Having an FPGA board?

[David83]

Hello all,

I have a revived interest in hardware design since my college. Now I'm working on Digital Logic design using Verilog. Is this enough, or working on an FPGA board is essential at some point? I recently leaned that an FPGA could be used for verification, and I guess that's an important part of the design process. Is it easy to obtain one with all the necessary cables and stuff to work on it at home?

Thanks

 

[sythe]

It is quite easy to obtain one. Just have a google.
Ebay offers used board. And probably both Aliexpress and Ebay offer chinese manufactured cheap and simple boards.
More expensive one come from Terasic or Digilent etc

Simon

 

[David83]

It is quite easy to obtain one. Just have a google.
Ebay offers used board. And probably both Aliexpress and Ebay offer chinese manufactured cheap and simple boards.
More expensive one come from Terasic or Digilent etc

Simon

Thanks. Do you think it's something essential to have to be a hardware designer?

 

[sythe]

Personally, I think it is.
It gives tons of learning experience.
Theory can be quite different than real world.

Next, investigate in getting a "free" vendor-locked (modelsim) or open source simulator like ghdl/icarus with GTKwave and invest time in getting proper test-benches and regression suites.

 

[David83]

Personally, I think it is.
It gives tons of learning experience.
Theory can be quite different than real world.

Next, investigate in getting a "free" vendor-locked (modelsim) or open source simulator like ghdl/icarus with GTKwave and invest time in getting proper test-benches and regression suites.

I'm sure I still have a lot to learn. I'm considering to get a board. I just want to familiarize myself with the basics first to be able to write a synthesizable RTL codes for little more complicated designs.

 

[TrickyDicky]

Hello all,

I have a revived interest in hardware design since my college. Now I'm working on Digital Logic design using Verilog. Is this enough, or working on an FPGA board is essential at some point? I recently leaned that an FPGA could be used for verification, and I guess that's an important part of the design process. Is it easy to obtain one with all the necessary cables and stuff to work on it at home?

Thanks

If you're not planning on putting your design on a board, whats the point doing HDL? You may as well do programming instead - its much faster and platforms are more readily available.

 

[David83]

I didn't say I'm not planning, I asked a question, and I'm willing to buy one.

 

[vGoodtimes]

@David83,

For a hobbyist or student, FPGAs tend to be "a solution looking for a problem". Basically, the person wants to solve any problem using an FPGA but doesn't have a specific problem.

There are a variety of low-ish cost FPGA boards each with a different niche. The main issue is that the boards are generally not competitive vs a chromebook or a raspberry pi or etc... The ones that are useful tend to be more expensive, or both expensive and application specific.

An example of a useful FPGA would be on some of the SDR boards. In this case, the FPGA actually serves a purpose. The boards are not nearly as cheap as things like the de0-nano or zybo or etc... These are at least cheaper than the pro dev-boards for networking or multimedia processing.



At the same time, you can do designs for free if you generate post-implementation simulations. This can get you the very important timing constraints aspect, but not the external-interface gremlins aspect, nor the bring-up experience.

 

[David83]

Thanks, but is it a purpose of FPGA boards to experiment with different problems? I mean, what's the problem in not having a practical problem? For me I just need to learn how to transfer a piece of software to a hardware. Solving a practical problem isn't my goal for now.

And if I can program an FPGA, why cannot I get experience? if I cannot get experience from owning an FPGA, where from can I get the necessary experience to have a good position in the job market?

OK, let me put this way: I would like to be a good hardware designer with a good chance in the job market, what are the steps to achieve this, from the theoretical knowledge to the practical skills and experiences that I need?

 

[TrickyDicky]

Thanks, but is it a purpose of FPGA boards to experiment with different problems? I mean, what's the problem in not having a practical problem? For me I just need to learn how to transfer a piece of software to a hardware. Solving a practical problem isn't my goal for now.

And if I can program an FPGA, why cannot I get experience? if I cannot get experience from owning an FPGA, where from can I get the necessary experience to have a good position in the job market?

OK, let me put this way: I would like to be a good hardware designer with a good chance in the job market, what are the steps to achieve this, from the theoretical knowledge to the practical skills and experiences that I need?

To be a good hardware designer with a good chance in the job market, you really need to have some experience in a digital electronic job. Hobbiests are unlikely to be taken seriously.
The job market is more specific - people in the video market will look for people with video experience. data processing will look for data processing experience. Anyone who understands digital logic can be tought an HDL - it is the knowledge of digital systems that is more important than coding.

Things that would help would be understanding interfaces: ethernet, axi, PCIe - things like that.

 

[David83]

I have a degree in Computer Engineering. I don't consider myself a hobbyist, but I switched majors in the advanced degrees, and I'm reading again about logic design and computer architecture, and I'm thinking to go to VLSI design as well. It's coming to mind again. I agree though coding is not the most important part.

Regarding the experience, if the job market needs experience and you don't have any, how can you obtain this experience? Most jobs ask very specific skills that is hard to get unless you get the job!!

 

[TrickyDicky]

I have a degree in Computer Engineering. I don't consider myself a hobbyist, but I switched majors in the advanced degrees, and I'm reading again about logic design and computer architecture, and I'm thinking to go to VLSI design as well. It's coming to mind again. I agree though coding is not the most important part.

Regarding the experience, if the job market needs experience and you don't have any, how can you obtain this experience? Most jobs ask very specific skills that is hard to get unless you get the job!!

Most engineers come in at graduate level and work their way through the system. The only others I have seen move into firmware are software engineers, often with an embedded background who also started at graduate level.

 

[dpaul]

Most jobs ask very specific skills that is hard to get unless you get the job!!

As I know it, while still within the university, students do internships (~6 months) to obtain a basic experience. This gives them the initial exposure. Then it is nice to do the final year thesis (~6 months) in the same domain, which helps to obtain more in-depth experience. So by the time they graduate, they already have 1 year of exposure to many of the phases involved in an ASIC/FPGA design along with exposure to various EDA tools + tips & tricks.

Regarding the experience, if the job market needs experience and you don't have any, how can you obtain this experience?

Yes, this is a very specialized domain and the companies always seek people with experience because I think the learning curve time here is steep and spread over a decent amount of time.

For you I can suggest that buy a cheap FPGA dev board at home and try out some simple projects. Understand the entire FPGA implementation flow. Then try to do some volunteering project work in *small* some company in the FPGA domain (it would be very difficult to get into ASICs). After that you can figure your way out how to proceed further.

There might be some projects in OpenCores that you can contribute to. This can bring in value to your CV (nice for those do not have any industrial experience).

 

[David83]

Most engineers come in at graduate level and work their way through the system. The only others I have seen move into firmware are software engineers, often with an embedded background who also started at graduate level.

Yes, I noticed that people with graduate degrees are asked for less experience years. I might consider getting a master in Computer Engineering to get professional education on the subjects.

- - - Updated - - -

As I know it, while still within the university, students do internships (~6 months) to obtain a basic experience. This gives them the initial exposure. Then it is nice to do the final year thesis (~6 months) in the same domain, which helps to obtain more in-depth experience. So by the time they graduate, they already have 1 year of exposure to many of the phases involved in an ASIC/FPGA design along with exposure to various EDA tools + tips & tricks.


Yes, this is a very specialized domain and the companies always seek people with experience because I think the learning curve time here is steep and spread over a decent amount of time.

For you I can suggest that buy a cheap FPGA dev board at home and try out some simple projects. Understand the entire FPGA implementation flow. Then try to do some volunteering project work in *small* some company in the FPGA domain (it would be very difficult to get into ASICs). After that you can figure your way out how to proceed further.

There might be some projects in OpenCores that you can contribute to. This can bring in value to your CV (nice for those do not have any industrial experience).

I agree. Unfortunately, I didn't do internships in my BSc degree. But if I pursued my master in Computer Engineering, I would definitely do this option.

Yes, I'm considering buying a board just to familiarize myself with how the design process goes.

The idea of projects is interesting, but with my beginner level experience in Verilog, I'm afraid I will be hindering right now. I think I will spend the couple of months ahead in reading as many materials as I can on Verligo, Digital Logic Design and FPGA, and then I'll go for this.