13: Gavin Baker's Intel Autopsy, Epic vs Apple, and Is Your Product a Kevlar Vest or a Plumbing Tool?

“If I read a book that cost me $20 and I get one good idea, I’ve gotten one of the greatest bargains of all time.”

“If I read a book that cost me $20 and I get one good idea, I’ve gotten one of the greatest bargains of all time.” — Tom Peters


"The man who doesn't read good books has no advantage over the man who can't read them." — Mark Twain

As promised in Edition #8:

Around 8 years ago I got into a new hobby. Folding knives.

Some people have a lot of uses for knives, stuff to cut around the farm or whatever, but that’s not really my case. I got interested mostly for the engineering details, and for the fun design aspects.

To most, it seems like a knife is a knife, but like with any sub-culture, if you dive in the rabbit hole, there’s an endless world of details and variations.

You can learn forever about blade geometry, the pros and cons of full-flat grinds vs hollow grinds, the trade-offs between various kinds of steels (including some advanced metallurgy, like compound steels that are made from metal powders using very cool processes), various blade-locking and opening mechanisms (each with their own trade-offs), liner materials (like G10 composite, which is similar to that often green material used to make computer circuit boards, actually), some have teflon-coated blades, assisted opening, some are ultra-light or especially water-resistant, woven carbon fiber or titanium handles, etc.

I never really got into high-end knives, like Chris Reeves and Striders, or the even higher-end custom shops. Mine are all pretty much low-to-mid-range, but they were fun to learn about and collect (until I had kids, and then locked them away somewhere safe). I’ll admit I used to lust after a Strider similar to this one

So here’s a quick look at what’s in my Pelican Case:


Investing & Business

Selling Kevlar Vests or Plumbing Tools

With some companies, the way to think about how their customers buy their products is: "If you were buying a bulletproof vest, would you pick the cheapest one, or the best one?"

The line comes from an Everbridge presentation, but it was reading about security companies, and the whole “mission-critical” thing I mentioned in edition #10 that made me think of it.

It’s basically about asymmetry, right? That’s how insurance is sold. “Yeah, it costs you $1,000, but if things go bad, they may cost you $1,000,000, so even if the odds are really low, you don’t want to be exposed to catastrophic failure…”

Some products are more like this, like insurance (Everbridge for critical events like a tornado or an earthquake or a crazed shooter in your building, stuff that hopefully never happens to you, though in 2020, you can never know).

Some other products are more on the “mission-critical” side, like having good tools that you use all day, every day, to do your work (ie. if you’re a plumber, you probably don’t buy the cheapest tools, because if they break or don’t work well, that $10 wrench may cost you hundreds or thousands of dollars in lost time, a lost contract, damage to reputation, etc). There’s asymmetry there too, but the usage pattern is different, so how you think about it is different.

So if you’re paying, say, a cybersecurity firm to protect your network, you expect to be low-grade attacked all day, every day, just by the nature of the modern internet. And you expect more sophisticated attacks and breaches to happen only once in a while. So IT security kind of combines both the kevlar/insurance and the mission-critical tool.

Though it’s also a really competitive, fast-moving space where everybody is doing the Red Queen’s “running to stay in the same place” thing.

But in spaces where things change often, sometimes there are changes that aren’t like the “regular” change, and every so often, the space can get reconfigured into a new “era” that may last a while (whithin which there will be lower magnitude change, of course).

That’s what I’ve been thinking about lately. No solid conclusions yet, but it’s interesting to think about.

Haven’t You Read Andy Grove?

Remember that Tom Peters quote at the top of this letter? Buy the book.

Gavin Baker: Autopsy of a Mistake

Gavin Baker, who you should follow on Twitter and listen to here and here, wrote a longer piece about a recent investing mistake he made in the semiconductor space.

Now first of all, more people should do public debriefings on their mistakes, not only because others can learn from them, but because writing is thinking, and taking the time to write about it is a way to both clarify and engrave the lessons into your brain.

You should read the whole thing because it’s all good, but I want to point out a specific passage that made me wonder about the same phenomenon elsewhere (kind of like the US dollar being the reserve currency, or Moody’s and S&P Global being kind of built-in the credit systems, etc):

Intel has the “exorbitant privilege” of having all x86 software code effectively optimized for their own architecture. No dominant digital processor company has lost the #1 market share position over the last decade because of this privilege. What this means in practice is that software just runs faster on Intel processors than AMD processors — it is not enough for AMD to be 20% faster in benchmarks, because real world software optimizations by developers mean that software generally runs faster on Intel. AMD needs to be much faster for a sustained period of years for software to be rewritten and optimized for their own particular architecture. As a sidenote, this is why Windows on ARM did not work for PCs and ARM server CPUs have been slow to take off — any ARM CPU needs to offer dramatically higher performance per watt per $ to justify rewriting software so that it can run on ARM instead of x86. On top of the “exorbitant privilege” conferred by software developers optimizing for Intel’s own architectural flavor of x86, Intel is much better at writing their own software — compilers, frameworks — than AMD and all of this software is essential for other software to run well. Having followed the GPU market for 20 years, I can confidently state that software has never been a strength of AMD and is the main reason they lost to Nvidia. In the CPU market, Intel’s inference software frameworks are far superior to AMDs and inference is a growing workload for CPUs. (Source)

The part about the software being important to sell the hardware is also important, and too often missed by non-technical (or even technical) people.

For example, most people think of Apple mostly as a hardware-maker, and software is only a distant second in people’s mindshare when they think about the company.

Yet if you went to most core Apple customers and told them that they could either have a Macbook running Windows or a similarly specced PC laptop running MacOS, or a Samsung Galaxy running iOS or an iPhone running Android. I think most (though not all) would pick the software over the hardware.

They’d rather have both working well together, but the software is clearly a huge differentiator for the hardware.

It’s similar with Nvidia. A lot of the software that they built (especially CUDA), or the third party optimizations that they did, helps sell their hardware. Some engineers on Hacker News may complain about it, but at the end of the day, a lot of customers just want things to work and care more about the package as a whole than about one or two elements on the checklist that may be inferior to some competitor.

Back to Gavin and what happened at Intel, I have to point out that this is just good writing:

I have tried to to think of a good analogy for this and the best one I can come up with is that Intel attempted the world’s highest pole vault in a torrential rainstorm while blind folded with no mat on the other side.

And because I like Dune too, I can’t help share this very good point, which contains a very good reference:

I think it would be an enormous mistake for America not to do everything possible to help Intel stay at the leading edge with its fabs in America and Israel. If Intel were to outsource manufacturing to TSM, Taiwan would be the most geopolitically important country in history. It would be no understatement to say that “Taiwan would be the new Arrakis” per Epsilon Theory. Modern semiconductor manufacturing is at least as important to the economy as oil was in the 1970s…

Read the whole thing.

Update on Nvidia & ARM

Talks between UK chip designer Arm and US firm Nvidia are accelerating with a $40bn (£30bn) deal expected to be thrashed out by the end of the month. [...]

The exclusivity period is expected to last 30 to 45 days, after which the companies would walk away or the talks would open to other parties. (Source)

I still hope it happens, if only because it’ll be fun to learn more about current ARM tech and financials.

Epic vs Apple: 1984 Revisited

Epic Games released this short parody of Apple’s iconic 1984 ad (which was directed by Ridley Scott, of ‘Blade Runner’ and ‘Alien’ fame) before suing them recently:


Science & Technology

124 Years Old Film Remastered by Deep Learning Algorithms

If you’re curious how much of an improvement it is, the original video is here.

Here’s San Francisco in 1906:

Tokyo 1913-1915:

The obvious comparison that comes to mind is Peter Jackson’s documentary ‘They Shall Not Grow Old’ (2018), which is made with WWI footage that has been cleaned up and colorized.

But that process was very labor-intensive, with lots of manual adjustments and coloring. It’ll be interesting when all this can be (mostly) automated at a low cost. How many thousands and thousands of hours of old footage could be given a new life?

Would future generations be better able to connect to the reality of the horrors of WWI and WWII with moving pictures that put less distance between them and history?


The Arts

Check out the video in the Tweet, it’s a good data-visualization, with the top track’s audio playing while it’s at the top.

Chaos vs Order

Listen to This: Hymn to Freedom by Oscar Peterson

Speaking of freedom… If you want to hear a track I really like, have a listen to this fine piece of music by legendary Montréal jazz pianist Oscar Peterson (known as O.P. by his friends — he released over 200 recordings, the first in 1945 and the last in 2007).

Oscar Peterson award to be handed out at York U gala ...

But you gotta listen to the end!

Don't miss that crescendo, and all the little musical jokes sprinkled throughout. When I’m in the right mood for it, this track makes my spirit soar (or whatever the cliché is) and helps me feel some of the goodness in humanity’s heart. Anything that helps us aspire to this level of joy and freedom for everyone, every day, is a good thing indeed.

Listen on: Spotify. Apple Music. Youtube.