DHaldane

5 ms·

1.
by DHaldane 2y ago
It's ok to build a compiler sometimes -- it's just very important to make that choice intentionally
2.
by DHaldane 2y ago
This is an old story, rehashed without new information. That 7nm chip from SMIC is what prompted the export controls -- nobody thought they were close before that.
3.
by DHaldane 2y ago
I've found the same thing - a little syntax example, some counter examples and generative AI does well generating syntactically correct code for PCB design. A lot of the netlists are electrically nonsense when it's doing synthesis
4.
by DHaldane 2y ago
Have you tried using more 4000x more samples? https://redwoodresearch.substack.com/p/getting-50-sota-on-ar...
5.
by DHaldane 2y ago
I've found that for speeding up design generation like that, most of the utility comes from the coding approach. AI can't do it itself (yet), and having it call the higher level functions doesn't save that much time...
6.
by DHaldane 2y ago
I promise that we want to stay a software company that helps people design things! Adding Skynetn't to company charter...
7.
by DHaldane 2y ago
Agree that PDF digesting was the most useful. I think Gemini could definitely do that microphone study. Good test case! I remember spending 8 hours on DigiKey in the bad old times, looking for an audio jack that was 0.5mm shorter.
8.
by DHaldane 2y ago
Right - LLMs would be a bit silly for these cases. Both overkill and underkill. Current approach for length matching is throw it off to a domain specific solver. Example test-circuit: https://x.com/DuncanHaldane/status&
9.
by DHaldane 2y ago
I doubt it too, but I notice that I keep underestimating the models. Do you have a challenge task I can try? What's the easiest thing I could get an LLM to do for circuit board design that would surprise you?
10.
by DHaldane 2y ago
My gut agrees with you that LLMs shouldn't do this well on a specialty domain. But I think there's also the bitter lesson to be learned here: many times people say LLMs won't do well on a task, they are often surprised either
11.
by DHaldane 2y ago
Update: Sonnet 3.5 is better than any other model for the circuit design and part finding tasks. Going to iterate a bit on the prompts to see how much I can push the new model on performance. Figures that any article written on LLM limits i
12.
by DHaldane 2y ago
It really depends on the type of question, but generally I'm between Gemini and Claude these days for most things.
13.
by DHaldane 2y ago
That's an interesting question - I'll take a few pokes at it now to see if there's improvement.
14.
Testing Generative AI for Circuit Board Design(blog.jitx.com)
353 points by DHaldane 2y ago|168 comments
15.
by DHaldane 2y ago
I'll note that the bandwidth doubled from PCIe 4.0 to PCIe 5.0 (it's now 30GHz). Dealing with that speed increase is causing a whole lot of problems for all the designers in the ecosystem. It's a bit of a nightmare, but peopl
16.
by DHaldane 2y ago
Without knowing exactly what you are working on, here are a few ideas: 1. Figure out if you can do fewer slow iterations. What's driving the need for a full PCBA run for each iteration? Might be able to split out to rigid assembly and
17.
by DHaldane 2y ago
Fascinating technology - if I interpret it correctly it looks like you're replacing lots of manual pipetting with a surface that can scooch little drops around using electrostatics. And then do a bunch of stuff you can't do with a
18.
by DHaldane 2y ago
Thanks for the feedback -- very helpful.
19.
by DHaldane 2y ago
Seems right - Renesas just acquired Altium after all. Varies a lot by customer.
20.
by DHaldane 2y ago
Good set of questions, and thanks for the feedback. Currently updating the website to add this stuff. Comments/answers to questions in order: 1. Yes, finding a single resistor is easy. Part solving is more about making parametric circu
21.
by DHaldane 2y ago
Welp - seems there are out of date parts of the website. We launched router last Tuesday so have to update that we are end to end now. Thanks for the pointer!
22.
by DHaldane 2y ago
Thanks Mike! We're still in VSCode -- thank you for kicking off that very first extension.
23.
by DHaldane 2y ago
Hey - you got this very correct. Good insights as to what we're up to. Grueling to build product for the pros, but once done it can support a lot of people. Good automation can dissolve false dichotomies.
24.
by DHaldane 2y ago
This is the classic SaaS CAD tool approach. But JITX runs locally - all design data stays local and we're not holding it hostage in our cloud. Definitely agree with a free version though - we want to help those just starting out as wel
25.
by DHaldane 2y ago
Schematics ended up being a hard requirement, so we put in the work to generate good ones automatically and the tools for EEs to adjust the aesthetics without ruining the correctness of the design. You are right that the schematic becomes
26.
by DHaldane 2y ago
Thank you. Talking to EEs that do this every day I hear "Well this would have to save me 1 day per month to be ROI positive". Turns out saving that day is incredibly hard, because the professional workflows are intense. Value of t
27.
by DHaldane 2y ago
Good catch - we can add some clarification to that form. Today JITX is for professional EEs mainly, and we don't have a lot of resources to help people learn electronics from scratch. (Still want to help the students though - this comp
28.
by DHaldane 2y ago
Right - "none of the other tools try to integrate nicely, so if we support it well it must be necessary" is the thought to address.
29.
by DHaldane 2y ago
Thank you for this comment! Would you pretty please help me debug my website - what makes JITX seem like a CAD plugin? (it is a freestanding tool)
30.
by DHaldane 2y ago
We are certainly focused on the professional market for monetization -- turns out this software is hard to build and we need to pay smart people to work on it. Not to say we won't have an open free/freemium offering soon. Hobbyist

More ›