A different way to say it is "If you found a gold mine, you can afford to do things others can't".
99% of companies don't sit on gold mines like Nvidia.
Some industries may have regulatory protection, but I don’t think it’ll be long before others start mass-outsourcing the kinds of jobs people thought were safe.
I used to work at a US company that had offices here in Mexico and had a GREAT QA team in India (despite the hatred they get here in HN, Indian techies are GOOD! ) .
The TZ problem is what ended up fracturing the relationship between the teams . Us/Mexico teams synergy was just great, while Indians were generally disconnected.
But yeah, US and Canada companies that turn to LatAm do get an "unfair advantage". I'm slowly but steadily trying to help this happen. Sorry?
I've seen more of a mentality and attitude problem than timezones. Often these other offices are just treated as 2nd class citizens without enough context to act.
My company has been trying to do this (pay top percentile latam salaries) but it’s just been a larger grind to find decent devs than when we were advertising US salaries. “You get what you pay for” seems to be at play for us.
I imagine remote work can give you some edge but we’re trying to get 2 devs for the price of one and it’s really hard. We’re probably losing money on all the time wasted (you get a ton of applicants at lower price points).
Agree on the quality - there were some devs we worked with who needed more hand-holding, but a majority were very capable.
Thinking that there will be a switch-flip moment (or that, because there hasn’t, there won’t be) is, I think, a limited view.
Average American devs just aren’t enough better, or won’t be soon, to justify the massively higher cost, and I think the push from a lot of devs around remote-first policy has forced employers to figure out timezone/asynch issues.
I will say that I’m really guessing about 3-5 years from now. Happy to see other perspectives!
You can also enforce it through trade laws, requiring imports to come from union shops.
Mainly they decided to manufacture their tech here in Mexico instead of in Asia. But here on Mex we dont have the skills to to such high tech manufacturing.
For the sake of argument, let’s reduce Nvidia (my employer but I don’t speak for them) to just a silicon vendor.
You’re not only ignoring the work of thousands who design the stuff that goes on that silicon, the billions of $ in R&D, you’re also dismissing the decisions about what to put in that design, way before other people do it. Some of those decisions were made years before the actual implementation started. It happens to be something a few people at Nvidia are very good at.
There were times in the past where others had a higher number of FLOPS, but it didn’t make a dent in the usage of CUDA.
My standard assumption about anything in tech (and beyond) is that you don’t win on first principles alone (“parallel design is well understood”). It’s the details that make you stand out.
Bill Daly gave a great presentation at Hot Chips a few weeks ago that explains how Nvidia’s AI performance was only in small parts due to process improvement. The rest was doing things smarter: matrix sparsity, different number representations and so forth.
They then had the good fortune that their competitors tried to leverage open source and cross company collaboration. While misc open source winning is sometimes thing, the industry built all their stuff on cuda instead of opencl or openmp.
Nvidia thus built their gold mine.
I don't think it's impenetrable. There are some design mistakes baked into their system that are going to be difficult to unravel without breaking backwards compatibility.
They don't make their own chips, right?
In 2023, this is a moot point. Neither does Apple. Neither does AMD. Neither does Qualcomm (which makes the vast majority of mobile phone CPUs). Neither does Broadcomm.To your second point: I am surprised that AMD hasn't tried harder to chase the high-end enterprise GPU market. Maybe that is a longer term secret plan after they gain more CPU market share. It is possible that a GPU ecosystem cannot out-compete NVidia's CUDA at this point.
Had media companies maintained their exclusivity on people’s screen time, then they would be in an even better position. But the supply of content exploded, from video games to YouTube to TikTok and Instagram and even forums like this one.
More supply relative to demand means lower prices.
I agree with your point about competition. On the other hand, mindless channel surfing has mostly been replaced at the same time. People do it to some degree with streaming and even channels setup for the purpose--but it's a less natural fit.
Try doing manufacturing work entirely remotely. Good product doesn't move when you can't actually make it.