Actions
Forum Thread Details
  • Replies 15 replies
  • Subscribers 562 subscribers
  • Views 2738 views
  • Users 0 members are here
Related

Is it worth it to kickstart an open-source FPGA IC project?

manili

Hello all,

 

We know that FPGA itself is a kind of ASIC, i.e. some companies build ICs which have the capability to be reprogramed. We have open-source CPUs, GPUs, Bus Controllers and so many different types of IP cores in the community. Why there is not any sort of open-source FPGA ICs out there, is it because of it's complexity to design or is it even not worth it?

I just saw a project here which was about creating our own EDAs for FPGAs but the project suffered from lack of informations about the FPGA internals so the creator (@johnbeetem) did some reversed engineering, as far as I understood. I asked this question in a Quora comment and Kyle's answer helped me a lot but I need more info, please.

 

Cheers,

Mohammad

  • in reply to michaelkellett

    Of course I'm not going to kickstart such project myself. Surly I do not have such ability even in theories!!! I'm just wondering to know if such project has its own benefits?

    As I watched the company's ads, it looks like Achronix is going to target startups for its new product, because the company claims that the product is 10x cheaper than real FPGA packages. However I did not find anything about the product's price yet.

     

    BTW thanks a lot for the reply.

  • in reply to manili

    I hadn't come across Achronix recently, I've been away from the FPGA world for too long clearly.... but.... Wow, those Speedster22i FPGA's are seriously impressive looking parts! So to me it looks like the eFPGA is essentially the basic Speedster22i FPGA fabric as an IP core which you can then incorporate into your own custom ASIC with your own mix of hard IP cores. So, yes it'll bring down the cost of developing an FPGA from scratch, but I think is right, it's going to be extremely costly still so not in range of the Open Source community (without backing of a large corporation willing to fund OSHW development) or small startups (without the backing of venture capitalists with deep pockets).

     

    I need to do some more research on who Achronix are and what they offer though, they look very interesting. Thanks for bringing these parts to my attention! image

     

    Best Regards,

     

    Rachael

  • in reply to rachaelp

    Well, I heard Intel is going to bring on-chip FPGAs (Altra) to its Xeon processors for high-end usages (e.g. datacenters). But I agree that this one is much more customizable at lower price (however we should not forget the cost of fabrication for customers).

    I think it's not a bad idea to take a look at Achronix's documentation about eFPGA.

  • in reply to johnbeetem

    I fully agree with you on this issue. For a student, this may take about 1 year of work. After all, this is a very difficult topic. I was working on a similar theme, I thought on the title. But in the end I found a formed topic on the site page with topics with business research proposal topics, because there are the best topics for all academic disciplines. You can find topics for essays, dissertations and other topics.

  • When considering whether it’s worth it or not, might be worth considering how long it took before RISC-V was a worthwhile addition to the microprocessor world.

     

    My feeling is pretty similar to several of the other replies - start with working up an open tool chain, worry about open hardware later.