Hardware

First look: The Proliant DL320

Dave Farquhar Hardware , , , , , , , , , , , ,

I’ve had the opportunity the past two days to work with Compaq’s Proliant DL320, an impossibly thin 1U rack-mount server. All I can say is I’m impressed.
When I was in college, a couple of the nearby pizza joints sold oversized 20″ pizzas. The DL320 reminded me of the boxes these pizzas came in. The resemblance isn’t lost on IBM: In its early ads for a competing product, I remember IBM using an impossibly thin young female model holding a 1U server on a pizza-joint set.

HP announced last week that Compaq’s Proliant series will remain basically unchanged, it will just be re-branded with the HP name. HP had no product comparable to the DL320.

I evaluated the entry-level model. It’s a P3 1.13 GHz with 128 MB RAM, dual Intel 100-megabit NICs, and a single 40-gigabyte 7200-rpm Maxtor/Quantum IDE drive. It’s not a heavy-duty server, but it’s not designed to be. It’s designed for businesses that need to get a lot of CPU power into the smallest possible amount of rack space. And in that regard, the DL320 delivers.

Popping the hood reveals a well-designed layout. The P3 is near the front, with three small fans blowing right over it. Two more fans in the rear of the unit pull air out, and two fans in the power supply keep it cool. The unit has four DIMM sockets (one occupied). There’s room for one additional 3.5″ hard drive, and a single 64-bit PCI slot. Obvious applications for that slot include a gigabit Ethernet adapter or a high-end SCSI host adapter. The machine uses a ServerWorks chipset, augmented by a CMD 649 for UMDA-133 support. Compaq utilizes laptop-style floppy and CD-ROM drives to cram all of this into a 1U space.

The fit and finish is very good. The machine looks and feels solid, not flimsy, which is a bit surprising for a server in this price range. Looks-wise, it brings back memories of the old DEC Prioris line.

The rear of the machine has a fairly spartan number of ports: PS/2 keyboard and mouse, two RJ-45 jacks, VGA, one serial port, and two USB ports. There’s no room for luxuries, and such things as a parallel port are questionable in this type of server anyway.

Upon initial powerup, the DL320 asks a number of questions, including what OS you want to run. Directly supported are Windows NT 4.0, Windows 2000, Novell NetWare, and Linux.

Linux installs quickly and the 2.4.18 kernel directly supports the machine’s EtherExpress Pro/100 NICs, CMD 649 IDE, and ServerWorks chipset. A minimal installation of Debian 3.0 booted in 23 seconds, once the machine finished POST. After compiling and installing a kernel with support for all the hardware not in the DL320 removed, that boot time dropped to 15 seconds. That’s less time than it takes for the machine to POST.

Incidentally, that custom kernel was a scant 681K in size. It was befitting of a server with this kind of footprint.

As configured, the DL320 is more than up to the tasks asked of low-end servers, such as user authentication, DNS and DHCP, and mail, file and print services for small workgroups. It would also make a nice applications server, since the applications only need to load once. It would also be outstanding for clustering. For Web server duty or heavier-duty mail, file and print serving, it would be a good idea to upgrade to one of the higher-end DL320s that includes SCSI.

It’s hard to find fault with the DL320. At $1300 for an IDE configuration, it’s a steal. A SCSI-equipped version will run closer to $1900.

Stand by your SCSI.

Dave Farquhar Hardware , , ,

The Storage Review recently ran a feature on the Seagate Barracuda 36 series, Seagate’s current economy-class SCSI drive. Like many low-end Seagate SCSI drives of the past, it is a converted ATA/IDE design. And Storage Review eats these kinds of units up, because theoretically they provide a nice way to demonstrate the difference between IDE and SCSI.
The result? The SCSI unit was actually slower than its IDE brethren in some of the tests.

The conclusion? SCSI isn’t necessarily faster than IDE.

That’s partially right. Taking the same drive mechanism and replacing the IDE circuitry with SCSI circuitry won’t result in a rockin’-fast drive. SCSI does have more overhead than IDE, so without some other changes, the drive won’t be an impressive performer.

The thing is, people don’t buy expensive SCSI controllers and then put retreaded IDE drives on them. Or at least they shouldn’t. The Barracuda 36 series is intended for people replacing SCSI drives in older equipment. Since the drive will frequently be replacing a five-year-old drive (or older), it doesn’t have to be a screamer. Anything made today will be faster than anything you can find from the mid-90s.

SCSI gives other advantages over IDE. First, with a modern host adapter (don’t call it a controller; you’ll get dirty looks) you can connect 14 devices and only use one interrupt. On today’s crowded PCs that try to be everything to everyone, that can be a real boon. Second, you have far fewer limitations over cable length. Don’t buy an IDE cable longer than 18 inches; you’re just asking for trouble. I know, I know, some of you have 36-inch IDE cables and they work fine. Trust me: Replace it with a shorty, and you’ll get fewer data errors, which means a more reliable system at the very least, and possibly a faster system as well due to fewer retransmissions. With SCSI, you can actually use the top bays in that five-foot-tall megatower you bought. Third, you can get external SCSI devices, in the event that you made the mistake of not buying that five-foot-tall megatower, or if you just like portability. This is less of a factor in these days of Firewire and USB 2.0, but it’s still a nicety you don’t get with IDE. Fourth and most importantly, SCSI devices sharing the same bus can talk at the same time. When you put two IDE drives on the same channel, one drive has to wait for the other to shut up before it can speak its peace. This limits the advantage of having multiple drives. With multiple SCSI drives, you can actually saturate all that bandwidth you paid for.

The fifth advantage of may soon fade: command queuing. SCSI drives don’t have to perform requests in the order received. If you’re constantly accessing two files at once, reading one, then writing to the other, in alternating fashion, the IDE drive will be jumping all over the place. The SCSI drive will figure out how to reorder those requests so it doesn’t have to jump around as much. IBM’s recent Deskstar drives can do command queuing as well, provided the operating system supports that mode of operation. But it’s not a common feature in IDE drives yet. This advantage usually won’t show up in benchmarks, but it’s significant. SCSI drives, to use a popular middle-management buzzword, work smarter. If you’ve got a Windows 2000 or XP system with a SCSI drive in it, try using the system while defragmenting the drive. The system will be slower, but not unusable. That’s never true of an IDE drive.

And the sixth advantage of SCSI doesn’t really have much to do with SCSI. With SCSI, you get cutting-edge technologies first. You can’t buy a 15K RPM IDE drive. You can’t even buy a 10K RPM IDE drive. There’s only one IDE drive on the market with an 8-meg cache on it. Caches that size are commonplace on contemporary SCSI drives, and the gargantuan Seagate Barracuda 180 has a 16-meg cache. It also costs as much as a nice computer all by itself, so it’s not exactly a consumer-class drive, but it’s available if you’ve got more money than patience.

Benchmarks are deceiving. Some changes will double the benchmark scores, but a user won’t tell much difference. Other changes barely register, but the user notices them. SCSI is one of those, especially if you multitask a lot.

It’s true that there’s no point in spending $400-$500 for a disk subsystem in a PC you use for word processing and e-mail. You’ll notice a difference, but it’s not worth the extra cost. Although if you’re buying a used system and have a choice between a system with IDE disks and SCSI disks, you should get the SCSI system, even if it means ponying up another 50 bucks. You’ll thank yourself for it.

As for me, I love my SCSI systems with 10K RPM drives in them. They’re wicked fast, and no louder than the IDE drives of four or five years ago. (I don’t have a current IDE drive to compare them to.) I can let my e-mail inbox fill up with thousands of messages without it dragging beyond belief, and my non-Adobe applications load in less than three seconds. Most of them load in less than a second. The drives themselves are small and expensive, but you’re buying performance, not capacity. I can’t fill up a 9-gig drive with applications anyway. Neither can most people.

So no, SCSI isn’t a magic silver bullet. But that doesn’t mean it’s not worth having.

News analysis

Dave Farquhar Hardware , , , , , , , ,

Short takes. Yesterday was a newsworthy day in technology, and I’m sure there’s going to be a ton of misinformation about it eminating from both coasts, so we might as well set the record straight.
Poor quality control drives IBM from the hard drive business! Yeah, whatever. IBM makes one questionable model (and many GXP failures sounded more like power supply failures than hard drive failures), and suddenly everything they’ve ever made is crap. Guess what? Seven years ago you couldn’t give me a Seagate drive, because the drives they were making back then were so slow and unreliable. Maxtors were worse–and my boss at the time, who has a very long memory, nearly disciplined me a couple of years ago for specifying a Maxtor drive in an upgrade. But he’s a reasonable man and saw that the drive held up and performed well. Western Digital has been so hit and miss I still don’t want to buy any of their drives. Though their drives started to look better after they licensed some technology from… Old Big Black and Blue.

And the truth about GXPs: Regardless of how true the quality control allegations are, the drives themselves are the most innovative and advanced IDE devices ever commercially marketed. The platters are made using different materials and processes than conventional discs, which was supposed to make them more reliable. Expect that technology to come of age in a generation or two. The drives even include SCSI-like command queueing (the newest version of Linux’s hdparm allows you to turn this feature on; I have no idea if Windows switches it on by default). The successor to the 60GXP is going to be worth a second and a third look.

Wanna know what’s really going on? Hard drives aren’t very profitable. IBM has a history of spinning off questionable divisions to see if they can survive as smaller, more independent entities. The most famous recent example of this is Lexmark. That’s what’s going on here. IBM and Hitachi spin off and merge their storage divisions, and each company takes a stake in it. If the company mops up the floor with the competition, IBM and Hitachi make lots of money. If the company continues to bleed cash, IBM and Hitachi get nice tax write-offs. Either way, the shareholders are happy.

A number of years ago, IBM was a large producer of memory chips as well. In fact, you can open up a Mac manufactured in the mid-1990s, and chances are you’ll find an IBM-manufactured PowerPC CPU, one or more IBM-manufactured DIMMs, and an IBM SCSI hard drive. Making memory had its ups and downs, and during one of the many downturns in the 90s, IBM got out of the business. There was a time when Intel and AMD were in that business too (I have some old AMD DRAM chips on an expansion card somewhere, and I’ve seen Intel DRAMs but I don’t know if I’ve ever owned any).

This news is a little bit surprising, but hardly shocking. IBM’s making tons of money selling software and services, they’re not making money selling hard drives, and they’ve got a new CEO and nervous investors. This is a way for them to hedge their bets.

And you can expect them to possibly start getting more aggressive about marketing their technologies to other drive manufacturers as well now. Seagate, Maxtor, Western Digital, Fujitsu and Samsung have just changed from competitors into potential customers. Expect disk performance to increase and price to continue to decrease as a result.

How to gauge hard drive reliability. This isn’t exactly news but it seems very relevant. Professional writers don’t see a lot of drives. They can recommend based on their own experience, but their recent experience is going to be limited to a few dozen drives. Message boards are very hit and miss. You have no way of knowing whether it’s a book author hiding behind that handle or a clueless 12-year-old kid. Find an experienced technician who’s still practicing as a technician (I’m not a very good example; at this stage of my career I no longer deal with large numbers of desktop systems–I deal with a handful of servers and my own desktop machine and that’s it) and ask what hard drives they’ve seen fail. When I was doing desktop support regularly, I could tell you almost the exact number of drives I’d seen fail in the past year, and I could tell you the brands. I’d prefer to talk to someone who fixes computers for a large company rather than a computer store tech (since his employer is in the business of selling things, he’s under pressure to recommend what’s in stock), but I’ll still trust a computer store tech over some anonymous user on Usenet or a message board, as well as over a published author. Myself included.

AMD withdraws from the consumer market! AMD mentioned in a conference call yesterday that it plans to discontinue the Duron processor line this year. It makes sense. Fab 25 in Austin is being re-tooled to make flash memory, leaving the Duron without a home. But beyond that, AMD’s new 64-bit Hammer chip is going to hit the market later this year. So they can sell a slightly crippled K7 core as their low-end chip, or they can make their high-end K7 core into the low-end chip and sell the Hammer as a high-end chip. This strategy makes more sense. Clock for clock, the Athlon is still a better chip than the P4. Hammer scales better and performs better. So AMD can pit the Athlon against the Celeron and give P4 performance at a Celeron price, and the Hammer against the P4, which will give P4 clock rates and deliver better performance for 32-bit apps, along with a 64-bit future. There’s not much room in that strategy for the Duron. AMD would rather cede the $35 CPU business to VIA.

Look for the Hammer to gain widespread use in the Linux server market, especially among smaller companies. The Athlon already has an audience there (in spite of some pundits calling AMD-based systems “toys,” you see far more ads for AMD-based servers in Linux Journal than you see for Intel boxes), but the Hammer will become the poor man’s Alpha.

Dell and Gateway upgrade caveats

Dave Farquhar Hardware , ,

I sent this message to Mike Magee of The Inquirer this morning:
Hi Mike,

I’m a freelance author, with one book published by O’Reilly to my credit and a few appearances in Computer Shopper UK.

I visited the Scott Mueller link you referenced at http://www.theinquirer.net/15040206.htm, and just to alert you, I’m not certain that Scott Mueller’s dates on the Dell systems are correct. In late 1998, I attempted to upgrade a Dell P133-based system with an AOpen AX59Pro motherboard, in order to get around the nasty memory limitations in Intel’s 430VX chipset. I knew the motherboard worked because I pulled it out of another working system. The board didn’t work in the Dell. Then, when I reinstalled it in the system I pulled it from, it didn’t work there either.

Fortunately I didn’t kill the power supply so I was able to get the system up and running again by replacing the factory board.

This leads me to believe that Dell has engaged in the practice of nonstandard wiring since 1996.

My recommendation to my readers has always been to replace the power supply when replacing a motherboard in a Dell, since standard ATX power supplies easily bolt into the Dell cases. Any brand-name power supply purchased at retail (Sparkle, Antec, Enermax, etc.) is likely to be of higher quality than the stock Dell power supply anyway, but that’s an additional upgrade expense people may not consider.

I suspect the reason this hasn’t been more widely known is that Dell mostly sells to corporations and has only recently gone after the consumer market in aggressive fashion, and corporations rarely replace motherboards. The labor involved in making the swap, then reinstalling the operating system and applications, costs too much. There’s less labor involved in replacing the system, and then you have a system covered under warranty.

Incidentally, while Gateway does use the standard ATX pinout, many Gateway cases use an odd-shaped power supply. So while an aftermarket power supply will function electrically, it’ll take some cutting and drilling on the case to allow you to bolt it in. Most people will prefer to just buy a new case if the power supply in their Gateway dies–and the power supply is usually the first component to go in a Gateway, in my experience.

How to get my job

Dave Farquhar Hardware , , , , , ,

I’ve had a couple of people ask me in the past couple of weeks how to break into the computer field. It was a tough question. I literally got into fixing these things because I couldn’t find a repair shop in St. Louis that I felt I could trust. So I started trying to fix them myself. I might break it beyond repair, but one time we had a repair done that cost more than replacing the unit outright would have cost. So what did I have to lose, right?
I took my Commodore 128 apart a few times. Usually it was for an upgrade, but once it was to clean the keyboard because keys weren’t working anymore. It was an adventure, and I had to learn how to solder first. My dad’s friend Norb taught me how. He was a building inspector. No wonder I still solder like a plumber, even to this day. So I de-soldered the 6 connections I had to in order to get into the keyboard, removed the dozens of tiny screws, cleaned up the printed circuit board, put it back together, re-soldered those connections, reassembled the computer, and held my breath. It worked. Cool. It didn’t impress the girls, but it saved me at least 50 bucks.

It was my uncle’s approach. I remember riding with him to an auto parts store once, then watching him work on his truck. “I don’t know what I’m doing when it comes to cars,” I said.

“I don’t let that stop me,” he said. “I just have to do it.”

His truck cost more than any computer I’ve ever owned.

Later on, I got an upgrade ROM for my Amiga 2000. So my dad came home one day to find me hovering above my Amiga, which was sprawled across his OMT table. The cover was off, the power supply was out, and the drive cage was out, and there I was, slowly prying out a chip with a screwdriver. Dad gave me a nervous look. “You gonna be able to get that thing back together?” he asked me. “Sure,” I said. I didn’t tell him how many times I’d had it apart before. So he stood there and watched me as I finished extracting the chip, popped the new one into place, and re-installed the power supply and drive cage.

Eventually I got smart and realized I shouldn’t be experimenting on computers that I cared about. XT clones cost about 20 bucks when people wouldn’t just give them to you, so I got a couple. I ripped them apart, figured out how a computer was really put together, and reassembled them. And yes, I even took parts from one and put them in the other to see what would happen. I was pretty sure it would work. It did. Eventually I did something stupid (I don’t remember what anymore) and I killed at least one of those XT clones, but it wasn’t important. I’d learned a lot from them, and I was only out 20 bucks. That’s assuming I wasn’t given the thing outright–I don’t remember that detail anymore either.

I needed that skill the next year. I was living in a fraternity house, and the power supply died in the house computer. I knew enough by then to diagnose it, and I headed off to the local computer shop for parts. They didn’t have any power supplies that would fit, and the motherboard was nonstandard. But they had a lineup of barebones systems sprawled across the floor. A bare 386 cost about $200. I knew the rest of the system worked. So I talked it over with the treasurer, then came back with a house check and bought a 25 MHz 386DX. I took it home, popped the case on the house computer, pulled out the video card and all the I/O cards, installed them in the 386, and found the computer wouldn’t recognize the hard drive. We eventually worked through that one (it turned out we had one of the very few 8-bit IDE drives ever made, and that 8-bit controller did not get along with our 386 one bit) and we got a working system up and going.

By the time I graduated there were at least half a dozen guys in that house capable of doing that job. Times changed (swapping a motherboard was much more of an endeavor in 1993 than it was in 1997, because by then so many components that had once been discrete and configured by jumpers were integrated and configurable through the BIOS Setup), and I’d like to think most of them learned at least a little something from me.

That summer, I got a job selling computers. An opportunity arose when the store technician developed a difficulty showing up for work. They never fired the guy, but since he was only there half the time, I got to be the tech the other half. When he was there, I learned a lot from him.

The next school year, I got wind of a job opportunity on campus. The journalism department had a batch of 300 new IBM PC 330s and 350s. Every last one of them needed to be unboxed and upgraded with extra memory and a NIC, then plugged into the network, where one of the more experienced techs could do a push install of OS/2. I got the job, and I learned a ton from those guys. These are guys who had seen prototypes of the IBM PS/2 Model 80, and who occasionally had to whip out a soldering gun and make a change to the motherboard with an engineer from IBM on the phone. You bet they had a lot to teach me.

That part-time job eventually grew into a full-time job when those guys recognized that I was willing to work hard and willing and able to learn.

That approach worked really well for me. But I had the advantage of being young and being able to wait for opportunities and take them as they came. I also had the advantage of growing up with the things (the schools I went to had computers and taught computer classes, all the way back to when I was in the second grade) and messing with them for the majority of my life.

Realistically, I don’t think that approach would work for an adult with minimal computer skills and a family to support. Or at least it wouldn’t work on a quick timeframe. I’ve tried to teach 24-year-olds starting from ground zero how to do this. It didn’t work very well.

It’s a lot easier to teach someone how to write.

Cult of Linux, and cheap cheap hardware

Dave Farquhar Hardware , , ,

Potpourri. I just have two little things to write about today, so that’s what I’ll do.
The Cult of Linux, and cults of computing past. In yesterday’s comments, Dave Thorarinsson brought up an interesting phenomenon. He observed, when talking about his new Linux box and losing track of time learning it, “It pretty much feels like the time I had my Amiga.”

I remember reading in the mail section of Amiga magazines, more than a decade ago, “I haven’t had this much fun since I got my C-64!” And the old Commodore magazines noted that their C-64s had a special place in their devotees’ hearts and asked, “Have you ever heard of anyone getting attached to a PC clone?”

The inferred answer is no. And that doesn’t seem to have changed. Today, the biggest PC enthusiasts replace their machines frequently, relegating their old, unbeloved machines to grunt roles, or scrapping them for parts.

The C-64, by contrast, was a simple machine. Although it only had one slot for expansion, the motherboard itself was simple enough that just about anyone with a little bit of patience could trace it out and understand it. There was some unused address space in it that you could add chips to. Common projects included speech synthesizers and an extra sound chip, so you could have six-voice, stereo sound. And it seemed like most C-64 owners had tried their hand writing at least simple programs.

The Amiga was similar. It was harder to program, and a little harder to hack, but I had a subdirectory on my Amiga’s hard drive that was called “PD Hardware,” containing makeshift schematics for stuff I could build. I wired in a PAL/NTSC switch so I could change video standards and run European software correctly. I even designed an add-on board for it myself, once, to give myself an extra ROM socket. And of course I replaced much of the operating system with replacement utilities written by hobbyists that were smaller and faster, had more features, or both.

You want to know why the Amiga and C-64 fanatics were so loyal? They knew their machines inside and out, and to a degree that just wasn’t possible with a PC clone.

With Linux, the hardware doesn’t matter anymore, but you can customize the software to whatever extent you want. The hardest-core Linux people are doing just that. At a level below that, people compile the OS from source, from scratch. At a level below that, they just replace utilities with smaller and faster or more functional ones, like I was doing with my Amiga.

The long and the short of it is tweakers have a lot more liberties with Linux than they’ll ever get out of Redmond.

People who liked my book can certainly find a lot to like with Linux. People who fancy themselves experts but don’t really want to know how their computers work and who think books like mine are a waste of time will never be tied to Linux like the fanatics are.

Cheap hardware. It’s a buyer’s market. I’m building an intranet server for my church, and they have lots of bits and pieces but not quite enough for a complete system, so I did some looking around at my favorite bottom-fishing holes. You can get an awful lot of computer for next to nothing these days.

I’ve bought things at least once from each of these vendors, and they got stuff to me reasonably fast (within a week, ground shipping) so I feel reasonably comfortable recommending them.

Slot-1 barebones systems — case, power supply, slot 1 mobo (66 MHz FSB, so you’re stuck with P2s up to 333 MHz or Celerons, and according to the manufacturer, Celerons will work), and a floppy drive. $24.50. $19.95 if you don’t mind some wear. Very nice.

At the same place, scroll down and you can pick up Tier-1 business-class P2s for under $200. If you’re lucky you can sometimes even find a dual-capable machine. They go fast, but the getting’s good. Lots of businesses are dumping these due to their 3- or 4-year upgrade cycles. For someone who just wants to do word processing and e-mail, these systems are overkill. If you want to upgrade in pieces over the course of the next couple of years, the P2s with a 100 MHz FSB are workhorses and you can add lots of cheap memory, nice video cards, and fast, cheap hard drives. Gatermann just picked up an HP Vectra P2-266, dual capable, with SCSI, for $117. It will serve him well. He plans to run Debian on it, but I don’t know if he’s thinking of it as a server or a workstation. It’d make a fabulous server.

72-pin SIMMs — if you’ve still got a system that takes them, nice deals on 4-16MB sticks, and good deals on bigger sticks.

Low-end Pentium I desktops — P75-120, 16 meg of RAM, 1 MB video, who knows what else. $29.99. Nice for a low-end Linux box, or for a Win95 system that’s going to see limited use. Put a fast hard drive in it and it’ll surprise you. After you get yours, check and see what CPUs the system will take; a P200 can be had for as little as $16 and makes a nice upgrade. It’s a pretty big step up from 75 MHz to 200 MHz.

Seagate 9-gig 5.25″ full-height SCSI HD — $12.95. If you’ve got two 5.25″ bays open and no spacer between them, here’s a cheap way to fill it with 9 gigs of storage. I know a couple of people who have these drives. They’re surprisingly quick (they hold their own next to 7200-rpm 3.5″ disks). Back up your data and buy some spares if the shipping doesn’t kill you. One of the guys I know has several of these, and he’s had one or two die on him.

FIC AZ-11 ATX mobo — $34.95. It’ll take Durons and Athlon Thunderbirds. The fastest chip I’ve put in one is a 1 GHz Duron. No ISA slots and just 5 PCI slots, but it’s a capable board. I use one in my video editing workstation, and I paid more than 35 bucks for it. Totally obsolete, but when you can get a 1 GHz Duron and fan on this board for $89, who cares? It’ll still be a good computer in three years. Trust me.

I keep seeing this board on the closeout places, so I expect even after the current supply dries up, there’ll be more.

Speaking of closeout motherboards, there’s a variety of them over at Just Deals. You can get a Soyo Socket 370 board for $28 and various Socket A boards in the $35-$40 range. If it’s Slot 1 or even Slot A you’re looking for (maybe you’ve got a CPU laying around), you can find stuff there too.

And if you need a cheap copy of Word 97, you can get the Works 99 suite for 30 bucks.

Need an operating system for that new old computer? Prices range from $25 for Win95 to $180 for WinXP Pro.

Heat compound and reliability

Dave Farquhar Hardware , ,

It’s known by many names. Heat sink grease. Thermal compound. Heat sink compound.

It comes in tubes or syringes. Cost varies from a buck or two to twenty. The cheap stuff is plain old grease. The expensive stuff is made of exotic materials including silver. Read more

Building an inexpensive PC

Dave Farquhar Hardware , , , , ,

Building an inexpensive PC. An old out-of-town friend I don’t hear from often called the other day. He wants to buy a computer and dabble in audio production. Some local guy quoted him $2,500 to build a system. He read me the specs, and all I can say is this guy had better be using Lian-Li cases and PC Power and Cooling power supplies (or I guess I’d settle for high-end Enermax), but I doubt it. I do know he’s using a top-end Athlon XP processor and an Abit motherboard, but he wasn’t pairing it with DDR, so he was totally killing the chip’s performance anyway. For two and a half grand, you’d better be getting DDR, and lots of it.
“You need a 32-meg video card because when the computer is drawing the waveforms, it has to be dead-on. You can’t afford for it to lag,” he said.

I got news for this idiot. When it comes to drawing simple line graphics like a waveform, the ancient ET4000 chipset in my 486 will have no problem keeping up with it. Even if you use a fill to make the waveform look pretty. And that video “card” (it was integrated into my motherboard) had 512K (K, as in kilobytes) of memory. Although anyone who wasn’t born yesterday knows that the amount of memory on a video card has nothing to do with its speed, outside of the realm of 3D gaming. Knowing kids these days, some of them may even know that at birth.

In other words, the guy’s a moron. Either he knows nothing about computers, or he knows how to skimp but he’s not a convincing salesman.

I know for a fact that audio editing doesn’t need a supercomputer. If I can do video editing on a 700 MHz Duron, I know a Duron CPU, paired with a decent supporting cast, is going to be adequate for multitrack audio recording and editing as well.

I asked him how much he could spend. He told me $800, not counting a monitor and the editing card/package. I squirmed. I spent way too much time shopping around. Here’s what I came up with (not counting the operating system):

1 GHz AMD Duron
FIC AZ11 motherboard (on closeout, so it was cheap)
ATI Xpert 2000 Pro AGP video card (with a blazing 32 megs–ahem)
Maxtor D740 20-gig 7200 RPM IDE hard drive
Maxtor D740 60-gig 7200 RPM IDE hard drive
512 MB Crucial PC133 SDRAM
Mitsumi 3.5″ floppy drive
Sony 52X ATAPI CD-ROM
Plextor Plexwriter 12/10/32A CD-RW
Enermax A1QX-6 mid-tower case with Enermax 300W power supply
US Robotics 2977 controller-based PCI modem
Closeout Dell-branded Logitech mouse and Dell-branded keyboard

I told him there are two brands of CD-RW I trust, especially for audio work: re-labeled Plextor, and Plextor. In all honesty, I would have much prefered to build an all-SCSI system, but for this kind of budget, that’s impossible. All-SCSI would have given much better disk performance, and it would have given access to the Plextor UltraPlex 40max CD-ROM, which is the only drive I trust for extracting digital audio. I imagine he’ll be doing a little of that. The Sony drive will do a decent job, but I’ve seen the Plextor work miracles. But the Plextor is $100, while the Sony cost around $25. I’ll definitely take a Sony over a Cyberdrive or Lite-On (which probably would have run $19).

I couldn’t get PC Power and Cooling on this budget. The price on the Enermax combo was good (less than a PCP&C 300W power supply alone) and the quality is respectable. The Japanese steel is a little lighter gauge than I prefer, but I didn’t cut myself on it. The fit is good, and it’s a good-looking case. Not show-off good like Lian-Li, but better-looking than most of the stuff in its price range. The cobalt blue trim compliments the lettering on the Plextor drive.

Finding a place to put the hard drives is a bit of a challenge. Modern 7200-rpm drives don’t run very hot, but I still don’t want them running directly above one another. I finally settled on putting a drive in the lowest 3.5″ bay and the other in the lowest 5.25″ bay.

The USR 2977 is the secret weapon here. A $20 no-name Winmodem would be a royal pain to set up, and chew up lots of CPU cycles. The 2977 was under $50 and won’t be a load on the system. That’s a speed trick I’m sure that local guy doesn’t know.

The 1 GHz Duron is still overkill, but that’s the slowest chip I could talk him into. I was starting to get annoyed with him. I don’t just know about computer speed, I literally wrote the book on computer speed, and my friend didn’t know what I was talking about when I said something about a boot floppy. And this year’s hot chip is next year’s budget chip, so if the budget chip is enough to get the job done this year, you can go buy more CPU next year. Besides, there was no way to cram any more CPU power into this tiny budget, other than sacrificing disk speed, which is more important unless he’s running Windows XP, which he won’t be. (I’ll drive 200 miles and take his computer away from him if he does.)

As for the two drives, any time you do multimedia work, you want to make sure your application and swap file are located on one drive, and the audio you’re working with is on a second drive. I probably could have gotten by with a 5400-rpm drive to hold the OS, but there isn’t much price difference between a 5400 RPM 20-gig drive and a 7200.

As for how the system runs, I’m sure it’ll smoke. The motherboard isn’t here yet. In all fairness, I ordered it Monday and it was shipped UPS Ground from California on Tuesday.

I ordered the motherboard from Just Deals and the memory came from Crucial. The rest of the stuff came from Directron and New Egg, who as always gave me great prices and fast delivery.

Linux and PC cubes

Dave Farquhar Hardware , , , , , ,

PC cubes! Yes, I want a cube-shaped computer, because it’s small. No, I don’t want one made by Apple, or an obsolete NeXT (I used those in college when I couldn’t get time on an SGI). I want something small and cheap, and if it’s reasonably good looking, that’s a bonus.
Enter the Shuttle SV24.

Unlike Apple’s cube, it has a brushed-alumninum case, so it won’t crack. Just like Apple’s cube, it generates extreme reactions, and not everyone who likes Apple’s cube likes Shuttle’s.

I admit, it doesn’t have Apple’s styling. But I like Lian-Li’s styling a lot better. I wouldn’t put this in Lian-Li’s league either. But it’s certainly no uglier than any of the PCs I own now, and it’s small and light. So yeah, it has me thinking.

Where can you get one? Two of my favorite vendors have it, at a price of $250: Newegg.com and Mwave.com.

I also saw on Ars’ forums that MSI makes a slimline PC called the 6215. Newegg has it (search for “6215”) for $210. It’s tiny, but has two PCI slots and is more conventional-looking. I’m thinking the 6215 would be great for a server appliance, seeing as it has two PCI slots so you could put a SCSI card in it. You could also disable the onboard Realtek NIC and replace it with a card like an Intel EtherExpress Pro that uses less CPU time.

More Linux. The biggest thing holding me back from migrating to SupaSite is its requirement of the Apache, MySQL and PHP trio. I’ve tried to get those three to work together before, and the setup wasn’t exactly trivial, especially when trying to do it from RPMs. It looks like it’d be a whole lot easier to just compile it yourself. But this past week I found Apache Toolbox, which downloads the source for those three, plus bunches of Apache modules and compiles them for you. It sounds like it even helps out with configuration. I’ve gotta give this one a shot.

It’s October…

Dave Farquhar Hardware , ,

October…
When the trees are stripped bare
Of all they wear
Do I care?
October…
When kingdoms rise
And kingdoms fall
–U2
I didn’t do a whole lot this weekend. I laid around a lot, I did some dishes, and Saturday night I went out with some friends. More on that later. I can’t tell the story properly right now.

Linux as a diagnostic. I remain convinced that compiling Linux is the best system-wide diagnostic in existance. Case point: I lost a drive in a Windows 2000 box a while back. I gave up on trying to get the data back; all I cared about was my Baseball Mogul stats, but I started another game, built up another dynasty, so I don’t care about it anymore. I reformatted the drive and put Sorcerer Linux on it. First things first, an all-SCSI Linux box with a fast CPU really rocks. The most time-consuming part of the boot time is bringing up the SCSI interface. That takes about 15 seconds. The rest of the process is literally instantaneous.

Well, there’s no point in having a great system without recompiling everything specifically for it to take maximum advantage of it, right? So I started recompiling. The controversial 2.4.10 kernel came down and compiled without a hitch, and yes, the system does run very nicely with it. The simpler packages that provide most of the standard Unix utilities came down and compiled quickly and easily. Then when it came time to recompile the monstrosity that is glibc (the key library of any Unix system, and it’s a 16-meg bzipped tarball–this thing’s huge), the system’s weaknesses showed up. The drive failed again. I got sector errors and the system crashed hard. I reset and tried again. It came back up, Reiserfs quickly fixed everything, and it looked good, so I recompiled. This time, I reached the end of the compile process, but when it came time to copy the files into place, files that are there stopped being there. The drive failed again.

So, I’ve either got a heat problem or a power problem. The drive’s kinda crammed in a spot where it doesn’t get much airflow, and I’ve got a PCP&C power supply, so I suspect it’s a heat problem.

Nothing stress-tests PC components like compiling an entire operating system. Besides, even under regular use Linux tends to push hardware harder than Windows, even Windows 2000, but I see that as a good thing. I paid for the hardware, so I want my OS to squeeze it for every ounce it’s worth.