Hacker News Reader

Comment by ChuckMcM | Hacker News Reader

Here's a fun fact (and I suspect shows what a loser I am :-)) Pat Gelsinger and I were colleagues at Intel in the early 80's. We both worked in what was called "MIPO" (Microprocessor Operation) but I was in Systems Validation and he was on the customer facing side with field engineering. My career went off through a variety of engineering roles, his went into management and up. When he joined EMC, I told him I had always thought he would have been Intel's CEO. He told me, "Hey, maybe I still will be :-)" That was like 10 years ago, and here he is.

There are a lot of things you can say about Silicon Valley, but one of the more interesting aspects of it for me has been how "small" it is in terms of people have an oversized impact. I have never been one of those people of course, just a part of the entourage. But it has been interesting to watch and learn from folks who are good (and bad) role models.

View on HN

Pat went to my high school and computer science program. It’s in a small rural, Pennsylvania farming town - pop 2000. Somehow we got an amazing math teacher turned CompSci and he turned out world class students - we always ranked high enough each year to compete internationally in ACSL.

A few years back when the teacher retired, we all gathered together to thank him, including Pat! World-class guy, and makes us proud! So great to share a story with students, who come from nothing, how you can leave our hometown and eventually be the CEO of Intel.

Wishing Pat the best of luck in his new role!

A single good teacher can make all the difference in the world.

Which is kind of sad, isn't it?

Doesn't that mean that a lot of people will never be on the receiving end just because they did not win the teacher-lottery (arguably the chance of getting such a teacher is rather slim)?

Shouldn't our systems be more robust? (Not talking about normalizing these effects down, but creating a better environment for everyone, using these effects en masse)

We must pay teachers on par with Senior Software Developers at least.

We need our very best to go into teaching. It's the ultimate multiplier role.

I hear my wife and her friends constantly talk about XYZ job that makes more than they do and it's very demoralizing to them. I can definitely see why there aren't many passionate teachers. My wife has been teaching 6 years and is just now at the $42,000 mark. The state health insurance is decent at least.

From an economics perspective, it's kinda insane that teachers are paid so poorly, since they are such force multipliers. A single brilliant teacher can positively influence the future educational trajectory (and future earnings) of thousands of kids. Investing in making that happen more often seems like it would have great dividends.

The problem, in the US, anyway, is more that either they can't be paid differently or that the bad ones can't be fired and the new ones paid more and held to higher standards. The cost per student is already quite high.

Nations like Finland and Poland have figured out how to make that change and seen their students flourish; hopefully we will too.

I have a few former colleagues who were equals (or close) with me early in my career, and are now senior executives at Fortune 500 companies, and it does always make me feel kind of like a loser!

But also I think about how much crap they put up with in those roles, and how relatively low-stress my life is, and then I feel better about it. They're cut out for that kind of job, and I most certainly am not.

If I may ask, any pointers on how is the stress level in your role currently/over the course of years and what is the difference in your compensation and your colleagues compensation cumulatively?

This question makes me groan.

“TC, bro?”

Not the GP, but I'd guess the difference is an order of magnitude for both stress and comp.

I see a remarkable number of people who have good personal experiences with Pat but I've got this concern.

Intel's technical failure with 10nm has gone hand-in-hand with financial success with 14nm. That is, without 10nm chips on the market in a meaningful way they've been able to raise prices for 14nm parts -- Intel put up better financial numbers than ever in a time when it has not been investing in future success.

VMWare was a big thing in 1998 but it was obsolete by the time Pat got involved -- a hypervisor is naturally part of the kernel and there is no way cloud providers are going to spend their margin on VMWare. Yes, many people in business are terrified of open source software and want a proprietary product so they have somebody to sue (VMWare) or they need somebody to hold their hand (Pivotal.) Either way, VMWare and Pivotal are units that can be merged and spun off whenever a company based in Texas (Dell) or Massachusetts (EMC) wants to look like it has a presence in San Francisco -- you see the vmware logo on CNBC every morning and somebody thinks the king is on the throne and a pound is worth a pound but that doesn't mean anything in the trenches.

Like Intel in the past 10 years, VMWare is entirely based on a harvesting business model. In the short term Intel made profits by pandering to cloud providers; but in the long term cloud providers invested their profits in better chips. (What if Southwest Airlines had developed a 737 replacement designed from the ground up for a low cost airline?)

Pat might be able to keep the game of soaking enterprise customers going for longer, but someday the enterprise customers will be running ARM and the clients will be running ARM and the coders will be thinking "did they add all of those AMX registers just to put dead space on the die to make it easier to cool?" and falling in love again with AVR8.

Speaking as an ex-Pivot (but only for myself), a couple of thoughts:

- VMware grew almost 2.5x under Pat. Obviously their core franchise is about harvesting, but new adjacent products like NSX and VSAN helped a lot with growth. They just took a long time to get traction given the customer base, who don’t like change. VMware Cloud on AWS has been surprisingly strong even to skeptics like myself.

On the other hand, companies would get rid of VMware if they could, and they tried and failed with OpenStack. The hypervisor is commodity, but the overall compute/network/storage private cloud system isn’t trivial. Turns out “holding hands” whether by software or services is pretty valuable?

Public cloud of course is a substitute, though private clouds and data centres have survived and thrived as well (for now).

- Pivotal had a boutique software and services model that would be difficult to scale as a public company given the current shifts in the enterprise fashion away from productivity-at-any-cost (PaaS and serverless) and towards perceived low cost building blocks (Kubernetes and its ecosystem).

But it would be an gross oversimplification to suggest this was merely a vanity project for EMC and Dell. It took in $500m+ annually on open source software (and another $300m in services), which is no small feat, though still minuscule given what surrounds it. But anything new has to start somewhere. Not enough to impact the mothership’s balance sheets, but there was something special there that could be nurtured. Whether it can be, or whether the differences matter enough is anyone’s guess.

Pat could take Intel in a direction we don’t anticipate. VMware was written off for dead when Pat came on but he managed to buy it another 10-15 years and at peak a doubling of the stock price. He’s probably learned from that.

This is spot on. And I think the TL;DR of the hypervisor argument goes as follows:

Hypervisor is a commodity, however management and support of hundreds or thousands of them is not. You can either pay people to support them and fix the software when it breaks or you can pay <vendor name here>. Given the former requires expertise and planning it's often more cost effective to go the latter.

Disclaimer: I'm employed by VMware (less than 1 year) and chose to come here based on pivots I feel they are making.

Any thoughts on what made him special (either from your perspective or from those choosing to promote him)?

Keeping in context that I've not worked with him in (gulp, 30 years!) at Intel he was one of the folks who put problems into the whole picture. So a customer would say, this chip doesn't work, over at Systems Validation we would get hopefully enough information to re-create their problem on an Intel built board, and Pat's job (at the time) was to co-ordinate between us, design engineering, and marketing to figure out how to tell the customer to proceed.

The "actual" problem could be anything from the customer misinterpreting the datasheet (Marketing/Comms problem), to insufficient testing (Factory/Production problem), to chip function (Design Engineering problem). As a "new college grad", or NCG in the lexicon, I admired how he dug out details from various folks to get to the real problem. He always had ideas for things Systems Validation (SV) could do to maybe trigger the problem that made sense to me. He really embodied the philosophy of fixing the problem not fixing blame on some group.

People that can ask the right question are much more impactful than people who can “only” find the answer. (NOT trivializing the smarts it takes to find the answer. )

My Pat Gelsinger story: I worked in a successor org to MIPO, in those days called MPG. One day during pre-silicon validation on the Pentium II, one of my NCG’s came to me and said: “An old test from the historical test archive is failing in the simulator. I want to find the author and ask him about it. Do you know a Pat Gelsinger?”

Me: “Well, he is an Intel Fellow now, so he might not remember what that test was supposed to do. “

> There are a lot of things you can say about Silicon Valley, but one of the more interesting aspects of it for me has been how "small" it is in terms of people have an oversized impact. I have never been one of those people of course, just a part of the entourage. But it has been interesting to watch and learn from folks who are good (and bad) role models.

I believe this is what you feel when you come to the industry when it is just starting in a given locality, and then it explodes. Senior workers who join in the earliest days tend to drift from company to company, but they rarely leave the place entirely, or change industries.

A sociologist I met described it as the "generation effect" which is that people live in a cohort of other people about their same age +/-10 years. Throughout one's life the composition of the cohort changes as you begin to specialize starting with college/first job and continuing if you focus on a particular area (for me it has been systems analysis).

This effect can drive people to over-pay to go to "good schools" or work for less than they think they should be paid at the "good places to work" because it lets them join a cohort with members who are statistically more likely to be "successful" (for some arbitrary definition of success). I personally never paid much attention to it but thought about it when talking with this person.

I know from experience that it happens in Silicon Valley that someone will say, "Oh I know someone at some previous company who is really good at that, let me see if they would be willing to change jobs." That is why companies pay people bonuses for "referrals."

I suspect it happens in LA in the entertainment industry as well. If you watch the old American television series "ER"[1]. Produced by Steven Spielberg's Amblin Entertainment production company, it is amazing to see people who were on that show first as guest actors and then show up later in series.

[1] https://www.imdb.com/title/tt0108757/

Reply on news.ycombinator.com