Home / Opinions / Power Mad

Power Mad

by

Our working life at TrustedReviews is split between products and technologies, which is similar to the difference between ‘cause’ and ‘effect’. Products that use a new technology tend to be more interesting to the reviewer than a version four product that is a slight revision of version three but the fact of the matter is that you buy a product rather than a technology. That’s the reason why Spode gets stuck in his corner running group tests of products that are very similar so that you the buying public can figure out how to spend your money.

Taking Spode’s AM2 motherboard group test as an example, the technical changes behind the move from Socket 939 to AM2 can be summed up in a short sentence. It’s a different socket because the revised processor supports DDR2 memory instead of DDR. Easy peasy. The trick, of course, is to figure out whether the changes are worthwhile and good value. In the case of DDR2, we were completely underwhelmed when it launched at 400MHz back in June 2004 along with Intel’s i915P chipset as the latency killed performance. As the speed of DDR2 has risen through 533MHz, 667 MHz and 800MHz the latency issue has diminished, making this a good time for AMD to adopt the technology and in a year or two you won’t be able to find a motherboard that supports 184-pin DDR memory.

There are some happy occasions when a new technology yields benefits that are blindingly obvious with no downside whatsoever. The use of perpendicular recording in hard drives crams more magnetic media into a given amount of platter space, which enables the manufacturer to boost storage capacity significantly. In the case of the Seagate 7200.10 it went from 500GB to 750GB. Result.

Sometimes we grasp the benefits of a new product without knowing exactly how it works, such as the time when Samsung increased the speed of its SH-S182D DVD writer from 16x to 18x. ‘Faster you say? Oh yes so it is, very nice. Thank you.’

Speed increases are relatively easy to comprehend but every so often we’re faced with a new technology, such as DDR2 memory, that is part of a platform. It takes a while to figure out how the different parts of the system are performing and history shows that the i915P chipset worked well, DDR2 needed a speed revision or two and that Prescott Pentium 4 was a stinker. We were helped towards that conclusion as the previous generation i875P/DDR/Northwood combination was similar in many respects to i915P/DDR2/Prescott so while we couldn’t do a back-to-back comparison at least we knew the correct questions to ask and in time we got answers.

comments powered by Disqus