Both are not final silicon yet, just there to kickstart the software infrastructure so that when the phones do show up, the software will be ready.
The first one was talked about last week, it is the third generation mobile graphics part, and has not yet warranted a marketing name. Not much to say here, bigger, faster, smaller, lighter and comes in a range of trendy southwest colours.
The more interesting of the two is the Vector GPU. When was the last time you heard of anything new in GPUs? OK, vector GPUs are not totally new, someone has probably done them before but this is probably the first one that will be made in quantities that you can't count to on one hand.

Why vectors? Two main reasons, scaling and flash. Imagine having a program on a mobile where you don't have to worry about font size, graphics size or even screen aspect ratios. With a fully vector UI, you can just write the graphics and give it a scaling factor. Better yet, you have that built into the phone, and voila, you are done.
OK, Sun killed this whole concept with Java, but it should be significantly easier to support new hardware and software with vector based GPUs. Write once, run anywhere, hopefully debug a lot less, but in reality something will bite your ass anyway.
The other interesting one is flash acceleration. With the advent of mobile flash variants, little sites like YouTube and others using it for video, and other little no-names picking up the concept, hardware acceleration has merit. If you can't see the obvious worth here, I won't belabor the point.
The two DAAMIT mobile chips are totally distinct and both worthwhile in their own way. It will be really interesting in a generation or two when you start seeing converged devices, anyone for flash video textures on 3D surfaces.... on your phone? ยต