Marking technology is going to be quite similar to marketing a game -- in fact, if you have games made with that technology, that would be a great boon to adoption. Show them off.
With SlimDX, we started by hosting the project on Google Code, advertising very narrowly -- within a single community of developers (GDNet). That got us a few high level bits of feedback that we iterated on. Most importantly, though, we used the framework ourselves to build our own projects so we knew what the user experience was like and tweaked that until it was a good user experience. This includes things like making the end-user runtime installers as painless as possible, coming up with a decent documentation build, and providing at least basic "getting started" samples.
Then we kept plugging it whenever it was relevant in forums, we started a Twitter account to watch for people mentioning it and interacting with them, and slowly built up a community around the library, at which point the community was able to take on some of the burden of mentioning it all over the place.
You have to be careful when plugging your project, of course; we tended to be able to say, "well, you could use XNA if X, Y and Z or you could use SlimDX if A, B and C." It helps, I think, to be honest about the advantages and disadvantages of your product.
The most important thing is probably going to be establishing yourself as quick, responsive, and client-focused early on. We were able to turn around bug reports and feature requests extremely quickly in our first few months and invested a lot of time in making the framework easy to get and start using. Establishing that solid initial relationship with your early-adopters will really help propel you forward.