DX10.1 is 'trivial'

FOLLOWING our story last week revealing that DX10.1 is going to require a hardware upgrade, we had all manner of crazy emails from readers who were incredibly cheesed off, shall we say, that their hardware was being superceded for no particular reason.

We also had plenty wondering whether developers would even bother supporting a standard likely to have such minimal adoption.

Here, then, is an interesting perspective from a high-ranking game developer, who is currently working on a highly-anticipated DX10 game, but who wished to remain nameless, for obvious reasons.

"It's funny," he whispered, "I recently had the whole "why not DX10" discussion with Nvidia and MS. But basically, if the API is not changing a lot, providing optional support for 10.1 is trivial, just a few lines of code in how you initialise the DX objects.

The larger issues of whether to go with DX10 or not are related to market share - and interestingly the rate of OS upgrade seems to be the limiting factor. There are about 4x as many gamers with a DX10 graphics card and XP out there as there are with a DX10 card and Vista, and although even with the XP users in there right now we're still only talking about 10% of the player base overall.

MS would do well to encourage uptake of Vista among gamers, preferably by lowering prices of Vista Premium upgrades."

So there you have it - 10.1 support is trivial, but the main problem is the minimal DX10 adoption in hardware and the even more miniscule DX10 hardware to OS ratio. Progress is a slow thing. µ