So you want to upgrade your computer? More memory, a bigger hard drive, maybe a new microprocessor?
The concept is tempting. Besides the old computer you may own, there are an awful lot of old computers floating around these days, available free or at nominal cost. Companies routinely throw out 486s, as well as PCs based on first-generation Pentiums. Surely we should be able to tweak one of these guys into performing acceptably as a second PC for Junior -- especially since all we're doing with it is browsing the Web, one of the least demanding jobs you can give a PC.
As for last year's computer, my K-6 II 333, well, what would it take to turn it into a screamer? Or even build a computer from scratch?
I've been toying with these kinds of projects for the past few weeks. I only ruined one of three systems, which is a pretty good average, actually, and I should have Nos. 4 and 5 done shortly.
The biggest lesson I learned from this little operation? Simple stuff, like cleaning out old software and updating Windows, works the best.
Upgrades that should work may not. And upgrades that seem to make sense from a technical perspective may simply cost too much. It may be better to buy a new computer.
The economics of computer manufacturing vs. personal assembly resemble those of auto manufacturing vs. chop-shop prices: The individual pieces cost two-thirds more than the complete unit. True, this is a deeply discounted, no-name system, but you'll find similar bargains in name brands at auction sites. Makes you wonder whether there's a business in buying PCs and stripping them for parts, eh?
Sure, a 300 MHz system is considered pretty pokey these days, but it makes a perfectly adequate Web terminal. The costs of upgrading to higher speeds are even more suspect, however, and the returns even less certain.
With that little bit of background, let's look at some popular upgrades:
More memory. Almost always a winner. Taking a PC from 16 or 32 megabytes of RAM to 64 or 128 usually gives you a major apparent boost in speed. That's because the computer can hold more data and programs in memory, and does not have to read from the disk drive as frequently. Newer memory chips that use 168-pin DIMMs (dual in-line memory modules) cost about a buck a megabyte. Thus, if your computer uses this style of memory, an upgrade from 32 to 64 megabytes of RAM is cheap and effective.
Increasing RAM from 64 to 128 megabytes isn't so clear cut, however, unless you're running lots of programs simultaneously. Nor is the benefit of upgrading memory clear within older systems. Computers of 1995-1996 vintage used 72 pin SIMMs (single in-line memory modules). Memory manufacturers are phasing this style out, and prices have subsequently risen to more than $2 per megabyte. In addition, SIMMs tend to be configured in ways that force you to throw out all your old memory before you can add new.
Thus, kicking that old P75 up to 64 megabytes of RAM could easily cost you $170. For another $170, you could buy a new, albeit low-powered system, complete. At that point, you have to wonder whether you're better off getting a new computer.
A bigger hard drive. Maybe, if your current drive is stuffed. You can get a 1.7 gigabyte drive for as little at $70, and a nice 10 gig model for about $110. Most computers (check your documentation) make provisions for a second-hard drive, and have all the cables and firmware support built in. Since newer drives tend to be faster than old, you should see a small increase in apparent speed. And if you ever dump the old computer and buy a new one, you should be able to transfer the drive, though you may have to reformat it.
The downside? I've never seen a computer -- apart from systems used by MP3 collectors, hard-core gamers or professionals for graphics, engineering and other high-end apps -- that actually needed a new hard drive, as opposed to a ruthless and thorough housecleaning of unnecessary software and data. In most cases, if all you're doing is writing letters and surfing the Net, you can wipe the hard drive, reinstall Windows, Microsoft Works, a browser and e-mail.
A new microprocessor. There are two ways to go about this. If you bought a computer that was designed to be upgrade-friendly, simply buy a faster version of the original processor. Thus, you can -- assuming you have instructions and are willing to throw some internal switches -- easily and cheaply replace the CPU with one that is up to twice as fast. Alternately, there are more expensive kits that can, in theory, increase the speed of a 75 megahertz system to 400 megahertz.
The problem with processor upgrades: The more likely they are to work, the less performance they tend to deliver. Modest 2x upgrades almost always succeed, but you won't necessarily notice an apparent improvement. On the other hand, you may run into compatibility problems with the typical 3x or 4x.