Disclosure: AMD x86 engineer in late 80s and early 90s.
Why can’t Intel put AMD out of business by buying them?
- Customers like to have multiple vendors… having one supplier is typically not good for business or innovation. Back in the 90s, Dell purchased a huge number of Intel chips, but they started to carry a few with AMD just to keep Intel honest. Also, when AMD introduced their Am386 processor, Intel rapidly dropped the ASP on their own product to suck the air out of the market; at the same time, they came out with the intel 486 to move customers away from the 386… all of this was to minimize how much money AMD would be able to make on that product. For vendors, this competition was a win/win … new generation of products that they could introduce, and cost reductions on existing PCs.
- Vendors also like to have second-source options for continuity of supply issues. A few years ago, many of the HDD companies struggled when Thailand was flooded. This had a ripple effect on PC companies which couldn’t ship PCs without HDDs, and were having to pay significantly higher prices for HDDs that they did buy.
- Governments like competition and discourage monopolies. One of the largest anti-trust suits ever was US vs Microsoft in the 90s. Microsoft had a dominant OS for the PC and was starting to bundle other programs (think Office, web browser, etc) along with it. The US government was threatening to break up Microsoft and the other half of the Wintel (Microsoft Intel) monopoly was probably feeling threatened. I suspect that Intel saw what was being threatened on the software side and said we can’t risk that on the hardware side of the PC. Having AMD there gave them cover that they didn’t have that monopoly.
- End customers like to have options and competition. AMD introduced the AM386 with a wopping 40MHz at the same price as Intel’s state of the art 33MHz processor. A few years ago, Opteron processors from AMD gained a significant hold on the server market when Intel struggled. This enabled the server market to innovate and improve, even when a key vendor did not.
- AMD is irrelevant in two of the hottest and most threatening markets for Intel. These days, considerable investment is being made in AI processors; this is a threat to Intel, but AMD doesn’t appear to be the right answer to the threat. In the mobile space, a huge market for Intel, companies like ARM and Qualcomm are the main competitors; again, AMD wouldn’t help them in this rapidly growing space. Acquire AMD? Why make an investment to gain market share in a segment (PCs/notebooks) that is rapidly declining?
PCs have been manufactured throughout most of my 30 year career, and while there has been a lot of consolidation, no monopolies exist. (Arguably three HDD companies, 2–3 GPU options, 3 DRAM suppliers, 5 flash companies, and numerous suppliers of commodity components.) The customers, markets, and end-customers all don’t want it and see the benefit of having competition.
And I’d also point you to Jeff Gruszynski ‘s answer on this, as I think it correctly highlights how complacency was not good for Intel wrt mobile and embedded, and opened the door to ARM in cell phones.
PS. It will now be interesting to see whether RISC-V unseats the monopoly that ARM currently has, or whether ARM remembers how Intel’s complacency opened up the door for them. Similarly, it will be interesting to see whether NVDAs strong position in GPUs gets upset by next generation AI processors. Companies’ focus on their core business often makes them blind to new and emerging threats (read “The Innovator’s Dilemma” for more on that)