Although Chipzilla has a lot of cash, such a big purchase is particularly risky, so what did it really want?
Altera is an expert in Field Programmable Grid Arrays (FPGA) which is exactly the sort of thing that Chipzilla wants. Intel is planning to continue FPGA development and integrate FPGAs into some of their future products.
Its problem is that Altera is one of the two major competitors, and alongside rival Xilinx the two companies comprise the bulk of the market. Intel has seen a need to keep Altera sweet, had hand been letting it use its 14nm process for their latest generation of Stratix FPGAs.
Altera's FPGA's have been a natural complement to Intel's ASIC business and its FPGAs are good candidates for early production on a lower-yielding node. The buy out meant that greater number of chips to be shipped earlier than on more complex ASIC. Intel could get its foot in the door early.
Purchasing Altera allows Intel to bring its FPGA production entirely in-house, not just producing FPGAs but fitting them into the rest of Intel's development plans. Remember Chipzilla wants to build an integrated device manufacturing model and access to top-tier FPGA technology will augment its other products nicely.
So we can expect Chipzilla to be at the centre of the IoT and datacentre markets in the future. We can also expect to see FPGA technology under the bonnet of semi-customised Xeon products. FPGA fully on-die is certainly interesting and something which is now possible thanks to the buy out.
All this means that Intel has to worry less about failing to beat ARM in the mobile FPGAs are a high-margin business that will increasing in revenue as ASIC development costs continue to rise. Still it is all a big gamble for huge stakes if it pays off.