I am hiring for senior or principal developer technology engineers (DevTechs). At some companies, they are called application engineers. Nvidia has a very high bar of quality for these engineers, as they need to both have deep experience and understanding of software engineering and development, as well as excellent communications skills to work with external devs.
They are distinct from solution architects in that every DevTech is expected to be a seasoned professional engineer. It is a special forces engagement model where we approach high value strategic partners with deeply skilled engineers to accelerate their adoption of our platform.
What you will be doing: Working closely with our strategic partners and internal teams to get the most out of our enterprise platform hardware/software stacks. This involves everything from serverless computing to optimizing end-to-end cloud workflows. It helps if you spent a few years grinding out C++. We are specifically looking for people who have front-end experience to help fill team gaps.
Also, if you are not super skilled at cloud infra, but are a steely-eyed computer engineer who has GPU accelerated your cat, feel free to apply to one of the roles as well. We have needs beyond just the cloud stuff.
How to apply: Email me directly at mclive at nvidia dot com, and apply to these roles on our career site:
Location: Nvidia is open to remote work. There are benefits to being near HQ in Santa Clara (nice offices!) and benefits to being in EU (No 7:30am meetings!) I am interested in applications from all geo regions.
I don't know if that is true. Stuff just changes. To solve similar problems to this, my lab developed sapphire metal bonding techniques that could be run with CNC machined parts. Everything became nice and reliable and repeatable.
We lose techniques all the time but gain new ones. I designed a scientific instrument once that had a long precision bore through it. I designed it with the belief that I needed clearances for a large stiff boring bar to make the feature.
I showed the prints to my machinist, and he told me they were just gonna put it on the five axis and hit the bore on both sides with an end mill. That it was as accurate as how the old guys did it.
It's entirely possible that this aerospike engine is actually lighter because they don't need large nozzles and nozzle extensions and are more efficient
Standard nozzles do not need a cooled support structure right in the middle of the combustion chamber. I love these things, I made a linear aerospike (Never fired), but this thing just looks heavier than a standard TCA/nozzle
This is almost identical to the process of setting up a manufacturing facility here in the bay area. Difference is that you never do all this yourself. You use permit sherpas and general contractors to handle most of the paperwork. Source: done it three times
I have done it both ways! I left a great career in computer graphics to go through a long journey through the aerospace and energy sectors, and now I am on a return trip to CG.
During this process, I
-Built my own machine shop
-Worked at <Cutting edge aerospace company>
-Built up an entire material science research lab
-Learned how to manage an amazing team
-Got really into machining, manufacturing and facilities
Whoever said that most of your time as a mechanical engineer is spent buying stuff and sorting through boxes/messing with supply chain/inspection is right. Your tolerance for cardboard boxes must be very high to do physical engineering.
Machining is really fun and soothing at times, but once you have learned it enough, you realize that Xometry and Protolabs and Front Panel Express and SendCutSend are just... better than you. One thing I learned is that machining was a fun hobby, but it is not fun to have it as a part of another hobby. It is always cheaper and quicker to outsource that labor. I want to make a telescope eyepiece tray, and I know I can design it and find a makerspace with a laser cutter and make it myself, but I am probably just going to order it from FPE so it will show up to my door, done, without any painting etc. I almost bought a mini lathe again (I sold off my first machine shop), but I realized that the thing I was thinking of making with it (tiny model biprop rocket engines) were hard to make and if I did pick up that hobby, I would rather just pay my buddy to make it on his superlathe and spend that time on designing the stuff he would make.
Another huge part of physical engineering is relationship management. You need to have someone to do everything. You need a good relationship with your sputter coater, with your weld shop, with your general contractor.
There is not actually that much senior level design engineering work to be done. Usually design work is done by about 10% of the org, and the rest of the org does support activities for that. There is nothing wrong with that, and there is plenty of interesting work in the support activities (I made a career of doing those activities)
I did all this with no formal education in the sector. I spent a lot of time in machine shops, reading books, going to rocket launches, networking. John Carmack was my principal inspiration (I met him at a tiny rocket conference once)
As far as comp is concerned, it is decent but not MANG like. I got lucky at <Aerospace place> and got in early enough that my TC was actually pretty solid due to relatively large share grant. I did have 5 room mates when I worked there because the cash level was low. My first aerospace job was paying me the same salary as an In N Out employee makes now. (They were taking a chance on me, so no hard feelings at all, it was a privilege to make that sacrifice!). You can make a living as a physical engineer, but you might not make a killing.
So, if you are committed to a huge downlevel for 5 years, you can have some fun adventures!
PS: Knowing how to write python is a superpower in these fields, I would guess than less than 10% of mechEs at the senior level know how to code. I imagine new grads know how to code now. I always was able to sneak a quick victory in various projects because I could write some hack code that did something tedious quickly.
PPS: Respect. Your. Technicians. Respect them. Treat them as partners. Do right by them. Value their feedback. Do their work from time to time. Pay attention when they speak, and act on what they say. Same goes for your machinists, material handlers, buyers, etc.
Theories: 1. I was looking through a bunch of old facebook photos of myself last night. Maybe I triggered a job that scans all old photos through the modern facebook image analyzer.
Problem with that theory: I don't think I was enough of an edge lord to post that ever. I don't remember ever posting anything like that. I am Jewish. Also, there were no likes, no comments, and there was not date to the post, just said 2 hours ago
2. Someone figured a way to hack a facebook API such that it looks like pictures are posted by me. Possible. Maybe he found the API keys for Pixelpipe or some other defunct system. Thing is, the picture never actually showed up on my wall, and was not hooked into the dbase as being posted by me as there was no profile photo, name or timestamp. It never made it to the public, it just appears to be sent to me, directly, through an obscure conduit of a notification of censorship.
Also: Why just me, and why a nine inch nails song as the comment?
