ua709
Joined 9 months ago
Karma
134
Submissions
72
Member since
Feb 7
Recent submissions
Comment · 5 months ago
Fair enough but I don’t work for AMD.
Comment · 5 months ago
> Are you seriously suggesting hobbyists should tapeout an ASIC instead of use an FPGA?
No. I said the low-end of FPGA sales is getting eaten by microcontrollers and the high-end of FPGAs sales is probably about to get eaten by custom ASICs.
Although the cost of making an ASIC is high, in the larger nodes it's not that high, and getting ever cheaper at FPGA performance levels and logic densities. FPGAs are terribly inefficient with their HW they're very easy to beat with an ASIC. They only get away with it because the NRE today is lower. But it's not an order of magnitude lower and I'm not sure how much longer that will be the case in nodes at 28nm and larger based on what I know Universities pay in tape-out classes.
Will there be very low qty projects where the NRE of developing an ASIC overwhelms that of an ASIC, sure. But will there be enough business in that niche to sustain the business of AMD, Intel and Lattice? Not obvious.
And I don't think the FPGA hobbyist market of people who "want to learn HDL" spends enough money to affect what's coming and this decision from AMD reflects that.
> 1. For one-off designs (quantity=1) ASICs will never beat a high end FPGA on unit price.
Never say never. These guys were able to convince investors you're wrong about that. :)
P.S. If you're a hobbyist who wants to make an ASIC... https://www.tinytapeout.com
Comment · 5 months ago
Just of out of curiosity, what parts of the University Program don't appeal to you?
https://www.amd.com/en/corporate/university-program.html
You can get free licenses and donated hardware through this program.
Comment · 5 months ago
I don't think AMD can say this but I think the reality is for most hobbyists this is the prevailing attitude:
"The Harsh Truth about FPGAs (You Should Avoid Them?!)" https://www.youtube.com/watch?v=l3d8uFKsJiY
a.k.a. Just use a microcontroller. And for the vast majority of hobby projects I suspect that is good advice. Low end FPGAs don't compete well with low end microcontrollers and more people know how to use microcontrollers.
Universities are fine as they can sign up for the University Program and get the licensees they used to get. https://www.amd.com/en/corporate/university-program.html
I think the reality is the niche that FPGAs occupied is getting hit hard on the high and low end. Cheap Chinese FPGAs are prevalent, cheap microcontrollers more so, and on the high-end making an ASIC that compete with a high-end FPGA has never been cheaper, and is getting cheaper and easier everyday. 65-28nm is very easy to use now (relatively speaking) and is very low cost with tons of tape outs and there is good competition. Beating an FPGA with an ASIC is not all that hard. Grad students at CMU, Stanford, Georgia Tech, etc. do it all the time in their tape-out class. Making an ASIC is not as easy as an FPGA for sure, especially if you need DDR and serdes. And NRE for ASICs for small volume ( <1K units) is higher. But it is getting easier and cheaper everyday. And it's now feasible for small teams (say ~6) to do it. I think they need to look very hard at where they spend their NRE now to stay relevant and they need to start getting brutal because I am sure the amount of revenue they're bringing in is under serious attack.
As to why Windows and not Linux? It's probably cheaper for them to maintain Windows for one reason or another. Maybe they don't even do it an just contract it out and Windows contractors are easier to find, but I'll bet it's just a basic cost issue at the end of the day.
Comment · 5 months ago
Schools can join the AMD University Program and get back to where they were, and more.
https://www.amd.com/en/corporate/university-program.html
As for hobbyists, in the world of $0.03 microcontrollers, strong competition on the low end from Chinese manufacturers, and where few people learn HDLs, is NRE money in the hobby market really money well spent?
I'm a HW designer and even I use microcontrollers now for most things, but not everything, because it's usually cheaper and faster.
With semiconductor prices coming down as far as they have I think the world has probably fundamentally changed for FPGAs and the niche they occupied is shrinking fast.
Comment · 5 months ago
If you have central control you might even be able to get away with changing the rules. i.e. most roads are now one-way leading out of the city. voilà we nearly doubled outbound throughput. Even just for commuting that would be awesome, not that it is happening anytime soon, but one can dream, especially while sitting in gridlock traffic.
Comment · 5 months ago
“ Minnesota House File 1606 would allow survivors to sue the owners of nudification apps for damages “
I have no idea how the law works in this case but if you could prove damages couldn’t you already sue? This is a guess but maybe this law clarifies what is considered damaging which is what enables them to prove damages where under previous definitions they couldn’t?
Comment · 5 months ago
Only 30? Those guys need to get their act together.
https://www.seattletimes.com/business/boeing-aerospace/congr...
Comment · 5 months ago
Thanks! I'm going to give this a shot on a nasty simulation I'm presently working on... :)
Comment · 5 months ago
SOTA = State of the Art? Like say Claude Opus 4.5? I actually want to try this out.
Comment · 5 months ago
> STM32 cubemx starter repo and ask for a feature
I'm confused, isn't the whole point of using the STM32CubeIDE that all the peripherals, like say setting up an ADC on pins 4 and 7, are checkbox features?
Comment · 5 months ago
> use cases related to root causing complex simulation failures.
That's a pretty interesting use case. I assume this is for RTL simulation given the thread, but how do you connect the output of the simulator to the AI?
Comment · 6 months ago
I wonder if that even works. Kinda like when kids play telephone I think it’s unlikely the input and output sentences actually match.