Jason Dunn

Apple’s M1 Max CPU: Heat, Fans, and Silicon Voodoo I Don’t Understand

My first computer with an Intel 286 processor was three decades ago, and my last was the MSI gaming laptop I split the cost of with my son featuring an Intel Core i7-11800H. In between I’ve had various flavours of Intel and AMD CPUs, overclocked most of them, and all have behaved in basically the same way: the harder you push them, the more power they need, the more heat they generate, and the louder the fans get to keep the chips from heat damage. As the Mandalorian would say “This is the way.”

When I recently purchased my Mac Studio with the M1 Max processor (10 CPU Cores, 32 GPU Cores), it was my first foray into the world of Apple Silicon in the desktop/laptop realm (my M1-powered iPad Pro 11 doesn’t really count). I’d been using an aging Core i7 iMac workhorse from 2015 until this purchase, and it behaved like any other machine with an Intel CPU: max speed = heat + noise. h265 video exports from Final Cut Pro would take forever.

Enter the Mac Studio. It behaves in ways I don’t understand, seemingly breaking all the rules we’ve accepted as immutable over the past 30 years. For instance, when the Mac Studio is idle, the fans spin at around 1326 rpm. In a perfectly silent room, I can hear it – but just barely. At first this irritated me slightly because my 2015 iMac was 100% silent when idle. It’s certainly not loud — the fan on my Google wireless charging stand is louder — but I can hear it.

That slight irritation faded immediately when I put the M1 Max processor under load…because what I saw didn’t make sense.

What you’re seeing above on the left is me using Handbrake to transcode a 50 GB 4K MKV file to a smaller h.265 4K 10-bit file. It’s punishing to the CPU — all cores are firing hard, but temps are only up about 20%…and the fan speed increased by a mere 12 rpm. This isn’t a one-off glitch either: I’ve done 50+ MKV transcodes on my Mac Studio and no matter how hard I push it, I never hear the fan get any louder. And the total power draw while under max load? About 52 watts. 🤯

Apple’s M-series silicon isn’t the best possible maximum performance silicon out there — the top-tier chips from Intel and AMD will match or beat it in some circumstances — but on a per-watt basis, and while staying cool and quiet…Apple has done something really special with their M-series chips and I’m thrilled with my Mac Studio so far.

It’s the guns. It’s always the guns.

Here we go again. Another day in America, another mass shooting. 😢 Below is something I wrote in response to a gun owner I know thinking out loud about solutions to the problem.

People buy guns because they are afraid and believe a gun will help them when they need it. A gun in the home is far more likely to be used for suicide or to accidentally shoot a family member then ever be used to “stop a bad guy”. Self-delusion is sadly common among gun owners.

A large number of guns are stolen every year from gun owners. Gun owners are a significant source of gun supply for criminals. These law-abiding gun owners are simply arming the criminals they’re trying to protect themselves from.

Buying a gun, any gun, simply helps fuel the gun industry that lobbies against gun restrictions and causes death and destruction in this country. The gun industry, and their puppets like the NRA, are a uniquely American monster that helps kill 45K+ people a year and wounds another 40K.

No other country has the same level of problem the USA has, but every country has criminals, mental health problems, etc.

What the USA does have is guns.

Lots and lots of guns.

And yet no matter how many people are killed every year, many Americans are puzzled by the source of the problem.

It’s the guns. It’s always the guns.

Moka the Friar

After much spiritual searching, my dog Moka has decided to begin her monastic life: I present to you Friar Moka.

(I don’t tend to post things like this often, but putting a little joy into the world is never a bad thing.)

T-Mobile 5G Home Internet: A Complete Failure [Part 1]

The headline already tells you where this is going, but come with me on an August 2021 journey that started with hope and ended in despair.

By way of background, I have two choices for home internet service where I live in a suburb of Renton, Washington: CenturyLink or Xfinity. Because I run a Plex server at home, and do a lot of video/photo uploads, having decent upload speeds and no data cap is important to me. Xfinity is off the table as even their fastest plan in my area (1200 mbps for $80/month) still has a 1.2TB data cap and the upload speeds are 35mbps (and Xfinity hides this fact). Plus, Comcast is the devil 👿 and I don’t want to give them my money.

So for 10 years I’ve stayed with CenturyLink’s DSL-based product. My community of 38 homes is too small for them to run fiber to, so I’m stuck with 80mbps down and 40mbps up. I long for more competition in the Internet provider space and it’s frustrating to me there isn’t more choice — though I often remind myself at least I have two options, many people in the USA have one or no access at all.

On the plus side, the CenturyLink connection is very stable, costs $50/month, and there’s no data cap. 80mbps is generally enough for my household, though I of course long for something better — and when I received an email from T-Mobile sharing that their 5G home Internet service was available in my ZIP code I signed up immediately…after I confirmed I could cancel if it didn’t work because I was hugely skeptical of their claims.

Why? People using TMO that come to my home get very weak signal. In early 2021 I tried a TMO mobile WiFi hotspot and had terrible signal and speeds. So it was with a large grain of salt I took TMO at their word their 5G signal was going to work great in my home. At the time, their speed claims were a minimum of 100mbps up and down, with the reps telling me verbally I should speeds around 300mbps. This was enough to lure me into testing it. It’s worth noting their web site now says typical speeds are 35-115mbps which is a dramatic downgrade to their initial marketing.

What the TMO site now says about speeds.

Their sign up process was incredibly invasive – at one point I wondered if I was being phished because they were asking for so many pieces of ID and information.

It took a few months for me to get the combo modem/ WiFi router (more demand than supply), but when it finally showed up I was excited to give it a try. Setup was fast and easy; after a couple of minutes I’d booted it up, applied a firmware update, and was ready to test it. I put it in the window of my main floor office and was immediately concerned, but not shocked, then I saw I only had two bars of signal.

Signal strength and speed are not linked though (despite what you might think), so I ran a speedtest on my phone connected to the TMO router. The results? 😩

This isn’t the 5G speed I was hoping for.

10 Years in the USA: Looking Back

10 years ago today, my family and I drove across the border, leaving our life in Canada 🇨🇦 behind and embarking on a new adventure: working and living in the 🇺🇸United States of America. It’s gone by so quickly, my brain can’t quite grasp it’s been a full decade of my life. It was such a scary 😱 decision at the time — leaving behind our friends and family, moving to a place where we knew no one, and where we’d be starting over from scratch. Keep in mind that neither my wife nor I moved away from home to go to college: our entire world was in our hometown of Calgary.

So much has happened in the past 10 years! Some of the highlights that come to mind, more or less in order, are:

Getting to work at HTC and having my first real corporate job (before that it was just vendor contracts with some big companies, never full time employment). Special thanks goes to John Starkweather for trusting that some guy from Canada was the right person for the job (even when the job ended up being a different sort of animal). Thanks also to Jason Gordon for his wise words as I was thinking through the big decision to move.
Being able to build a new home with space for people to come visit.
Having the joy of creating HTC elevate, traveling the world and meeting some truly fantastic HTC fans. That community program and the people in it hold a special place in my ❤️. #IBleedGreen
Working with some truly fantastic people at HTC, some of whom are still my friends to this day.
The birth of my daughter Alanna, who’s a real spitfire and destined for great things!
Learning the valuable lesson that green card sponsorship should have been part of my HTC hiring contract. 🤷‍♂️
Getting the opportunity to work for John Starkweather again, taking a contractor role with K-Force to work for AT&T and get my green card (and so did my wife and son).
Learning that I could step up into big roles and take on leadership where there was a gap; that being the one who says “Yes, I’ll help” is a great way to learn and grow.
Using my role at AT&T to move through a variety of experience-expanding roles such as learning Adobe Experience Manager and other tools.
Discovering a great school for Logan that grew into a career path for my wife and a great school fit for my daughter.
Experiencing the totally new scenario of having to earn a job interview, and failing to get the role; reinforcing the importance of interview preparation and confidence in telling my story to an employer.
Watching my son grow from a little guy who fit in my arms to a kind-hearted, creatively brilliant kid who’s getting taller by the day!
Once again learning that truth in the maxim of “It’s not what you know, it’s who you know”; leveraging my network and past recruiter contacts to open the door to a role at AWS that was absolutely perfect for me. I had to work very hard to get through that open door mind you…
Being reminded that relationships matter; building and maintaining a professional network is critical to continued success.
Jumping into the developer community space with my role at AWS, something I knew nothing about, but realizing that people are the same no matter if they can code Javascript or not: everyone wants to belong to something special and online community has been in my blood since I first heard the screech of a dial-up modem.

Looking back on the past 10 years, I feel a tremendous sense of gratitude for the blessing that have been showered upon my family and I. I’m so glad that every other door in Canada was closed to me, and that Ashely and I had the courage to leap through the open door that led us to this life. Here’s to the next 10 years!

The EcoFlow DELTA 1300: The Most Badass Battery Ever*

In 2019 I backed a crowdfunded product called the EcoFlow DELTA. While it was promoted as a “battery powered generator”, the name was misleading: fundamentally a generator creates one form of power by consuming another, and this was a battery that stored and outputted electrical power. It’s a battery. A really, really big battery: 31 pounds and 1260 Wh of power to be exact. It has six AC outlets supporting up to 1800 watts of output, 3300 watts of surge protection, pure sine wave output, four USB-A ports (12 or 18 watts per port), two USB-C ports (60 watts per port), and it charges via AC power, solar (up to 400 watts input), or 12V car adaptor.

I backed the project for $799 after we had an 18 hour power outage at my home and I found it frustrating how many things wouldn’t function. I was looking for a specific solution to allow us to continue using our on-demand hot water heater, which uses natural gas but requires electricity to operate. When the power goes out it’s relatively easy to create light and bundle up if you’re cold, but the immediate lack of hot water is an uncomfortable problem for a family with two kids. I had a quote on a natural gas-powered generator, but the $6000 price tag was too high for the rarity of the outages. I’d need to lose power several times a year for 20+ hours each time to justify that expense.

The EcoFlow DELTA arrived in January 2020, and it exceeded all my expectations. I took a bunch of photos because I thought I’d write up a long, thoughtful review of it…and didn’t. That review never quite got written, but I had all these pictures and a few thoughts I wanted to share, so here’s a photo essay of sorts for anyone interested in the DELTA.

* This was the most badass battery that EcoFlow made in 2019, but in mid-2021 they release the monstrous DELTA Pro, a 99 pound battery with 3600Wh of power! 🤯 But it’s also $3599, so…🙃

The 1260 watt-hour battery charges directly from AC wall power, and it charges fast: pulling over 700 watts from the wall power, it will go from 0% to 80% charge in under an hour (often all the way to 100% in 60 minutes). It arrived mostly charged. An audible fan kicks on when it’s charging to keep the heat down.

Chromecast with Google TV + ESPN+ = Usability Nightmare

I admit it: I’m an armchair product manager.

Every time I use a new product or service, I either applaud it or I’m critical of the user experience. Often both! I wrote product reviews on various tech web sites (mostly my own) for ~15 years, and when I worked for Spb Software I took on the role of a product manager for Spb Imageer, so I’ve experienced both sides of this coin to some extent (though much more on the reviewing side).

Working at HTC also gave me interesting opportunities to learn more about the decisions that go into creating hardware and software. I understand every product is a series of trade-offs; most teams don’t have enough developers to build things they way they wish they could, and timelines are never quite long enough to fit in every feature and testing.

But…

Sometimes product managers and UX designers will make such inexplicably awful choices, you have to wonder what they were thinking. You also have to wonder if they tested with actual customers in real-world use, or if it was never tested by anyone other than an internal QA team with a checklist and no knowledge of real-world use. The ESPN+ app on Google TV is one such app.

When I bought a Chromecast with Google TV late last year (what a mouthful of a product name!), I was genuinely excited about it – this was the first truly new execution of Google’s Chromecast platform since the first one launched. I’ve done a fair amount of tweeting about my impressions of the hardware/software from Google – I wish Twitter had a better search function, but here are a few – so this blog post is focusing on one very specific scenario: how utterly terrible the Chromecast with Google TV is for watching long-form content on a poorly designed app. Walk with me through this real-world scenario…

Epson-ET3760 Review: Muted Colours, Awful User Experience, but Affordable Prints

The Epson-ET3760 is a decent printer, but it’s hobbled in a few ways that keep it from being an exceptional product. It was more a bit more expensive ($279 + tax on sale from Costco.com) than a comparable HP printer, but Epson makes less on the ink so it’s expected that you’ll pay more in hardware costs. It’s a “pay extra for the razor because you won’t need to buy as many blades” scenario. The bundle I bought from Costco included two extra black ink bottles, so I expect to not need ink for several years. Epson touts costs as low as 1 cent per ISO colour page.

The print quality is crisp – no complaints there. Compared to my HP though the colours are muted and don’t pop as much. Colour accuracy is significantly off as well. Red is more orange, blue is more grey, yellow is more orange. This means, unfortunately, that all that cheap ink you’re getting doesn’t measure up to what you get on an HP printer.

LECTRON Tesla CCS Adapter: First Impressions

My family and I have a road trip to Canada coming up soon, and for the first time I’m going someplace — a small town in BC, Canada — where there aren’t Tesla Superchargers within range once I arrive. Even with a full charge from an overnight L2 charge, I can’t make it back onto the Supercharger network to continue my journey to Calgary (due to elevation and the fact that I bought the Model 3 SR+ rather than the LR which had 70 more miles of range). 🤷‍♂️

So I’m going to head east, and along the way there are no Superchargers, only CHAdeMO and CCS L3 chargers. Typically charging at around 32 miles per hour of charging, L2 chargers are too slow for a road trip unless you really don’t care when you get there. I need a way to connect my Model 3 to one of these L3 charging standards, and because:

CHAdeMO is an old standard, being rapidly replaced by CCS
Tesla doesn’t have their own CCS adaptor sold anywhere but South Korea (!!?)
The Tesla CHAdeMO adaptor is sold out online

…I figured I’d buy the LECTRON CCS adaptor off Amazon that, at the time I ordered it, had zero reviews. And it was $600. 😮 What could go wrong? 😜

This is a simple, bullet-point first impressions post meant to get something online fast, not a detailed, well thought-out review. I have too many of those already sitting in my WordPress drafts folder. 😉 So here we go…

I tried to update the firmware before using it, and not only is it Windows-only (no macOS option), the updater just crashed and gave me an error over and over. Not a great first impression.

It has to be used in a particular sequence (powered on, wait until green, connect the adaptor to the charger plug, wait 10 seconds, then connect to Tesla). The first charger I went to, I didn’t follow that sequence exactly, and it failed so I thought it was the EVgo charger and left. This was after two phone calls to EVgo, them rebooting the whole charger, etc. We tried everything. I did charge via a Tesla plug at that same charger, so we deduced it was probably the LECTRON CCS adaptor that was at fault (but really it was me that was at fault…classic PEBCAK error even though I was standing). 🙄

It’s comically huge and surprisingly heavy. It takes a reasonable amount of strength to connect and disconnect it, and it’s clumsy as hell. A smaller person with smaller hands may find it very challenging to use.

I’m concerned about the amount of pressure it puts on my M3. The weight of the adaptor + the CCS charger, all pulling down at a particular angle, puts a lot of pressure on a very small part of the car. I have no idea if Tesla engineered the charging port for that kind of weight.

I went to an Electrify America charger, followed the right sequence this time (RTFM!), and charging started. Oddly it stopped about four minutes, no explanation. It could be because I started multiple charging sessions back to back – it told me to move my car each time – and the charger was confused, but I’m not sure.

For the first time ever, I was a bit envious of other EVs I saw there that could just plug into these chargers without adaptors.

It charged at 42kW. That was giving me about 181 miles of charge per hour, which is 5x more than I can get from my L2 charger at home. Good stuff. That’s far slower than a Tesla Supercharger, but fast enough for reasonable road trip times.

After the successful partial charge, I went back to the EVgo charger to see if using the correct sequence would result in it charging. It did. I got 11 minutes of sustained 41kW charge before manually disconnecting. No issues.

The next test is to find an L3 CCS charger that puts out more power to see how much power the adaptor can take and if it’s stable (according to LECTRON, it’s limited to 50kW on my M3). That’s my big fear taking this thing on a road trip: that it will just flake out. I’m thinking about paying over market value for the Tesla CHAdeMO adaptor just to have something I *know* will work instead, though I consider it a bit of a waste of money given CCS is the future. I don’t know anyone with the Tesla CHAdeMO adaptor so I can’t borrow one.

I’m irked that Tesla doesn’t have a CCS adaptor for us owners and we’re having to use these third party adaptors in the first place. Why aren’t they selling a first-party adaptor in North America?

It uses microUSB to charge, which in 2021 for $600 is ridiculous. It should be USB-C. But also it’s one more thing to keep charged, using a deprecated connection standard. I also can’t help but chuckle at the idea of 40kW+ of electricity flowing through this adaptor, but it won’t work unless you plug a tiny microUSB cable with a trickle of electricity to charge it before hand. It’s probably too complex to siphon off the incoming charge to power the adaptor itself, but a guy can dream. 😉

So will I keep this or return it? I want to do more testing. It’s not a great product, that’s for sure, but it may be good enough for what I need.

Is Tesla’s Market Share Limited Due to Competition?

I was having a WhatApp chat with family member about Tesla and competition from other EV makers. I figured it was worth sharing here.

This is a fun topic. 🙂 Globally, EVs as a share of the market is in the low single-digit %. Likely around 1-2% (edit: it’s 3%). Some countries like Norway it’s much more. So as more car makers bring more EVs to the market, it’s not about them competing only with Tesla, it’s about the whole market making the shift. There are going to be billions and billions spent transitioning from gas to EVs in the coming decade: cars, trucks, vans, etc. The weight and cost of batteries has, in the labs, hit a point where battery powered planes are within reach. The single biggest problem every EV maker has is battery supply. So Tesla will be one of many companies with products to take advantage of this buying wave. It’s a BIG pie.

Because the big car makers are so many years behind Tesla, it seems like it’s “Tesla against everyone”, but it’s not – Tesla will be one of many options in the EV market. Just like there’s no expectation that one gas car company would ever be able to have 100% of the market, neither will Tesla. They just control a large % of the EV market as it stands today, but the EV market is so tiny still globally.

Think about Toyota. In 2020 they were about 9% of the US market in sales. You’d never think Toyota was a failure as a car company – you see their cars everywhere – so let’s say Tesla “only” sells 9% of all EVs in the USA. That would still be an incredible business! I personally think they’ll control 2-3x the market share of Toyota if they can keep going.

They’re like Apple from a brand perspective. Few people are buying a Nissan Leaf because they want to; if they could afford a Tesla they probably would buy one. Tesla has never paid a cent in advertising. It’s really hard to overstate how much of an advantage that is. And if they can get a $25K car into the market by 2022-2023, that will open up a new market for them.

Remember Tesla is fundamentally an energy company trying to get the world off fossil fuels and they chose to make cars their first product to show people that battery-powered vehicles could be exciting and efficient.

Share:

Share:

Share:

Share:

Share:

Share:

Share:

Share:

Share:

Share: