I am probably going to be a bit out of the mainstream in my advice.
If you are thinking of going tech in the future, then you are looking at a temporary device. Do not buy a computer with the intention of using it for tech diving. One reason is that it is very possible that you will train with an agency or individual instructor who does not use one, and you could end up having spent a lot of money for what will be essentially a simple bottom timer. A second reason is that if you do use a computer for tech, but the time you get to that point in your training, a whole new generation of computers will be out there, and anything you buy now will be obsolete.
So what should you look for in a computer?
One thing people will push on you is the ability to switch gases. You probably don't have any idea why yo would want to use multiple gases during a dive right now, but that is needed for introductory level tech diving and for some cave diving. that means, again, you won't need it for a long time. If you do go into tech, you will eventually want to switch to a gas mix that includes helium, and you will find that many of the multi-gas computers do not use helium. So ask yourself when you would ever use a multi-gas function before deciding you want that.
The next issue is whether or not to have an air-integrated computer. That is a legitimate question, and I will only suggest that you weigh the pros and cons carefully for that.
I think pretty much all computers do nitrox now, so you don't have to think about that.
One feature I do recommend if you think you may want to go into tech is the ability go into what is called gauge mode. If your computer has that ability, you will be able to use it effectively with decompression dives without having to worry about it saying you are bent all the time. Not all basic computers can do this.
A suggestion has been made about being able to use a heart rate monitor feature on some computers. I have no idea why anyone would need that. Heart rate has no established effect on decompression rates, so any attempt to tie it into the computer algorithm is speculative at best, and I am told that in the computers that do it, the differences are pretty minimal.