PortalPlayer also much smaller in terms number of employees, which means much smaller ongoing investment in the form of salaries.
Yeah, and PortalPlayer had less revenue and higher margins, so they didn't show up in financials as much as VIA potentially would. However...
None of of NVIDIA's acquisitions have been all that large in terms headcount. If they really wanted VIA's x86 maybe they would buy that piece or do something like the 3DFX deal.
VIA has plenty of cash reserves, so if you want to wait for them to go bankrupt, you're going to wait long. Also, it is key to understand that they are pretty much losing all their markets and that soon they'll be a x86 company anyway.
Their Intel FSB license expired in April 2007, which means they can no longer develop any new Intel chipset and that existing ones could no longer be produced as of April 2008. At the same time, they have lost the vast majority of their market share on the AMD platform in recent years due to NV, and I suspect AMD's acquisition of ATI is just going to squeeze them to near-zero eventually.
As for GPUs, S3 still hasn't released their DX10 stuff, and I'd be relatively surprised if it was super-competitive (but who knows). The rest of VIA's business is basically a bunch of commodity peripherals and analogue. While I doubt NV would care much about being able to sell that stuff (let alone because it's not super-high-margins), I think it would be in their best interests to have the IP in order to be able to integrate that eventually (possibly on-package rather than on-die?)
If your arguement is "VIA has too many employees", then the obvious answer to that is to wait some more until they get rid of more of them in a desperate attempt to be adapt to their new x86-centric business model, or in order to make themselves a more desirable acquisition target. By early 2008, things might be quite different and it is true that now may not yet be the best time to buy.
It's seems likely they have in-house x86 expertise with some of those Stexar people. The question is, as JHH posed at some financial event, are they going to build x86, go with ARM, or will it be something different?
Yes, but x86 designs take time to develop obviously. Even if they started, say, 6 months ago... That still puts availability in 2H10, I'd suspect.
VIA's CN will literally be available within 6 months apparently, and if an acquisition happened early enough it should be realistic to achieve an integrated solution by mid-2009. That's a significant head start compared to the alternatives and the competition.