and could have whatever components I wanted, I’d probably look for an Asus P55T2P4, easily the best Socket 7 motherboard ever manufactured. (In 1997 when I was in the market, I opted for an Abit IT5H instead and I’m still kicking myself.) That board is most naturally paired with a Pentium-MMX/233, but with unsupported–but widely-documented online–voltage settings, you can run more recent K6-2 CPUs on it. The P55T2P4 allows an FSB of up to 83 MHz, but for stability’s sake, I’d keep it at 66 MHz, or possibly 68 MHz if the board supports it (I don’t remember anymore). You can run a K6-2/400 with a 6x multiplier at either of those settings and be very close to its rated speed. Then I’d use an ATI Xpert 98 video card. Yes, it’s a bit old, but it’s probably the best all-around PCI card that’s still reasonably easy to find. Win95 won’t recognize it without manufacturer-supplied drivers, of course, but that’s not so bad. This combination would give you surprisingly good performance, stability, and minimal difficulty of installation.

Anyway, that adventure reminded me that a Pentium-120 can still be a viable computer. Vintage software like Win95 runs well on it. Office 95 has more features than most of us use, and it’s faster and more stable than the recent incarnations. It also has fewer strings attached. IE 5.01, although recent, would run decently on a P120, as long as you left out Active Desktop. Acrobat Reader 3.0 will still read the majority of PDF files on the Web, and it’s smaller and faster-loading than more recent versions. Do a Web search; you can still find it online.

Don’t get carried away with what you install, and a P120 can certainly surprise you.

Two chipsets from the AMD front

Yesterday AMD formally unveiled and shipped the AMD-760MP chipset. Right now there is one and only one motherboard using it, the ritzy Tyan Thunder K7, which runs about $550 minimum. (Wholesale cost on it is rumored to be $500.) Considering its 64-bit PCI slots, two built-in 3Com NICs, onboard ATI video, onboard Adaptec SCSI, and four available DIMMs, that’s not a half-bad price. It’s obviously not a hobbyist board. This dude’s intended to go in servers.
At any rate, reviews are all over the place and the quality varies. Far and away the best I found was at Ace’s Hardware, where he tested the things people actually likely to buy this board would do with it: workstation-type stuff.

Anand does his usual 10 pages’ worth of butt-kissing and he’s living under the delusion that people will buy this board to play Quake. However, he does test the board with plain old Thunderbird and Duron CPUs (they work, but AMD won’t support that configuration). Skip ahead to page 11 after reading the story at Ace’s. His tests suggest that for some purposes, a dual Duron-850 can be competitive with a dual P3-933. That information is more interesting than it is useful at this point in time, but we’ve all been curious about dual Duron performance, so if and when an inexpensive AMD SMP board becomes available, we have some idea what we’ll be able to do with it.

All the usual hardware sites put in their two cents’ worth; by the time I read Ace’s and Anand’s and Tom’s reviews I stopped learning anything new.

Some of it bordered on ridiculous. One site (I forget which) observed that the AMD 766 northbridge looks just like a K6-2 and said they must have made it look that way just to remind us where the Athlon came from. Whatever. The AMD 766 northbridge and the K6-2 use the same heat spreader. The intention is to keep the chip cool. It’s not there just for looks–the chip runs hot. But that’s the kind of quality information we get from most hardware sites these days, sadly.

More immediately useful and interesting, but not yet available, is the nVidia nForce chipset. You can read about it at Tom’s and elsewhere. This is technically nVidia’s second chipset, their first being the chipset in Microsoft’s X-Box. This chipset is a traditional two-chip solution, linked by AMD’s high-speed HyperTransport. It includes integrated sound better than anything Creative Labs or Cirrus Logic currently offer (now we know what nVidia was doing with those engineers they were hiring from Aureal) and integrated GeForce 2MX video connected via a high-speed port that would be equivalent to AGP 6X, if such a thing existed. And nVidia pairs up DDR controllers to give dual-channel, 128-bit memory with a bandwidth of 4.256 GB/sec. Suddenly DDR provides greater bandwidth than Rambus in addition to lower latency.

Just for good measure, the chipset includes Ethernet too.

What’s all this mean? High-speed motherboards with everything integrated (and with integrated peripherals definitely worth using) for around 200 bucks. By the end of the summer, last summer’s monster PC will be integrated onto two chips and priced for building PCs at the $600-$800 price point.

This summer’s computer revolution won’t be Windows XP.

And, in something not really related, here’s something you probably missed, unfortunately. Start rubbing your hands together if you enjoy the Mac-PC or Intel-AMD wars. This is a hard benchmark comparing AMD Athlon, Intel P3, and Motorola PowerPC architectures and their relative speed. The methodology: under Linux, cross-compile a Linux kernel for the SPARC architecture (compiling native isn’t a fair comparison; this way they’re all creating identical code and therefore doing the same work, or as close to it as you’re gonna get). You know those claims that a Mac is twice as fast as an equivalent-speed Pentium III running Photoshop? I always countered that with Microsoft Office benchmarks, where a Mac is about 1/4 the speed of a PC, at best, when doing a mail merge. Neither is a fair test. This benchmark resembles one.

Anyway… Yes, a G4 is faster than the equivalently clocked Pentium III. How much faster? Roughly 10 percent. And an Athlon turns out to be about 20 percent slower than the equivalent P3. Of course, the Athlon reaches clock speeds the P3 never will, and the Athlon is also much more than 20 percent cheaper than the equivalently-clocked P3, so who really cares?

This still isn’t a totally fair comparison of CPU architecture, since chipsets vary (and it’s entirely possible that the difference between the P3 and the Athlon in speed is due to chipset quality), but if indeed the G4 was twice as fast as the P3, it would surely outperform it by better than 10 percent in this test. But it’s a decent comparison of real-world performance, because it doesn’t matter how much better your CPU is if it’s burdened by a chipset that doesn’t show up to play on game day.

Most telling is the end, where he gives the cost per speed unit. AMD wins that chart handily.

Enough of my babble. Read all about it here.

More Like This: AMD Hardware

Plextor bargains, and Year 2000 in review

A bargain Plextor CD-RW. I just spotted this great tip in a link to a link to a link in the StorageReview forums. The Iomega ZipCD 12x10x32 appears to be a relabeled Plextor drive, and it sometimes sells for around $100. So if you’re looking for the best CD-R on the market at a great price, go get it.
Details are at www.roundsparrow.com/comp/iomega1 if you want to have a look-see.

The $99 price seems to be a CompUSA special sale. Check local availability at www.compusa.com/products/product_info.asp?product_code=280095 if you’re interested.

Incidentally, the IDE 12x10x32 drives from TDK and Creative are also reported to be re-branded Plextors. Regular retail price on these four “twin” drives is similar, around $300. The TDK and Creative drives come with Nero Burning ROM, however, making them more desirable than the Plextor model. Iomega bundles Adaptec’s CD suite.

Happy New Year. An ancient Chinese curse says, “May you live in interesting times.” Well, 2000 certainly was interesting. So, my toast to you this year is this: May 2001 be less interesting than 2000. Boring isn’t always bad. Just usually.

Linux 2.4 almost made it. Yesterday, Linus Torvalds released linux2.4-prerelease and vowed there won’t be a prerelease1, prerelease2, etc.–this is it. Bugs get fixed in this one, then the final 2.4 comes out (to be immediately followed by linux2.4ac1, no doubt–Alan Cox always releases a patched kernel swatting a couple of bugs within hours of Linus releasing the new kernel. It happened with 2.0 and with 2.2, and history repeats itself).

Anyway, the 2.2 prerelease turned into a series in spite of Linus’ vows, so Linus isn’t always right, but I expect 2.4 will be out this month, if not this week.

Linux 2.4 will increase performance, especially on high-memory and SMP machines, but I ran a 2.3 series kernel (basically the Linux equivalent of an alpha release of 2.4) on my P120 for a long time and found it to be faster than 2.2, even on a machine that humble. I also found it to be more stable than Microsoft’s final releases, but hey.

I ought to download 2.4prerelease and put it on my dual Celeron box to see how far it’s come, but I doubt I get around to it today.

Other lowlights of 2000. Windows 2000 flopped. It’s not a total disaster, but sales aren’t meeting Microsoft’s expectations. PC sales flopped, and that was a disaster. The Pentium 4 was released to awful reviews. Nvidia bought the mortal remains of 3dfx for a song. Similarly, Aureal departed from this mortal coil, purchased by longtime archrival Creative Labs after bankruptcy. (In a former incarnation, before bankruptcy and being run into the ground, Aureal was known as MediaVision. PC veterans probably remember them.) A federal judge ordered the breakup of Microsoft, but the appeals process promises to at least delay it, if not prevent it. We’ll hear a lot about that in 2001, but 2001 probably won’t bring any closure.

Hmm, other highlights. Apple failed to release OS X this year, and saw its new product line flop. Dotcom after dotcom shuttered its doors, much to Wall Street’s dismay. Linux companies didn’t topple MS, much to Wall Street’s dismay. And speaking of Wall Street, Larry Ellison (Oracle) and Bill Gates (Microsoft) flip-flopped in the rankings of richest man in the world several times.

And two of my favorite pundits, Bob Metcalfe and G. Burgess Alison, called it quits last year. They are sorely missed.

And once again, 2000 wasn’t the year of the NC.

I know I missed a few. But those were the highlights, as I see them.

01/01/2001

Mailbag:

Partition; IDE/SCSI; Lost CD ROM; Optimizing ME; Win 98/ME

A bargain Plextor CD-RW. I just spotted this great tip in a link to a link to a link in the StorageReview forums. The Iomega ZipCD 12x10x32 appears to be a relabeled Plextor drive, and it sometimes sells for around $100. So if you’re looking for the best CD-R on the market at a great price, go get it.

Details are at www.roundsparrow.com/comp/iomega1 if you want to have a look-see.

The $99 price seems to be a CompUSA special sale. Check local availability at www.compusa.com/products/product_info.asp?product_code=280095 if you’re interested.

Incidentally, the IDE 12x10x32 drives from TDK and Creative are also reported to be re-branded Plextors. Regular retail price on these four “twin” drives is similar, around $300. The TDK and Creative drives come with Nero Burning ROM, however, making them more desirable than the Plextor model. Iomega bundles Adaptec’s CD suite.

Happy New Year. An ancient Chinese curse says, “May you live in interesting times.” Well, 2000 certainly was interesting. So, my toast to you this year is this: May 2001 be less interesting than 2000. Boring isn’t always bad. Just usually.

Linux 2.4 almost made it. Yesterday, Linus Torvalds released linux2.4-prerelease and vowed there won’t be a prerelease1, prerelease2, etc.–this is it. Bugs get fixed in this one, then the final 2.4 comes out (to be immediately followed by linux2.4ac1, no doubt–Alan Cox always releases a patched kernel swatting a couple of bugs within hours of Linus releasing the new kernel. It happened with 2.0 and with 2.2, and history repeats itself).

Anyway, the 2.2 prerelease turned into a series in spite of Linus’ vows, so Linus isn’t always right, but I expect 2.4 will be out this month, if not this week.

Linux 2.4 will increase performance, especially on high-memory and SMP machines, but I ran a 2.3 series kernel (basically the Linux equivalent of an alpha release of 2.4) on my P120 for a long time and found it to be faster than 2.2, even on a machine that humble. I also found it to be more stable than Microsoft’s final releases, but hey.

I ought to download 2.4prerelease and put it on my dual Celeron box to see how far it’s come, but I doubt I get around to it today.

Other lowlights of 2000. Windows 2000 flopped. It’s not a total disaster, but sales aren’t meeting Microsoft’s expectations. PC sales flopped, and that was a disaster. The Pentium 4 was released to awful reviews. Nvidia bought the mortal remains of 3dfx for a song. Similarly, Aureal departed from this mortal coil, purchased by longtime archrival Creative Labs after bankruptcy. (In a former incarnation, before bankruptcy and being run into the ground, Aureal was known as MediaVision. PC veterans probably remember them.) A federal judge ordered the breakup of Microsoft, but the appeals process promises to at least delay it, if not prevent it. We’ll hear a lot about that in 2001, but 2001 probably won’t bring any closure.

Hmm, other highlights. Apple failed to release OS X this year, and saw its new product line flop. Dotcom after dotcom shuttered its doors, much to Wall Street’s dismay. Linux companies didn’t topple MS, much to Wall Street’s dismay. And speaking of Wall Street, Larry Ellison (Oracle) and Bill Gates (Microsoft) flip-flopped in the rankings of richest man in the world several times.

And two of my favorite pundits, Bob Metcalfe and G. Burgess Alison, called it quits last year. They are sorely missed.

And once again, 2000 wasn’t the year of the NC.

I know I missed a few. But those were the highlights, as I see them.

Mailbag:

Partition; IDE/SCSI; Lost CD ROM; Optimizing ME; Win 98/ME

10/30/2000

Leading off, some baseball news. Baseball and network execs are puzzled over why this was the lowest-rated World Series ever. (Story here.) Could it be that no one’s interested in watching $200 million worth of spoiled brats from New York throw temper tantrums? Nah, couldn’t be.

Baseball needs a Cinderella story. Bad.

Athlons are dirt cheap. Don’t buy one. Dan Seto noticed and mentioned that AMD Athlons are now cheap as dirt, at least compared to their once-stratospheric levels. He cited a 1 GHz Athlon for $320. So I hopped on the Web, and sure enough, you can easily find one in the $300 range. Some of the bottom-feeder vendors are selling them for as little as $260.

The rest of the lineup? 700/$99, 750/$108, 800/$129, 850/$146, 900/$166, 950/$224.

Remember, though, before you rush out to buy a supercheap gigahertz CPU, that CPU speed is but one factor in performance. Match it up with a video card that treats you right, and with a sound card that isn’t going to suck up all your CPU cycles (the SB Live! MP3+ is an outstanding inexpensive choice), and most importantly, with a hard drive that doesn’t hold you back. If you’re building a performance system, particularly one that’ll be running Linux, NT, or W2K, give serious thought to a SCSI disk. You’ll be happier with a SCSI-equipped 700 MHz system than with an IDE-equipped GHz system.

If money were no object, here’s what I’d get today and why (then I’ll tell you why I still wouldn’t buy it, even if money were no object):

  • Asus A7V mobo — most stable Athlon board available, and every time I buy something other than an Asus I regret it later
  • AMD Thunderbird 1.2 GHz — strictly for braggin’ rights
  • 256 MB Crucial PC133 RAM — Micron memory, the best in the business
  • Adaptec 29160 Ultra160 SCSI PCI host adapter — hey, it’s Adaptec
  • Seagate Cheetah X15 18GB 15K RPM hard drive — Who cares about drive size? This bad boy has a 3.9 ms seek time, a 4-meg buffer and 15,000 rpm spindle speed. It’ll heat my apartment, it’ll wake up my neighbors, but I won’t wait on it (much).
  • Plextor UltraPlex Wide 40X CD-ROM — I love my Plextor drives
  • Plextor 12X CD-R with Burnproof — no coasters with this drive
  • Sound Blaster Live! Platinum — same as the MP3+ but with a nice front-mounted breakout box for my audio gear
  • 3Com 3CR990 NIC — this is the coolest NIC on the market, far and away. It has an onboard processor that handles much of the TCP/IP encapsulation itself, freeing CPU cycles. Same principle as 3D acceleration on your video card and DirectSound acceleration on your sound card. A hundred bucks, but probably worth every cent. Nobody seems to know about it, so I’m telling you.

I wouldn’t worry so much about the video card. My two-year-old STB Velocity 128 frankly is enough card for most of what I do. I suppose I’d get an nVidia GeForce256-based model of some sort. Since the nVidia Riva128 chipset has long since been sent to the gulag, the value chipset is the TNT2. Hot tip if you’re building a value PC: I’m seeing Creative Labs OEM TNT2-based cards for $60, and that’s more than enough card for all but the most die-hard gamer.

Amazingly, you could have this system for well under $3,000. I figured buying the best of everything would run into the $4500 range easily.

I suspect AMD slashed prices precisely because this is a good time to wait and they don’t want you to. Those in the know know that the AMD 760 chipset, which supports DDR SDRAM (basically 266 MHz SDRAM) comes out this week, so anything available today is old hat. This isn’t the multiprocessor AMD 760MP though — we’re looking at January for that. Sorry.

So why not buy now and replace the motherboard later? The 760 introduces a newer, faster front-side bus. If you want to exploit its full potential, you need a new CPU. No one is going to want these old ones now.

I spent a good part of the weekend working on an article. Essentially, I’m distilling chapter 2 of Optimizing Windows into a 3,000-word piece. That’s hard. The tips fit into that, but with very little explanation and very little flair. So much for the difference between it and every other “21 Ways to Speed Up Windows” article, except mine may be more complete for lack of explanation and flair.

Some argue they don’t want flair. They’re lying. Without flair, it reads like an economics textbook. Without explanation, you haven’t done anyone much good.

The line I really don’t want to lose: “I hate screen savers. I hate them so much, when I was once invited to make an appearance on a US television program called The Screen Savers, I turned them down.” Then I go into explaining why screen savers are the cause of everything wrong with the world today.

I was at 3,600 words Saturday, down to about 3,200 by Sunday afternoon. I can cut the two least important tips, leaving 20, and be at 2946, which might leave room for some screenshots. I’m half tempted to ask him if I can do the page layout for this thing as well… That’s not likely, but worth asking.