There is a major problem facing the average person when buying any kind of computing device: they don’t know what they are buying. Geeks don’t have the same problem directly, they check the hardware sites every day so they know what’s good. But every geek has had a non-geek friend come back from a big store with a “bargain” Celeron laptop to play games or an equivalent.
There is an easy fix for this but it needs the marketing teams of the biggest players in the industry to grow up and bury the hatchet on a few things. It’s time for an independent performance metric.
Every time an ordinary customer is disappointed with the performance of a device, the marketing teams in companies like Intel, AMD, Nvidia and the various ARM licensees should hang their heads in shame. It’s a good bet that everyone in those teams has got friends or family who’ve been shafted at some big store too. At least AMD doesn’t have to hang its head quite as low as the others. Its Vision badges are a start but which bright spark thought that Ultimate should be the mid-range and not the, well, ultimate?
Ordinary people get shafted every day and most of them can’t afford it. For a supposedly mature industry, the computing business is embarrassing. It’s nearly 30 years since the original IBM PC was launched. In that time, people have learned lots about computers but computers and the computer industry have learned almost nothing about people.
What’s needed is a proper, agreed, independent performance metric that works across the board. It needs to be simple enough that non-geeks understand it. There can be no politics because the industry doesn’t need another fiasco like the Vista Ready one caused by Intel’s marketing team. So it needs to be run by a separate, non-profit organisation. The test to produce the numbers needs to be freely available so that small PC builders can give their customers an accurate metric too.
It’s not a difficult thing. To prove it, here’s a simple proposal as to how it could work using just four numbers: the year, processing performance, graphics performance and power consumption/battery life. Make the metrics comparative, say 1 to 9, and we’re there. So a desktop made this year with a mid-range processor, an integrated graphics chip but very low power consumption might look like this: 2010-539. Not so obvious by itself but it works pretty well when you start comparing it with other machines:
2010-539 – mid-range processor, integrated graphics, low power
2010-991 – a serious games rig or workstation
2010-239 – a good netbook
2010-118 – an iPad
2010-x74 – a decent graphics card
2010-3×8 – a mobile processor
Any marketing teams worried about having a 1 series machine need only look at the sales of the BMW 1 Series cars to know that it’s not all about the numbers.
The average customer can easily figure out that the year makes a difference too. Just like a car, a 2 year old 991 won’t be quite as good as a new 991:
2010-991 – a serious games rig or workstation
2008-991 – a serious games rig from two years ago
The metric works when buying a PC, laptop, netbook, tablet, graphics card, processor, motherboard, hard drive, monitor and almost every other geeky item. It allows an ordinary person to easily know which one fits their needs best. Companies can still keep their obscure model numbers for us geeks but the rest of the human race stands a chance for the first time.
The same metric also works in reverse. If a piece of software like a game says it needs at least a 2006-561 system instead of a huge list of graphics chips, processors and clock speeds, everyone wins.
Some questions for the marketing departments of Intel and AMD:
When a customer can’t decide between an Athlon II X2 P320 and an Intel Pentium Dual Core SU410-based laptop, buys the one unsuited to them, regrets it and bad-mouths it for the next few years, do you really think that it helps the industry in any way? If they could look at a single metric for each machine and make a good decision, would they be more likely to spend the next few years singing your company’s praises?
Some questions for the marketing departments of Intel, AMD and Nvidia:
When a customer wants to buy their teenage son a machine to play games on, do you think they know that an Intel GMA 4500M won’t cut it and will that customer ever want to buy anything with Intel inside again? How many disappointed ex-ATI/Nvidia customers are there because they bought a graphics card that was too cheap for their needs?
Some questions for the marketing department of ARM:
Do you really think anyone knows how your licensees’ processors compare with each other? Or an Intel Atom? Or if anyone outside a few geeks really know how good the power consumption is in comparison?
A request for all of those marketing departments: go ask someone who works on the checkout of your local supermarket what their computing device buying experiences have been like.
The point of the metric above is that it would be easy for the average customer. There would be much less confusion and that would mean ordinary people buying machines that actually suited them. If average people can easily buy machines that suit them, that can only be good for our industry.