Years ago I worked for a start up that had Intel Ventures as one of our investors. During a break in board meeting I was chatting with the Intel board member and I asked him what was Intel’s secret to their success. Mind you, this was when Intel was in full power and one of the dominant companies in our industry. I expected his answer to be along the lines of technology advancements, or engineering talent or maybe even their manufacturing advantage. But the answer was none of those things.
What he actually said was, “At Intel, we eat our young.”
It took me a second to unpack this as at first blush it sounds like generic corporate speak. But the more I thought about it the more I realized that there was a dark secret under that statement. And that was the fact that the chip industry is a never-ending arms race. The very bones of the business are built around constantly punching out chips that are faster, cheaper, or more energy efficient. It is truly on of those industries where if you are not getting ahead you are falling behind.
The truth behind that simple statement was that Intel was brutally aware that every chip they released was going to be quickly matched and then surpassed by other chips, sometimes even faster than we consumers would think. And so their attitude was that if someone was going to eclipse their chip, it might as well be them. It sounds very simple, but it also is incredibly stressful for an organization in that you have to constantly be willing to reinvest money, time, and headcount to destroy your own products. If you take you foot off the gas for a moment, you risk someone else suddenly getting the next chip out and now you are chasing them.
I was thinking about this conversation this morning as I watched Intel CEO Pat Gelsinger on CNBC. Intel is about to be the beneficiary of government funding that will help it build factories here in America. And to be be clear, I am all for this, as COVID and the resulting shutdown showed how vulnerable we were to relying on offshore manufacturing. So this part of the interview was fine, but it was the next part that bothered me. Gelsinger was asked when the new plants would be coming online, and he very excitedly announced that it would be within five years.
Five years.
Now understand that I am at Game Developers Conference in San Francisco this week, where NVIDIA has been drilling out more press releases per square inch that you can imagine. One of those has been announcing new chips that will cost tens of thousands of dollars. And a roadmap that is packed full of new products on both the hardware and software side. And that’s when it dawned on me.
NVIDIA is running the Intel playbook. In fact, they are running it better than Intel is. They know if they let up they risk someone else surpassing their chips, and they are not going to let that happen. And intel is left with plants that are coming on line with chips that will be generations behind what NVIDIA and others are producing by then.
This is core of the challenge Intel faces right now. They know they need to get back to the manufacturing prowess that powered their success during their glory years. But are they too late? The math they are faced with is that by the time these new plants come on line, they risk the chips coming off them being comically behind their competitors. And this means they will be releasing chips that by then will be near commodity if not commodity items. Which not only means that they will be priced lower, but more crucially will carry lower per unit margins. Those lower margins are the problem, as they will generate a cash flow that may not justify the Capex invested in those new plants. And that’s the problem with taking your foot off the gas. Sometimes you never get the chance to get back in the race.
