You know you've made it as an engineer when you switching jobs gets an article in reuters.
So is he 10x or how much?
10000x at least. Whole staffed departments at rivals can't bring to market what he did repeatedly.
Jim Keller moves to AI chip startup
11–20 of 94 posts
Re: Jim Keller moves to AI chip startup
#12This page has some details of Tenstorrent's current chips [1]. Looks like a manycore simd design focused on tensor ops from a brief look. Apparently they also have some sort or compression scheme to boost memory bandwidth. 1: https://www.tenstorrent.com/technology/
PiedPiper for machine learning?
Re: Jim Keller moves to AI chip startup
#13Earlier quoted context omitted.
So is he 10x or how much?
10000x at least. Whole staffed departments at rivals can't bring to market what he did repeatedly.
Nobody does that stuff alone. I think a more accurate description would be that he seems to be a force multiplier for the team he leads, which can have a bigger impact than any singular engineering feat.
Re: Jim Keller moves to AI chip startup
#14Earlier quoted context omitted.
10000x at least. Whole staffed departments at rivals can't bring to market what he did repeatedly.
Nobody does that stuff alone. I think a more accurate description would be that he seems to be a force multiplier for the team he leads, which can have a bigger impact than any singular engineering feat.
That's what Xx means often. Leads can depress teams' output, maintain it, or enhance it, and he presumably enhanced hugely.
Re: Jim Keller moves to AI chip startup
#15> The chips also operate on the assumption that future software will involve programmers giving high-level directions while artificially intelligent computers write much of the nitty gritty code required to implement those human ideas.
I can't wait to stop writing machine code.
Re: Jim Keller moves to AI chip startup
#16Predicted. https://news.ycombinator.com/item?id=23498142
What was your prediction based on?
Re: Jim Keller moves to AI chip startup
#17I often find it hard to believe that a single person can make much of a difference in such intricate problem domains as chip design but in his case the evidence is overwhelming. Also goes to show what a shit show Intel has become since even he was not able to right that ship. I think in the CPU space it will be all ARM and RISC ten years from now and since Intel never really managed to become a dominant player in any other (relevant) field, they are pretty much done for.
Re: Jim Keller moves to AI chip startup
#18Jim is a tour de force: https://www.youtube.com/watch?v=Nb2tebYAaOA
This was one of the most impressive things I've ever seen. The sheer mind puzzle exploration of "well..if we had a CPU the size of the sun, here's why it still wouldn't work". Guy seems unbelievably brilliant.
Re: Jim Keller moves to AI chip startup
#19I often find it hard to believe that a single person can make much of a difference in such intricate problem domains as chip design but in his case the evidence is overwhelming. Also goes to show what a shit show Intel has become since even he was not able to right that ship. I think in the CPU space it will be all ARM and RISC ten years from now and since Intel never really managed to become a dominant player in any…
Intel died with Andy Grove. Every CEO after that has been living off the momentum he created.
If his successors applied his management style, mobile phones would have an Intel processor in them.
Re: Jim Keller moves to AI chip startup
#20I often find it hard to believe that a single person can make much of a difference in such intricate problem domains as chip design but in his case the evidence is overwhelming. Also goes to show what a shit show Intel has become since even he was not able to right that ship. I think in the CPU space it will be all ARM and RISC ten years from now and since Intel never really managed to become a dominant player in any…
It is hard to believe. However out of my own experience: Execution, decision making, vision and goals, people, alignment, setbacks. A very complex mixture and the more people there are, the more important a driver is.
PS: Nice nickname. Read the book from Th. Mann - over a period of 7 years I think.