What to look for when deciding what computer model to buy? a system
What people are often looking for when buying a computer is the fastest computer (MHz) they can find, in the hopes that this means that their system will be obsoleted less quickly. Unfortunately speed is misleading, different machines running at different speeds have different levels of performance. But even if it weren't, performance alone is not the best indicator of the most value.
Buying a computer system is about cost tradeoffs. Putting too much money in buying the fastest box leaves less for the peripherals that will make their computer really useful to them. You should balance the entire system costs and not just the main box. I'd gladly move down to a slower computer if it meant I could do more. In fact, I have. I usually use a laptop, which is slower than a desktop, because I care more about usability, convenience and portability than just pure speed.
Computers are like cars: faster is not automatically better. A Ferrari may make a worse commuter car than a nice Honda, it will definitely be a lot less cost efficient, and you would do much better to buy a pickup truck for hauling manure. Computers are complex systems and looking at one feature while ignoring the rest is not an intelligent way to buy a car or computer. Figure out what you need to do, and focus on getting that done.
Where should you buy in the product line? At any given time there are low-end, high-end and midrange computer systems.
Many sales people will always try to start you at the top. Like that little old lady from Pasadena who trades in her Dodge Viper. One has to wonder, "who the hell sold her that car, and what did she think she needed it for?" Most sales people are overselling you on the box.
You probably do not want to buy the top-end because you pay for a premium for that performance edge, and get diminishing returns on investment for each dollar invested. And the low-end machines have a lower life span and may not run the top end stuff fast enough. So there is a sweet spot in the middle. Midrange machines almost always offer the best value, with a combination of flexibility and performance, as well as have a reasonable lifespan.
If you are going to be using your machine professionally or for many hours a day, then you should be starting near the top end (fastest) computers. The same if you keep machines longer; the less often you replace machines, the more you can afford to buy-up the line. The machine costs, amortized over the years of use (lifespan), will be nominal compared to the cost of your time, so buying up pays off.
Gamers claim they need the fastest, but strategy, role playing, 2D, board games, and so on, are on not really performance demanding at all. It is only the 3D, virtual world and Flight-Simulator games that really require speed. But almost all those types are still playable at mid end or lower end. And gamers are the types that replace their machines often, so they are going to stay in the sweet spot and would save more by spending less on the CPU. So most gamers buy higher than they need to for ego and status as much as requirements. Everything else is a rationalization.
If you are a newbie (just starting out), and you want to learn computers, but you are going to use it mainly for word-processing, web-browsing and email, then you should start thinking towards the bottom of the product lines. All machines are overpowered for most people's basic needs. If it takes you 2 or 3 years to outgrow the current machine, then by then you'll be able to buy a new machine that will be far faster, at less cost, than you would have 2 or 3 years earlier. So you can afford to sell your machine and trade up when know better what you actually need. Also if you're likely to replace machines often, then buy down the line because you'll be replacing it with something faster soon anyways.
In general processors are so much faster than people need that I tell them not to focus on speed. I recommend that people buy less processing power (speed) if they will spend that extra money on other things that will improve the usefulness of the machine. Things like a better display, wireless network, or better peripherals will probably make a better purchase overall. Remember you're looking for a system (entire solution), not just the hottest box on the planet; so figure out your budget, and find the best system, with all the things you want, that fits in that budget.
For example, you're going to be on the machine a lot; so the keyboard and display are important. Too many people hardly think about them; but consider moving down in pure processing speed or hard drive, if it means you'll get a larger display or clearer display that you'll be staring at for years. People often keep displays for a few machines, so there's a much longer payoff on them, and some other peripherals than the machine itself. Some jobs require seeing more, which means a larger display will require you to scroll less, which will make you far more productive than just having a faster processor. For others, they need portability and don't care about the display size; they can think smaller or even about laptops.
Portability can cost. Laptops cost more for a given performance, but they add convenience that returns that value. I often recommend people consider portables if they want to use their computer in many places. Laptops are incredibly convenient for people who are on their machines a ton, or who want their information always with them. Know you're paying for that convenience, but it can still be a good value.
Laptops aren't as good in ergonomics as a good desktop; so for full time or professional usage, desktops are easier to have setup well (comfortable placement and so on). Or at least buy full sized monitors and keyboards that you plug into when at your workspace. I also recommend that laptop users buy a few power chargers, one for work, home office, couch, etc., and just leave them there; then they don't have to move them around and carry them with them as much; an easy way to increase the convenience.
If you have a portable, and you want to use it multiple places in your house, then you definitely want to think of a wireless network (and factor in the costs). Wireless networks are often cheaper and easier than trying to run wires into a few rooms of the house, if you're going to have more than one machine. And in fact, sometimes a couple machines are better than one. Decide what you need first.
Don't skimp on the RAM. Prices are falling quickly so you don't want to over-buy; but a memory starved system is much slower than it needs to be. Drive space is cheaper and easier to add later, and most people have far more than they need. So unless you're doing video or something, put more into RAM and you'll probably be fine with the off-the-shelf HD.
Digital Cameras are critical for many people; they allow real estate agents to shoot homes, people to send photos, and others to have fun laying out newsletters or doing websites. The computer is valuable; but the computer with a digital camera is a solution. If you like picture, you'll love digital pictures. The same if you're doing video; when you look at systems you need to consider integrating video cameras, DVD players, and software to do what you want. Again, think solutions.
Printers are important. How are you going to display what you create? Do you need low capacity ink-jet for printing photos, or a high capacity laser printer for those high volume technical documents or reports that you generate in your home business? Don't over-buy, because prices drop so fast. Do you need an integrated scanner and copier (All-in-one), or can you get away without one for now. And remember, for $.50 a month in copies, or just running them off at work, it is probably not a good justification to spend $100 more on a printer if you don't need it. On the other hand, if you're doing a ton of page layout and want to scan articles or photos, then you might want a dedicated scanner, and not something on your printer. Get what you need now, with a eye towards the next couple years at most; but don't go crazy. And you need to think of solutions, not just the computer box.
Storage is unimportant; hard drives are cheap and you can add them later for little cost, and most systems have more than the users need. Most drives that are filled just need some cleanup, organization or archiving (throwing unused stuff in some backup and then cleaning it off). Unless you know you need the space (like you are doing video production, or have a thousand CD collection that you want to convert to MP3's, etc.), then don't worry about it.
While drive space is seldom an issue, having something to copy your data to is very important. Get a machine with a writable CD or DVD and don't forget to backup semi regularly. How often you use your machine, and how much you can afford to lose, will dictate how often; but once a month for occasional users, and once a week for regular ones is probably a good rule of thumb. Everyone will have a hard-drive die at some time; so decide how much you can afford to lose, and then backup accordingly. But definitely think of getting something that will allow you to archive and backup so you don't lose it all; and factor it into the system price.
Don't forget the software. Too many people will buy a computer, and then suffer only with the software that comes with the machine. Make sure you budget in what you need to really enjoy and use your machine.
One thing that is a less important option, for most people, is upgraded speakers. If you love music, and are going to listen to MP3's all day, or you're a home musician and going to mix your own, or you love playing games at full volume, then have fun with it. But for most everyone else, the cheap speakers are more than enough. Plus they're easy and inexpensive to add later anyways.
Remember, in most people's systems the actual computer (the box) should be under half the cost of the entire system. It's the other stuff, the RAM, Monitors, Printers, Modems, and software that cost the rest. You can upgrade the box later, without having to update all the parts, and the parts usually have a longer life span than the box, and are usually dropping in cost less quickly than the machine itself; so you see a better return on investment by putting more money there. And you can always replace the box later for fairly cheap, if everything else is in a good state. If you have to replace everything at once, well then you're screwed.
Buy what you need now. Waiting is a way of wasting your own potential and stalling your own learning. If you need it, budget for it. If you don't, then don't get it.
The trick to buying a computer is not the highest Mhz, even if that was the fastest machine (and it isn't). You need to think systems and solutions. Ask yourself, "How can the computer become more useful for me?" Think about all the different categories and uses, and what you really need. Try to meet those needs. Speed is nice, and there are minimum requirements for certain tasks; but don't get lured to the cutting edge (and price premium that comes with it) if you don't need it. Buy a slightly downgraded model (especially for your first machine), and spend the extra money on making the computer more useful.
So just like buying a car isn't about buying the biggest motor, buying a computer isn't about just the CPU. Balance all the costs and accessories, and decide what you need in an entire system and how you're going to use it.