Intel recently announced a plan to eliminate 15,000 jobs after historically bad quarterly results.
It’s been rehearsed to death that Intel missed the boat on mobile and AI, but execution issues seem just as important to the story. When Apple ditched Intel for its own processors in 2020 to glowing reviews, the root cause was quality assurance. Even if Apple was a small, difficult customer, more alert leadership might have seen their departure as a canary in the coal mine and moved heaven and earth to fix QA. Instead, four years later, Intel’s desktop chips are still having serious stability problems, and other laptop vendors are jumping ship.
Meanwhile, AMD in gaining ground in both the server and desktop markets, but from what I can see, AMD is not exactly firing on all cylinders. Consider the case of the 8600G and 8700G, released in early 2024. They’re both APUs (“Accelerated Processing Unit” — who comes up with these names?) with CPUs and GPUs on a single die aimed as the low-power desktop gaming market. Both got pretty tepid reviews because the all-in price, when considering the new motherboards and faster RAM they require, was too high. Last month, AMD released a laptop processor that outperforms both desktop parts handily with lower power consumption.
So why did AMD release overpriced products and then make them obsolete 6 months later?
One explanation is that big companies are like octopuses, and each product division/arm is essentially its own brain with its own reward structures. I think this model works pretty well for Google, where product shutdowns are depressingly regular and generally attributed to misaligned incentives. But AMD’s issues seem to be about incoherent go-to-market (GTM) strategy. Chipping away at Nvidia’s market share at the low end with low-power, integrated chips is sensible. But the details need to work.
Some other places where we see confusing GTM:
AMD’s naming scheme is truly a marvel of impossible to remember terms. Why make this hard for buyers? (This is a surprisingly common problem among manufacturers — what does “G34WQi” evoke for you?);
Its most recent flagship desktop processor is getting mixed to negative reviews on price, performance, and a serious driver bug that’s lasted for over a year;
The company continues to lag badly in the AI/GPU market in large part because its software isn’t competitive.
Basically, I see a company that capitalized on its main competitor’s struggles and continues to release some very successful, specialized products, but that seems to be its own worst enemy in terms of GTM. (Some might attribute this to having a CEO who specializes in hardware rather than business.)
I think this has gotten less attention than it might in light of Intel’s more obviously existential struggle. But inability to market and sell the right products to the right customers is also a serious crack in the foundation.
If I were tasked with fixing things at AMD, I would say:
Get the product marketing right. Name the products simple, memorable things. Look at Apple.
If the price proposition doesn’t work, subsidize. Sell the 8600G at a loss, make future processors backwards compatible, and encourage upgrades.
Make the AI software work.
But who knows? I am just seeing a small part of the puzzle, and they still release great products. I will continue to watch with interest.