Beating the System: The AnandTech Guide to Economic Upgrading

Quantifying Price

Something that Anand and I like to talk about is buying the best hardware that you can for your dollar, today. If you think about the long run, your time is expensive. Even if you don't have a lot of money to spend on upgrades or a new configuration, purposely procrastinating can sometimes be the worst thing. Waiting for the "next best thing" always results in a perpetual cycle of more waiting, and even more so in the computer industry. Instead of beating around the bush, let's go back to the original statement that we made about the ideal time to upgrade.

The right time to upgrade can usually be modeled around how valuable additional Quality is to you.

The principle seems fairly obvious, but it's easy to get lost in the NDA launches and rave reviews. A former executive to a very large motherboard company once told me that selling computer hardware was almost identical to selling produce. Every day that hardware sits on the shelf in a warehouse is another day that it loses value (to the company). With the exception of the short term LCD and DRAM spot markets, virtually all computer hardware sells for less today than it did yesterday. Sometimes we see small fluctuations in price where demand out-weighs supply, but for the most part, we don't see significant changes in computer hardware pricing like other commodities. Due to the extreme pace at which hardware evolves, buying computer hardware can almost always be summed up as, "What you buy today is cheaper tomorrow."

Prices don't fluctuate to the point where a product introduced today will cost more six months from now, and the path from today until then continually decreases (though not necessarily linearly). Then why do we recommend not waiting for the next best thing when buying a new component? If you feel that it's time to buy a new component, your value of that component should out-weigh the Price – even over time. If you plan on buying an Athlon 64 3200+ today for $200 versus $100 a year from now, hopefully your value of the component can be quantified at more than $100 per year. To state the obvious, if an upgrade doesn't have any value to you, it isn't worth buying.

Now, let's back up for a second. When we buy new upgrades, we tend to look at singular components – this can actually be a costly habit. As we stated on the previous page, even if you have a slower processor, the money may be better spent on a video card than on a processor. Instead of picking out a particular component set from which to upgrade, a more economic upgrade path may be to consider all components that fit in a particular price range and determine their relative Quality.

Quantifying Quality

Even though Quality is a numerical representation of performance, features and service, finding a method to determine an exact value of several different components to fit your user habits shouldn't be too hard. If you spend most of your computer experience gaming, then relying on gaming benchmarks for the games that you use is the most practical step. If you feel that the time to upgrade is now, so that you can get better performance out of Half Life 2, then the first step would be to quantify the Quality that a new hardware upgrade will bring to your system. A certain video card might double performance, so its relative Quality in the computer is 200%.

On the other hand, if we feel that we need to upgrade to maximize our storage space, then quantifying Quality over different components becomes even easier. Any video card that we add to the system would increase storage space by 0%; thus, the relative Quality of a Radeon X800 Pro when maximizing storage space is 0%. Subconsciously, we all do this same process at some level or another when buying hardware.

The dicey part gets when we start quantifying data that doesn't have a real world correlation – or perhaps data that isn't entirely complete. Determining the exact performance increase from a Radeon 9600 Pro to a Radeon X850XT is not something that you'll readily see published on AnandTech or anywhere else. The quick and dirty trick to determine relative performance on that scale is to find an intermediate product and look at the performance between those two. For a certain application, a Radeon 9800 Pro might be twice as fast as a 9600 Pro, and we also might know that an X850XT is twice as fast as a 9800 Pro. The resultant relative Quality should be about 400%.

We call that the quick and dirty method of finding relative Quality, but the only true method of determining the relative quality of one component compared to another is to have a test scenario. This is possible for more mainstream configurations as you can see in a lot of our motherboard reviews, but it's a little harder in real life. Our best advice is to pick conservative estimates for your quality assumptions.