315: CHIPS Act, AMD Q2, Amazon Q2, Datadog Q2, Transdigm Part 4, Thoma Bravo + Ping, Intel's Xeon Bugs, and Competitive NES Tetris
"it often takes a jolt from the outside to trigger new insights."
Motivation comes from passion, not comparison.
–’Dragonfly’ (A top-level Doom mapmaker)
🚲 📢 🏙 Rush hour in Amsterdam:
This reminds me that we don’t talk enough about urban noise-reduction as a huge quality-of-life win (one more reason why diesel trucks eventually being replaced by EVs will be great).
Humans have not evolved in environments that were constantly full of random loud noises, over a background of constant low-level noise. It has to have an impact on chronic low-level stress/alertness for people living in such environments, even if subconscious (I know city-dwellers consciously tune out a lot of it after a while, but it doesn’t mean that the sensory input isn’t there). ¯\_(ツ)_/¯
I’ll have to see if good studies have been done on this (cortisol levels as proxy?)
💚 🥃 I'm thinking of it kind of like we're sitting in a pub somewhere, and I'm talking about various things that interest me or that I've learned about recently.
If you like it, once in a while, you send me a scotch to keep me talking and show appreciation. That’s what becoming a paid supporter is. All very casual & civilized. And a jolly g'day to you too, mate.
Liberty’s Highlights is reader-supported. To support my work, consider becoming a paid supporter. (this gives you bonus access to the private Discord where 85+ friends-of-the-show hang out) 🤖
🧩 A Word From Our Sponsor: Heyday 🧠
Do you have 100+ browser tabs open right now? 😬
Give your memory a boost with Heyday, the research tool that automatically saves the content you view, and resurfaces it within your existing workflows. 👩💻
It’s like cheat codes for your memory. 😲💡
🏦 💰 Liberty Capital 💳 💴
🇺🇸 CHIPs Act, Semiconductor Nationalism, Defense, etc 💣
War went from throwing as much *of everything* at the enemy as possible pre-Cold War — which meant that it was all about making as many bombs/tanks/planes/ships/etc as possible — to the semiconductor era where precision matters a great deal too.
One reason precision weapons are so powerful is because of a basic fact of geometry. The damage an explosive does drops off very quickly as the explosion becomes more distant (generally speaking, the radiative heat drops off as the inverse square of the distance, while the strength of the shock wave drops off as the inverse cube). So making a weapon 10 times as precise means that you make it somewhere between 100 and 1000 times as destructive. (Cluster munitions help to solve this problem, but precision is better, especially if you have precision cluster munitions.) The superiority of precision weapons is one reason why U.S.-made HIMARS rocket artillery is so effective against the Russian Army in Ukraine, despite only firing a small number of projectiles.
Another reason precision weapons are powerful is that they let you hit moving targets from very great distances. Planes, tanks, ships and other armored vehicles simply aren’t safe in a world where missiles know where they’re going. This is why U.S.-made Javelins and Stingers, as well as similar weapons from elsewhere, have been the scourge of Russian tanks, planes, and ships in the Ukraine war.
So how do we get all this precision?
Computer chips are essential in both delivering precision weapons and evading them. They are a crucial part of guidance systems, cameras and computer vision, other sensors, and so on. And computer chips have become an essential component of modern military vehicles themselves. Of course this all goes double for drones, which are playing an increasingly central role in warfare. And computer chips are absolutely indispensable for lots of other aspects of modern war — electronic warfare, satellites, communications, encryption, you name it.
This makes it very obvious how important it is for nations to control their supply of these chips.
The chips don’t *all* have to be cutting edge to do their job, but some of them need to be, which is what makes TSMC (and Samsung) so crucial these days.
Ergo, the US Government is subsidizing
Intel the US semiconductor industry so that more cutting-edge foundries are built in the country, and more expertise is either home-grown or brought back to friendly soil.
The US is also playing offensive to try to block, or at least slow down, China’s own semi industry, blocking the sale of the most cutting edge extreme ultraviolet lithography (EUV) equipment from ASML in the Netherlands, among other suppliers to fabs.
The US is still way ahead when it comes to the *design* of chips — most of its companies are now fabless, with exceptions like Intel and Texas Instruments — but if they want to also rapidly ramp up the fabbing, they’ll have to reform immigration to attract the talent needed (because it’ll take decades to form enough new engineers).
In other words: Chips need 🧠. The US only has 5% of the world’s 🧠, but if it can attract and cherry pick the best 🧠 from the rest of the world, it can boost its chips 🧠 capability.
🔥 AMD Q2 🔥
It’s a bit harder for Intel to blame macro factors out-of-its-control for its terrible results when rival AMD is going brrrr:
The data-center segment is up 83%! That’s where the highest margin chips SKUs are… 💰💰💰
Gaming is up 32%, which is solid 🎮
They’re also following Nvidia’s lead with the acquisition of Pensando System for $1.9bn, giving them their own DPUs to sell into the data-center space. I don’t know much about Pensando and have no idea how they compare to Mellanox tech (on which Nvidia’s Bluefield stuff is originally based), though ¯\_(ツ)_/¯
Amazon Q2 🚚📦📦📦
This came out last week. Is it fishwrapper already? I hope you don’t come to me for up-to-the-minute updates on earnings… I do VLFT, not HFT.
I’ll keep this short, a few highlights:
Our compound annual growth since the start of the pandemic stands at 25%, a growth rate higher than what we were seeing prior to the pandemic.
Prime members have meaningfully increased their spend since the start of the pandemic.
While there’s often the cynical tendency for companies to reframe time periods in a way that makes them look best, in this case, it does make sense to zoom out a bit because the pandemic period was so anomalous for Amazon.
It’s not every day that companies with hundreds of billions in revenue grow 40%…
Our seller community also had a strong Q2. Third-party sellers represented 57% of all units sold on Amazon in Q2, the highest percentage ever.
While this comes with its own set of trade-offs — it’s harder to control the quality of what is sold with 3P sellers — it’s more profitable in most cases for Amazon to keep increasing that 3P to 1P ratio over time, so it’s likely we’ll see it keep climbing…
Inflationary pressures remained at elevated levels in Q2, similar to what we saw in Q1. These include pressures from higher fuel, trucking, air and ocean shipping rates, which we expect will continue into Q3. [...]
on the seller fee, again, we added that fee grudgingly in May to compensate for some of the inflationary pressures we're seeing. I don't want to give you the idea that either of those fee increases came close to covering our costs.
They also mention electricity rates for AWS data-centers going up as natural gas gets more expensive.
Inflation is Amazon’s dragon, right now, but it can’t really slay it on its own… 🗡🐲
AWS, where net sales were $19.7 billion in Q2, up 33% year-over-year, and now represent an annualized sales run rate of nearly $79 billion.
AWS continues to grow at a fast pace, and we believe we're still in the early stages of enterprise and public sector adoption of the cloud.
It’s still bonkers to me that AWS can grow 33% at almost $80bn of run-rate while doing 29% in operating margin.
Imagine how much hardware and software — engineers and fiber optics cables and large buildings with very big power wires — you need to add to the very complex system to grow by that much in 1 year! 🤯
In 2021, we incurred approximately $60 billion in capital investments.
About 40% of that is comprised of technology infrastructure, primarily supporting AWS as well as our worldwide stores business. Another 30% of the $60 billion was fulfillment capacity, and a little less than 25% was for transportation.
The remaining 5% was comprised of things like corporate space and physical stores.
For full year 2022, we do expect to spend slightly more on capital investments than last year, but the proportion of capital spending shifts among our businesses.
We expect technology infrastructure spend to grow year-over-year, primarily to support the rapid growth in innovation we're seeing with AWS.
We expect infrastructure to represent a bit more than half of our total capital investments in 2022. For the worldwide stores business, we've continued to moderate our build expectations to better align with customer demand.
We expect the fulfillment and transportation dollars spent on capital projects to be lower in 2022 versus the prior year.
This capex is ridiculous. And of course, AWS is eating ever more of it.
Amazon’s capex is a bit like TSMC’s capex. I mean, the processes aren’t as dark arts magic as TSMC, but the absolute amounts mean that only giants can follow in those footsteps…
Big shoes to fill indeed 👞
this net loss includes a pretax valuation loss of $3.9 billion, which is included in nonoperating expense from our common stock investment in Rivian Automotive.
In the U.S., we have started making customer deliveries using the Rivian electric delivery vehicles
A reminder that mark-to-market investments are in there. Great to see that EV delivery vans are rolling out. I hope they ramp up really fast and competitors follow suit.
The majority of advertising revenue is in North America. But having said that, we are making great strides in international as well.
we're also, as you mentioned, expanding our array of advertising products from our consumer websites to video opportunities, Twitch and others. […] we, of course, introduced interactive ads last year for streaming video content, things like Freevee. We've got Amazon Music's ad-supported tier as well for audio ads. […] We're obviously excited about Thursday Night Football and Amazon streaming TV ads
Ads were up 24%, which is impressive at that scale.
It’s easy to only think of Google and Facebook when you hear “digital ads”, but Amazon is speed-running into that space pretty quickly.
Update: It just came out this morning that Amazon is buying iRobot for $1.7bn. Maybe they just needed lots of Roombas to keep distribution centers clean 🤔
🐶 Datadog Q2 🐾 🦴
So they grew revenue 74% at 81% gross margins, 15% FCF margin (but hey, it was 36% last quarter) with a net retention rate above 130%. 😲
Ok, not bad.
What’s the interesting trend, though?
As of the end of Q2, 79% of customers were using 2 or more products, up from 75% a year ago. 37% of customers were using 4 or more products, up from 28% a year ago and 14% of our customers were using 6 or more products, up from 6% a year ago.
So they keep adding modules and products and cross-selling into the installed base.
How are those new SKUs doing?
Our newer products, excluding infrastructure monitoring, APM suite and log management, continue to grow ARR more than 100% year-over-year
Ok, but are they just milking the existing customers, or adding new ones organically?
We had about 21,200 customers, up from about 16,400 in the year ago quarter.
2,420 customers with ARR of $100,000 or more, up from 1,570 in the year ago quarter. These customers generated about 85% of our ARR.
Why not both, I guess? ¯\_(ツ)_/¯
They’re a bit more cautious on the guidance, predicting 57% rev growth for the full-year, but that sounds like caution because everybody’s being cautious right now rather than a noted slowdown taking place right now.
Time will tell, there’s definitely a consumption component to Datadog products that can fluctuate faster than purely subscription businesses (as we’ve seen in the early days of COVID19).
Transdigm, part 4 🛩
They’re not kidding when they’re calling this a deep dive… 50x just released their 4th podcast on Transdigm (5th if you count Thorndike’s appearance on ILTB), the second one with Rob Small of Berkshire Partners and a long-time member of Transdigm’s board:
I know it’ll probably be a while between episodes, but I certainly am looking forward to the next company covered by Thorndike. What a great way to bust on the scene with a new podcast! Bravo 👏
Thoma Bravo buying Ping Identity for $2.8 Billion 💵
In April Thoma Bravo bought Sailpoint for $6.9 billion (did Elon Musk come up with that price tag?), and now they’re spending $2.8 big ones on Ping… Both companies operate in the enterprise identity space. 🆔
What are they building in there? 🤔
I don’t know, but I bet Okta is *really* curious about it…
The offer for Ping of $28.50/share in cash is 63% above the closing price on August 2, but still about 25% below the peak that it reached 2 years ago, so whether the glass is half full or half empty depends on your perspective (as usual).
🔌 U.S. Burning near-record amounts of natural gas for electricity ⚡️
Record U.S. electricity consumption is driving near-record combustion of gas by power generators, ensuring gas inventories remain under pressure and prices remain high. [...]
Net electricity generation between January and April amounted to 1,337 billion kilowatt-hours surpassing the previous seasonal records of 1,331 billion kWh in 2014 and before that 1,319 billion kWh in 2008.
Note that unlike in 2020 when more gas was burned because it was so cheap (displacing coal), the situation is very different today:
Gas-fired units accounted for 470 billion kWh in the first four months, their second-highest seasonal output after 500 billion kWh in 2020, according to data from the U.S. Energy Information Administration (EIA).
Gas-fired generation is surging this year despite high fuel prices – in contrast to 2020 when the surge was driven by collapsing gas prices, which caused gas generators to run in preference to coal ones. (Source)
Gas accounted for 35% of all power output in the U.S. in the first 4 months, compared with 21% for coal, 19% nuclear, 16% wind & solar combined, and 7% hydro.
🧪🔬 Liberty Labs 🧬 🔭
VR-to-Reality, 1:1 Match 🤔
One thing that has been hard to replicate in virtual reality (VR) is the sense of touch.
Engineers are working on various things (gloves and full body suits and nerve stimulation and whatnot, I’m sure), but the low-tech way to do it is to recreate the real world inside of VR on a 1:1 scale so that whatever you touch in the real world has a virtual twin.
Of course, it limits what you can build in VR to what you have in the real world, but I’m sure the illusion must be extremely powerful.
h/t Ryan Petersen (📦📦📦)
Intel's Sapphire Rapids Xeon CPU had 500 bugs, delayed again 😬 🗓
Not that Intel really needs a win lately, but the news about its next Xeon isn’t great:
So far, Intel has released A0, A1, B0, C0, C1, C2, D0, E0, E2, E3, E4 and E5 steppings of Sapphire Rapids processor to fix nearly 500 bugs. Given that modern processors integrate tens of billions of transistors, it is inevitable that have a certain number of bugs. They are called erratas and are mitigated with microcode or even software updates. But 500 erratas seems overwhelming, as does 12 respins considering that a respin costs tens of millions of dollars. (Source)
More expensive than the process itself is no doubt the delay — Intel now targets Q1 of 2023 for launch of this new 60-core chip… Meanwhile, AMD is vacuuming up market share. That’s gotta be painful for Intel 😩
🌏 Good News, Bad News 🌎
On the left is US CO2 emissions per capita, and on the right are US annual CO2 emissions total.
The good news is that, well, it’s going down.
The bad news is that I suspect that a lot of that is due to heavy manufacturing and various energy-intensive processes moving to places like China, so we also need to zoom out and look at the whole picture to determine what the real trends are.
But I’d still rather see this than see it keep climbing in the US while the rest is going on.
🎨 🎭 Liberty Studio 👩🎨 🎥
New technique introduced to competitive NES Tetris decades into the game… 🕹
This one is kind of esoteric, but I found it interesting.
It goes to show that even for what seems like a relatively simple (if difficult) game, there can still be innovation and tweaks even after decades of high-level play.
The in-group, old-guard starts thinking alike, so it often takes a jolt from the outside to trigger new insights.
Forget about literal Tetris — what’s the “Tetris” in your life that you’ve been doing the same way forever? How would you go about it if you had to re-learn it from scratch?