just everything that comprises real time, high speed and low latency. the developing time and skills needed are very high so it has to deserve the efford.
LOL Please define "real time" and "high speed". Points for "latency". "Developing time"??? I was putting out a new design every 3 months in my FPGA design. This was because I created an embedded micro-sequencer in my design. I only had to change a few parameters and POOF the entire response curve of the signal was updated to new customer requirements. I say "3 months" only because it took them a month to come up with the new requirements and another couple of months to do testing. Code changes were in the order of 1 week. HOWEVER, I do agree, the initial design took a long time. But divide that by total number of products created by the same design, and you have the equivalent of 2 months of FPGA design time per customer. So, it's not the technology which drives the "development time", but rather the skills of the engineer! The beauty of FPGA is that you create reusable modules. Once you have 100 or so modules, the design time is highly reduced. The same goes for PCB design. Think "architecture" and "reusability". Concerning skills, yes, you can learn (poorly) to write a few lines of code, but (good) FPGA design requires that you are taught by someone who understands how to think. The effort is likely 10 fold.
Latency, bandwidth, and IO flexibility are the only reasons we put an FPGA into a design. FPGA design consists of a very small percentage of the overall job market and the pay is not much more. The reason the pay is not much more is because most people want Embedded Programmers. They want embedded programmers because 99% of the people you work for will not even come close to understanding the purpose or the benefit of FPGA. I can create virtual parallelism in an embedded processor. I can also create sequential logic in an FPGA. Either way, the end result is latency. Sorry kiddies, the real answer is WAY more complicated than presented in this video - good luck! In a nutshell however- don't learn FPGA unless you also have general programming skills as backup. Otherwise you may find yourself traveling long distances to find a job. Personally I do PC level, embedded level, FPGA, and schematic level design. Being able to switch from one technology to the other has proved very helpful and interesting. However, if you can find a manager smart enough to appreciate FPGA, stick with that company because they are far and few.
@@IshikaDas-xw9jp It won't kill the job but it will reduce the demand. Any job be it hardware or software side, AI will always have a huge role to play. I still feel it can't kill the job as such, you will always need good RTL design engineers.
Considering Devon the latest and greatest AI Coding engineer. Are you aware of any efforts made in creating an FPGA Coding AI engineer? Thank you for the video, very informative.
Thanks for the comment, I'm glad you found the video informative. While there's definitely a lot of buzz around AI coding these days, like Devon, I'm not aware of any specific tools that can write full FPGA code just yet. There are some AI options, like ChatGPT, that can generate simpler code snippets, but FPGA design is very hardware-specific and requires a deep understanding of resource constraints and logic optimization. In my opinion, AI isn't quite there yet for complete professional-level FPGA development. It's an interesting area though, and who knows what the future holds!
Download my free PDF guide to begin your journey with FPGA design - "Get Started With FPGA In 20 Minutes" 👇
bit.ly/StartFPGA
just everything that comprises real time, high speed and low latency. the developing time and skills needed are very high so it has to deserve the efford.
LOL Please define "real time" and "high speed". Points for "latency". "Developing time"??? I was putting out a new design every 3 months in my FPGA design. This was because I created an embedded micro-sequencer in my design. I only had to change a few parameters and POOF the entire response curve of the signal was updated to new customer requirements. I say "3 months" only because it took them a month to come up with the new requirements and another couple of months to do testing. Code changes were in the order of 1 week. HOWEVER, I do agree, the initial design took a long time. But divide that by total number of products created by the same design, and you have the equivalent of 2 months of FPGA design time per customer. So, it's not the technology which drives the "development time", but rather the skills of the engineer! The beauty of FPGA is that you create reusable modules. Once you have 100 or so modules, the design time is highly reduced. The same goes for PCB design. Think "architecture" and "reusability".
Concerning skills, yes, you can learn (poorly) to write a few lines of code, but (good) FPGA design requires that you are taught by someone who understands how to think. The effort is likely 10 fold.
Latency, bandwidth, and IO flexibility are the only reasons we put an FPGA into a design. FPGA design consists of a very small percentage of the overall job market and the pay is not much more. The reason the pay is not much more is because most people want Embedded Programmers. They want embedded programmers because 99% of the people you work for will not even come close to understanding the purpose or the benefit of FPGA. I can create virtual parallelism in an embedded processor. I can also create sequential logic in an FPGA. Either way, the end result is latency.
Sorry kiddies, the real answer is WAY more complicated than presented in this video - good luck!
In a nutshell however- don't learn FPGA unless you also have general programming skills as backup. Otherwise you may find yourself traveling long distances to find a job. Personally I do PC level, embedded level, FPGA, and schematic level design. Being able to switch from one technology to the other has proved very helpful and interesting. However, if you can find a manager smart enough to appreciate FPGA, stick with that company because they are far and few.
It depends what the application is going to be. There are HFT firms who pay huge amounts to FPGA developers.
@@AshishPatel-vy7mn Sir will AI kill FPGA jobs in future?
@@IshikaDas-xw9jp It won't kill the job but it will reduce the demand. Any job be it hardware or software side, AI will always have a huge role to play. I still feel it can't kill the job as such, you will always need good RTL design engineers.
@@AshishPatel-vy7mn thanks
@@AshishPatel-vy7mn So will it be worthy off to choose career in this field in upcoming time? I am targeting HFTs like hudson, Tower research etc
Thank you so much for the information
Considering Devon the latest and greatest AI Coding engineer. Are you aware of any efforts made in creating an FPGA Coding AI engineer? Thank you for the video, very informative.
Thanks for the comment, I'm glad you found the video informative.
While there's definitely a lot of buzz around AI coding these days, like Devon, I'm not aware of any specific tools that can write full FPGA code just yet. There are some AI options, like ChatGPT, that can generate simpler code snippets, but FPGA design is very hardware-specific and requires a deep understanding of resource constraints and logic optimization.
In my opinion, AI isn't quite there yet for complete professional-level FPGA development. It's an interesting area though, and who knows what the future holds!
the perfect job for an artist ;)
what
@@hueheeuuehueuheuheuhueheeu1245fpga design is an art
Thanks . it’s help me more sir
Glad to hear that
Thank you
You the best
Are you Tomasz Adamek's brother ? :) pl.wikipedia.org/wiki/Tomasz_Adamek
Thanks so much for the information.