https://www.electronicdesign.com/technologies/eda/article/55408466/quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-designersMaybe AI can do PCB design.The graph midway down is interesting.
john larkin <jl@htigct.com> Wrote in message:rUntil you start worrying about cross-talk between traces, and voltage
https://www.electronicdesign.com/technologies/eda/article/55408466/quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-designersMaybe AI can do PCB design.The graph midway down is interesting.
Actually, there are not that many design rules to implement.
Should be fairly easy to do.
On 7/10/2026 6:15 am, Martin Rid wrote:
john larkin <jl@htigct.com> Wrote in message:rUntil you start worrying about cross-talk between traces, and voltage
https://www.electronicdesign.com/technologies/eda/article/55408466/
quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-
designersMaybe AI can do PCB design.The graph midway down is
interesting.
Actually,˙ there are not that˙ many design rules˙ to implement.
Should be fairly easy to do.
drops along traces.
There were a couple of occasions in my career where cleaning up a
printed circuit layout to deal with the unfortunate consequences of
these effects turned out to be essential to the success of a project or
a product.
The really horrible example was a ground layout which routed about 100mA
of logic supply current along a ground trace which subtracted about
100mV from a control voltage that had been carefully derived from 6.2V voltage reference generator. When the control voltage was at low end
it's range at it's source it had reversed sign by the time it had got to it's destination.
At least it was easy to fix.
https://www.electronicdesign.com/technologies/eda/article/55408466/quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-designers
Maybe AI can do PCB design.
The graph midway down is interesting.
On Mon, 05 Oct 2026 14:22:37 -0700, john larkin <jl@htigct.com> wrote:
https://www.electronicdesign.com/technologies/eda/article/55408466/quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-designers
Maybe AI can do PCB design.
The graph midway down is interesting.
That article is a load of crap. Written by the founder of an AI tool
company with a product which assists in PCB design, justifying the
need for such a tool.
There is no shortage of PCB layout staff. There is no shortage of
engineers full stop. Never have been. Pay enough and you'll get
staff.
As for AI in PCB layout the problem is not in routing, but in
optimising placement. Autorouters have worked for several decades -
if you spend time creating the constraints they work.
As for AI in design it is already offered by the big 3 EDA vendors for
both Analogue and Digital work.
Cadence for example have at least four agents that I know of
(viraStack ChipStack auraStack innostack) which specialise in
different areas. Virastack pairs up with Virtuoso (which is probably
used for 99%+ of all analogue chip designs) for analogue work.
Aurastack assists in PCB work, and the others in the 1's and 0's
stuff.
There is no shortage of PCB layout staff. There is no shortage of
engineers full stop. Never have been. Pay enough and you'll get
staff.
As for AI in PCB layout the problem is not in routing, but in
optimising placement. Autorouters have worked for several decades -
if you spend time creating the constraints they work.
As for AI in design it is already offered by the big 3 EDA vendors for
both Analogue and Digital work.
Cadence for example have at least four agents that I know of
(viraStack ChipStack auraStack innostack) which specialise in
different areas. Virastack pairs up with Virtuoso (which is probably
used for 99%+ of all analogue chip designs) for analogue work.
Aurastack assists in PCB work, and the others in the 1's and 0's
stuff.
On 10/11/2026 8:19 PM, Bill-Sloman wrote:
On 7/10/2026 6:15 am, Martin Rid wrote:
john larkin <jl@htigct.com> Wrote in message:rUntil you start worrying about cross-talk between traces, and voltage
https://www.electronicdesign.com/technologies/eda/article/55408466/
quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-
designersMaybe AI can do PCB design.The graph midway down is
interesting.
Actually,˙ there are not that˙ many design rules˙ to implement.
Should be fairly easy to do.
drops along traces.
There were a couple of occasions in my career where cleaning up a
printed circuit layout to deal with the unfortunate consequences of
these effects turned out to be essential to the success of a project
or a product.
The really horrible example was a ground layout which routed about
100mA of logic supply current along a ground trace which subtracted
about 100mV from a control voltage that had been carefully derived
from 6.2V voltage reference generator. When the control voltage was at
low end it's range at it's source it had reversed sign by the time it
had got to it's destination.
At least it was easy to fix.
In my experience -- albeit perhaps limited -- the business people really don't care if a project fails because the technical expertise to realize
it is insufficient in-house.˙ Nor do they care when people with good technical skills are let go [presumably to a different employer.]
After all, the business people continue to be paid, even when several projects fail.
On 7/10/2026 6:15 am, Martin Rid wrote:> john larkin <jl@htigct.com> Wrote in message:r>> https://www.electronicdesign.com/technologies/eda/article/55408466/quilter-the-us-is-reshoring-chips-but-its-still-missing-the-board-designersMaybe AI can do PCB design.The graph midway down is interesting.> > Actually, there are not that many design rules to implement.> Should be fairly easy to do.Until you start worrying about cross-talk between traces, and voltage drops along traces.There were a couple of occasions in my career where cleaning up a printed circuit layout to deal with the unfortunate consequences of these effects turned out to be essential to the success of a project or a product.The really horrible example was a ground layout which routed about 100mA of logic supply current along a ground trace which subtracted about 100mV from a control voltage that had been carefully derived from 6.2V voltage reference generator. When the control voltage was at low end it's range at it's source it had reversed sign by the time it had got to it's destination.At least it was easy to fix.-- Bill Sloman, Sydney
The point about AI extracting layout constraints is the interesting part to me. It suggests a useful division of labour: the system can collect details and explore arrangements, while people still decide what the design is for, what trade-offs are acceptable, and who is responsible when the constraints were incomplete.
The essay?s argument is that access to a capable system is not the same as freedom if the person cannot inspect its work, keep control of the tool, or share in the gains. In practical electronics work, where do you think the most important human boundary remains: specifying the objective, checking
the constraints, or accepting responsibility for the finished board?
On 10/14/2026 8:52 PM, Martin Davies-Roundhill wrote:
The point about AI extracting layout constraints is the interesting part to >> me. It suggests a useful division of labour: the system can collect details >> and explore arrangements, while people still decide what the design is for, >> what trade-offs are acceptable, and who is responsible when the constraints >> were incomplete.
Machines should always handle the "grunt" work -- the work that
is repetitive, error prone (due to its repetitive/boring nature)
and comprehensive.
This is as things have been done, historically.
Doing placement with rough connectivity algorithms (heuristics for which >humans would rely on ratsnests), autorouting, enforcing design rules, >validating layout against schematic, etc. All dreary tasks that
no one really looks forward to.
It's only recently that NLP suggests an AI could "read a datasheet"
(or an app note) and extract layout criteria from that -- instead of relying >on a human to express those constraints in a manner that an *existing*
tool could understand (if your tool *doesn't* understand them, then it
relies on the human operator to enforce them -- let the AI assume the role
of the human in that case).
Only humans can assess the relative value tradeoffs in a design;
how much reliability do you sacrifice to save on production costs?
how much performance do you trade for ______________? etc.
The essay?s argument is that access to a capable system is not the same as >> freedom if the person cannot inspect its work, keep control of the tool, or >> share in the gains. In practical electronics work, where do you think the
most important human boundary remains: specifying the objective, checking
the constraints, or accepting responsibility for the finished board?
I don't think anyone is equating AI with "freedom".
The problem with LLMs is you can't see their reasoning, just their >conclusions.
Ages ago, I recall listening to a grandmother lament the fate
of one of her friends recently diagnosed with cancer.
"Once the AIR gets to it, their gone!"
She refering to exploratory surgery (predating MRI) to ascertain
the extent of a cancer's spread. "Opening up" the patient
("exposing it to air"). But, this step often precedes death,
not because of the "air" but because of the procedure that
exposed the cancer.
What seemed to make sense to the grandmother (air -> death)
was based on a faulty rationale.
On Wed, 14 Oct 2026 22:44:44 -0700, Don Y
<blockedofcourse@foo.invalid> wrote:
On 10/14/2026 8:52 PM, Martin Davies-Roundhill wrote:
The point about AI extracting layout constraints is the interesting part to >>> me. It suggests a useful division of labour: the system can collect details >>> and explore arrangements, while people still decide what the design is for, >>> what trade-offs are acceptable, and who is responsible when the constraints >>> were incomplete.
Machines should always handle the "grunt" work -- the work that
is repetitive, error prone (due to its repetitive/boring nature)
and comprehensive.
This is as things have been done, historically.
Doing placement with rough connectivity algorithms (heuristics for which >>humans would rely on ratsnests), autorouting, enforcing design rules, >>validating layout against schematic, etc. All dreary tasks that
no one really looks forward to.
It's only recently that NLP suggests an AI could "read a datasheet"
(or an app note) and extract layout criteria from that -- instead of relying >>on a human to express those constraints in a manner that an *existing*
tool could understand (if your tool *doesn't* understand them, then it >>relies on the human operator to enforce them -- let the AI assume the role >>of the human in that case).
We have data sheets that run to 600 pages or more. I'm happy with the
ide of an AI tool that looks at my schematic and PCB and prowls all
the related data sheets and flags possible errors. Alspice claims to
do that and I think it may be useful. They showed mr one example that
clains to have discovered an open drain that needs a pullup, and that
it checked all the power supplies and in/out direction on all the
chips, stuff like that.
Data sheets seem to be getting progressively longer and worse.
Only humans can assess the relative value tradeoffs in a design;
how much reliability do you sacrifice to save on production costs?
how much performance do you trade for ______________? etc.
The essay?s argument is that access to a capable system is not the same as >>> freedom if the person cannot inspect its work, keep control of the tool, or >>> share in the gains. In practical electronics work, where do you think the >>> most important human boundary remains: specifying the objective, checking >>> the constraints, or accepting responsibility for the finished board?
I don't think anyone is equating AI with "freedom".
The problem with LLMs is you can't see their reasoning, just their >>conclusions.
That's fine if it's spotting possible mistakes.
Ages ago, I recall listening to a grandmother lament the fate
of one of her friends recently diagnosed with cancer.
"Once the AIR gets to it, their gone!"
She refering to exploratory surgery (predating MRI) to ascertain
the extent of a cancer's spread. "Opening up" the patient
("exposing it to air"). But, this step often precedes death,
not because of the "air" but because of the procedure that
exposed the cancer.
What seemed to make sense to the grandmother (air -> death)
was based on a faulty rationale.
That is a fair distinction. I did not mean AI itself is freedom; I meant
that handing repetitive work to a system might create more room for choice, but only if the system remains understandable and accountable. Your
datasheet example makes the practical boundary clear: an assistant that
flags a possible missing pull-up could save a search through hundreds of pages. I would want each flag to point to the exact datasheet requirement, the pin or net it matched, and a check the designer can reproduce. That
gives us something inspectable instead of asking us to trust an opaque conclusion. The person still makes the trade-off and signs off on the
design. Do current tools show that source-level evidence clearly enough, or does the engineer still have to verify everything manually?
| Sysop: | Jacob Catayoc |
|---|---|
| Location: | Pasay City, Metro Manila, Philippines |
| Users: | 3 |
| Nodes: | 4 (0 / 4) |
| Uptime: | 497972:48:57 |
| Calls: | 190 |
| Files: | 846 |
| Messages: | 71,498 |