> "I also supported cloud computing, participating in 110 customer meetings, and created a company-wide strategy to win back the cloud with 33 specific recommendations, in collaboration with others across 6 organizations."
Man people keep count of this stuff?! Maybe I should too, it does make flexing easier.
While I have a personal career document and have had one for years where I have all of my major accomplishments in STAR format. This seems a bit much.
When I was at BigTech, there was an internal system where you recorded your major accomplishments and the impact they had.
But I would never write it up on a public blog post like this. I am assuming the author of the post must be someone well known in the industry for it to make it to the front page of Hacker News. If his intent was to promote himself so he could get another job, I’m sure that he has a network where a few messages would lead him to one.
Even in my little niche of the world where in the grand scheme of things I’m a nobody, I was able to lean on my network at 50 after being Amazoned in 2023 and have three offers that were at least a lateral move within two weeks.
I had one fall into my lap last year too that I accepted based on my network.
Do you have a particularly easy way to track or are you kind of doing the same thing as consultant and logging your dailies? Always drove me a bit crazy having to do that admin piece every day.
Contrary to the usual opinion on HN, this provides a good reason to do an MBA!
You should learn enough economics that if you are even a bit insightful you will avoid Econ 101 thinking, you will learn about things like intrinsic and extrinsic motivation, and a lot of other things relevant to management.
Damned is this industry, when even _you_ say you have to show that "remoteness works".
I also measure meetings (counts, lengths, and mostly meeting minutes/outine jotted down by myself) and keep track of other metrics, exactly for this reason. However, I also don't happen to have written best selling books and stuff, so I really must do this, and you really shouldn't have to :-)
If only I had known that in the past, I even once received the completely wrong advice to "not stand out, since your work will speak for itself and you will get recognition".
(Fancy US tech companies like to be very selective, have a competitive mindset, hire "the best" according to their filters, and then want people to show how amazing they are, uu, so much impact, woah... and in effect people need to constantly manage upwards.
While in many other companies, or "orgs", having a good team cohesion is more important. To blend in a bit, get accepted even if it means foregoing some ambition.)
It’s more that it takes so long to get anything done, the effort and results need to be recorded because it most often won’t be obvious from the impact. It’s hard to make a splash on a production system maintained by 30 other people, but you can usually make things better, but it won’t always be obvious.
I guess they don’t know how or don’t bother to evaluate people on what they actually contribute? Just number of meetings attended, number of tickets closed?
Joe implemented feature A. Sandra implemented feature B. Raj implemented C. All launched in July. Since then metric X is up 20%. Who gets credit, and what does that credit really mean?
Now say all 3 did that in 3 different products. One produced a 200% improvement in an internal product, one a 40% improvement in a product with thousands of users, one a 1% improvement in a product with a billion users? Compare *that*.
200% improvement may just be the result of feature A and 40% may just be the max performance gain from feature B. Comparing developers over the effectiveness of features they implement is mostly rating the PMs or the leadership they work with. Its a dice roll and its a measure of one's luck to be at the right place & right time to work on the right task.
People complain about using metrics. People complain about rating performance based on what your manager or coworkers say about you. Performance reviews are an unsolved hard problem.
Not the OP, but my proposal is to acknowledge that unsolved hard problems are... unsolved. Instead of inventing bullshit and pretending it's constructive. Some people make a career out of that bullshitting, others complain about it because it actually has a negative impact on them.
"What's your proposal", in my experience, is often used as a defense against someone calling bullshit. My proposal when I'm calling bullshit is that the bullshitter start being professional, but it's not exactly something I can say.
Keeping track of actual value would require actually rewarding people proportionally; all jobs ever only really care about how often you're on time or your meeting attendance record.
Rewarding people proportionally is a macro-level unsolved problem. Kropotkin wrote it about it and his solution was to throw his hands in the air and say fuck it, labor value is impossible to accurately evaluate, and thus he invented anarchist communism.
Just look at all the weird quirks our world does to labor value: the same exact job in two different locations for a global employer (say, Google), selling to a global market, pays differently depending on "local labor market prices." In 2025 for engineering what on earth is a "local labor market?" An optimization coming from an engineer in Taiwan saves you the same money as if it comes from an engineer in SF but the SF engineer gets 8x the reward for doing the work. Luxury goods and electronics cost the same in both places. Buying property is only slightly cheaper in Taipei vs sf (yes really), vehicles cost more in Taiwan. Food and healthcare is cheaper in Taiwan, and that alone I guess means the Taiwanese engineer is worth 1/8th the SF engineer, to make sure the sf engineer can afford 16$ burritos?
Many other quirks. You point out another one: labor often isn't rewarded based on real value to a company, for many reasons but one of which is that managers often don't understand the job of the people they're managing and so apply management relevant KPIs to disciplines where those KPIs don't make any sense. Engineering, for example, doesn't correlate actual value add to the company via meetings attended or customers met, but that won't stop management from applying those KPIs and thinking it does!
I'm torn between thinking we keep things this way out of ignorance vs we keep it this way maliciously so the management class (which sets the rates) doesn't get written out of labor agreements altogether because they're often useless vs if we didn't keep up this charade, capitalism would just collapse entirely.
Agreed. There's the additional point that I think many people don't appreciate, which is that those managers and many people lower down in the org chart merely exist because somebody else needs to be responsible for a system or a liability regardless of whether they do anything measurably profitable, and aren't necessarily incentivized to do anything more productively; they're just there to take care of it or be blamed if it's not, and have a low ceiling for what that job can possibly be worth with no measurable way to argue for more, and so in the case of managers, try to invent clout-generators at any cost and with no connection to how the assignees might accomplish it.
> An optimization coming from an engineer in Taiwan saves you the same money as if it comes from an engineer in SF but the SF engineer gets 8x the reward for doing the work.
Suppose you have a thousand engineers and those thousand engineers generate ten billion dollars in annual profit. How much do they each get paid? They amount they're worth? Nope, the amount they'll accept.
If you live in the US and you have the wherewithal to be an engineer then you could also have been a doctor or a lawyer or some other high paying occupation. And many of those can't be fully remote because they have to see domestic patients or interact in person with local courts or clients. Which means that if you want someone in the US to be your engineer, you need to pay them an amount that makes them want to do that instead of choosing one of those other occupations. Whereas the one in Taiwan doesn't have the option to become a doctor in San Francisco and is therefore willing to accept less money.
So why don't companies just hire exclusively the people in Taiwan? There are all the usual reasons (time zones, language barriers, etc.), but a big one is that they need a thousand engineers. So they and their competitors hire every qualified engineer in Taiwan until Taiwanese engineers reach full employment, at which point the companies still don't have all the engineers they want. And when the average engineer is making the company ten million dollars, paying San Francisco salaries is better than not having enough talent.
So then why doesn't every smart person in Taiwan become an engineer? Because the companies hiring engineers there are only paying Taiwanese wages, and then they're not any better off to do that than to become a doctor or a lawyer in Taiwan. And if they would pay higher wages there, the local economy would have to start paying local doctors and engineers more to keep them all from becoming engineers, and then you would only get a modest increase in the number of engineers for a significant increase in compensation. Which is still what happens, but only slowly over time, until the wages in Taiwan ultimately increase enough to no longer be a competitive advantage. And companies don't want to make that happen faster because then they'd have to pay higher salaries in Taiwan.
> An optimization coming from an engineer in Taiwan saves you the same money as if it comes from an engineer in SF but the SF engineer gets 8x the reward for doing the work.
An underappreciated difference is that it's hard to schedule meetings between people in SF and Taiwan, because of time zones.
I know some of them do this, but ours doesn't. There is a once yearly self-review, and as far as I can tell it has literally no impact on your actual performance review and compensation, which are basically entirely up to your manager's observations of you.
So it is important to keep your manager informally up to date on what you're doing, at least during the weeks they're thinking about performance.
If you look at many of his recent blog entries, it is clear he has felt the need to quantify his impact to prove he isn’t less effective as a remote employee in Australia working for a company in the US.
Intel's management did not appreciate (as likely did not understand) tech skills/talent lately, which likely contributed to them squandering their lead.
I mean maybe. We often have weekly customer meetings. One of my programs has 2 customers, we meet with both weekly. So do I put idk 200+ customer meetings? That seems like a weird metric because it's like "compiled code 400 times." I've seen resumes that have the same vibe. We did not hire them. Sometimes it's very telling what people think are accomplishments.
The fact that they were busy keeping count during those 110 occasions and for every other activity clearly tells that they nothing better to do. You have to be loud about such numbers when you have very little meaningful work to show for.
A "goodbye" post after only 3.5 years. Hard to relate.
In my world it's hard to imagine an impact after that short of a time. And in fact, reading the list of accomplishments ("interviewed by the Wall Street Journal") makes it clear it's a good PR piece.
I'm perfectly willing to believe he's fabulous, but this didn't move the needle for me.
Clicking through his links to various posts about e.g. stack pointers or flame graphs, my takeaway is he's an outlier in productivity, and got a lot done in 3.5 years at a monstrously large organization.
I'm pretty envious of his capabilities, in 3.5 years I can ship a couple webapps, I would never personally get JVM compilation flags added.
You are tired, but you still read, still commented (or worse, commented but didn’t read).
I always give people benefits of the doubt. He posted it on his personal blog, going back so many years. Most of his content are technical in nature, the kind of things that would never be on the front page of HN.
For other people, they're going to be thinking "some other company is going to get one of the most effective and impactful performance engineers on the planet".
Dude shipped flamegraphs (which he also created in 2011) for cloud GPU loads and persuaded internal stakeholders to release the code as open source.
The "interviewed by the WSJ" line is for managers. Reading between the lines, I'd say he did really well and, if he didn't do better, it's because the organisation didn't let him.
> if he didn't do better, it's because the organisation didn't let him.
The last few sentences to me read like he knows for sure that the organisation is actively working against what he sees as his important goals. Carefully worded (and likely personal lawyer approved) to avoid burning the bridges as he mic-drops and deftly avoids having the door hit him in the arse as he struts out.
I felt like he avoided saying anything negative about Intel just in case it would be taken that way. Intel doesn’t have the best reputation so we are all interpolating a much more negative message than he actually said.
> The "interviewed by the WSJ" line is for managers.
It’s a green flag for hiring managers for sure. Even a lot of valued employees wouldn’t be allowed to represent a big company to the WSJ for various reasons, even with a PR person sitting next to them.
Good at self-promotion == just good in most cases for most practical purposes whether it's factual or not, arguably. His books seem substantial though, I don't know many people who've read or written 800 pages on system performance
> Good at self-promotion == just good in most cases for most practical purposes whether it's factual or not, arguably.
This does not seem true to me. Most popular programming YouTubers are demonstrably great at self-promotion but tend to be mediocre or bad programmers who know very little, even about the topics they talk about.
If anything we have plenty of examples of where being good at self-promotion correlates inversely with actual skill and knowledge.
With that said, I wouldn't classify Brendan Gregg as being good at self-promotion.
Flamegraph is literally just a perl script that visualizes the stack traces collected by perf/dtrace (kernel). It's a good tool though but it doesn't need to be oversold for its capabilities, the hard work is done by the kernel. And honestly, many times it is not that useful at all and can be quite misleading, and not because of the bug in the tool but because how CPUs are inherently designed to work.
If you've been there 16 years, I'm sure you employer feels your impact has been worth the investment. Are you really saying that you don't feel you have made the impact you would have liked to make? Do you feel under-utilized?
That's interesting; I feel like like it's the opposite: What used to be great work is basically unfathomable today and what used to be regular productivity is seen as almost superhuman. People get almost nothing done nowadays and I've never felt like expectations were ever really at the level they ought to be at, especially with how much money people are getting.
He doesn’t mention it in this post, but in another post he talked about the toll of needing to frequently attend meetings in the middle of the night in his time zone.
Whatever his reasons for leaving, I hope that he finds a better balance in his new role.
This was the takeaway I had taking to a colleague about his time at Intel - they're a genuinely global company with engineering teams in practically all time zones who are still expected to collaborate with each other. No matter what time of day the meeting was scheduled for, it was the middle of the night for somebody, and no, just working on written docs async for everything didn't cut it, and they couldn't just fly people out all the time. So that's apparently just part of what it means to take a job at Intel these days.
On modern AWS instance types so much is offloaded to dedicated hardware that the only shared (noisy) components between VMs is memory bandwidth and higher levels of CPU cache (and I think graviton doesn't even share CPU cache now)
I would suspect your performance difference is mostly likely showing that on metal you are sharing the same software wider so not polluting caches as much as a vm neighbour running unrelated software.
The virtual instance is the exact same size as the metal one, which covers an entire physical machine - I guess this is pure overhead rather than noisy neighbors.
That's a surprisingly large overhead. I've not measured that large an impact on AMD, particularly for compute heavy.
Did you profile at all? And have you observed if it's not compute-bound? If it's memory or IO bound it can be due to other virtualization overheads, such as memory encryption.
Calling them AI flamegraphs is really naming them after the workload they are likely to be used on. If you want to make workloads more efficient it’s useful to know where they are spending their time.
Flamegraphs are great, GPU flamegraphs are an obvious good idea. But his choice to work on "AI" instead of general GPU compute, and advertise this "AI" work at the very top of the list of things he's done of note, tells me what I need to know.
Masterclass in turning a goodbye email into a hire me after my next gig ends. I’m not being sarcastic, this is a great example of highlighting the value they added.
If my back of the envelope math is right, in the last 6 months he’s been attending more meetings at possibly odd hours; he lives in Australia and Intel is based in the USA.
Two different numbers, no? The resignation posts specifies 110 customer meetings, the blog post you linked to about meetings during odd hours does not.
Yeah, different numbers, 110 customer meetings, the other post tracked 1-6am meetings. I'm glad I tracked 1-6am meetings since I've shared that number when people think that remote workers aren't making an effort.
I really have no idea how IBM is still in business, or the other big toxic techs like Oracle and Salesforce. Just goes to show I don’t know as much about the industry as I think.
Oracle basically runs HR and finance services for like every large company in Europe. They also run a scary amount of healthcare stuff and other government tech type stuff.
It sucks but I see why they do it. If you don't have the technical/managerial talent to handle procurement then it's the safest bet.
They bought Red Hat, which has OpenShift and all their other "DIY Cloud" bits. This stuff is popular in government or old businesses that may have been slow to (or unable to for regulatory reasons) jump to AWS/GCP etc.
To say nothing of the banks and others still using the IBM big iron.
The American hyper scalers are not necessarily the place to be. Modern can mean Non-hyper scalar as well. Can this sentiment just die please? Great that its working out for you and you replaced good sysadmins with aws admins, but it should not be the default strategy perse.
This is true economically but in reality if you have much larger cost savings than that for sale then these companies mostly say "we would be happy to buy that for $0 while we pay you a million a year to move to the united states"
His bio says he was an Intel Fellow, which is like a VP-level individual role, and yes that's what I expected too… but apparently not? These are kinda low.
Id expect his comp even before Intel to be way above that (he came from Netflix), perhaps levels info is not entirely correct for Intel or doesn’t apply to exceptional hires, fellow level compensation at FAANG seems to be more accurate there though
Yeah, behind datasette it looks like there's C64 C parked, and above is a laser 300 (which makes sense if guy is australian) and we can also see 1541-ii behind that, on the top.
Right, laser 300 was called the VZ300 here. I'm out of desk space so I had to put the VZ300 on a stand above my C64C. Maybe AI can finally help me code some C64 and VZ games. :-)
If only! It's kind of a blessing and a curse for us who still code for c64 (demo scene). It looks like llm may help you, but it's usually gibberish 6502 asm. I've seen similar with z80 but on spectrum.
I mean I understand if someone like Keller writes such posts but some dude claiming to have hosted conference events and some kind of process flame graph which could have been done by anyone…
Maybe you should read something about him before you call him that. I recommend the "Contributions"-section on his Wikipedia-article. And if it is of any relevance to your work: his "Systems Performance: Enterprise and the Cloud" is a comprehensive and excellent guide.
Man people keep count of this stuff?! Maybe I should too, it does make flexing easier.
reply