The first hurdle for any system with an FPGA is: finding and learning the FPGA development tools. For many companies (cough Xilinx cough) that's a huge download, that sits on a PC. If the FPGA Dev tools can downloaded AND run from the Beagle 5, that's a major point and needs to be covered.

Follow up with a "blink the LED" simple type of FPGA graph and cover the tools to synthesize and put into use? Way over my level of "FPGA sounds scary" understanding but I would love to see more examples of this at an introductory level. There must be some people viewing that might become interested in development with FPGA that just don't know how useful it is yet.

Finally the wait is over. Thanks for this masterpiece

Suspect the thermal imaging is viewing the metal/reflected heat rather than the actual surface temperature. Try placing a sticker on the chips and try again?

Cockpit is mentioned a fair bit in the homelab community, I guess it’s a modern webmin equivalent. Really interesting product! Would love to see the watchdog and fpga in action!

I like that management UI, could see it and openwrt getting together and sharing notes.

Really nice it's a modern distribution and kernel.

Thanks for another great video and I look forward to grabbing some cases for my Milk-V Duo's once you get them in your shop. Beagle have been ahead of the curve supporting RISC-V and the inclusion of an FPGA makes this board very interesting 🤔

Thank you for this video. I did not know how to flash it yet...

Thank you for sharing. Looking forward to seeing an example Blinky FPGA dev flow on this system! I'm tempted to buy one if installing a dev system is easy. I spent way too much time setting up Xilinx on my X86 Linux kits.

Never trusted beaglebone since I got one with the Omap processor years ago. There was a bunch of boards that had a revision just before the new model that could not even boot linux. Not cheap either. They just blamed Ti for stopping to support the chipset, and ignored me.

Well, I am in USA Wisconsin and I don't mind buying from your shop if it supports you. You work on lots of the same stuff that I do.

Ohio loves free pcb rulers. 😜 Great stuff man. Im loving the content. Cant wait to play along at home here in a couple weeks. Untill then much love.

Ooooooooooo. Want it!!!

That is a really cool board 😎👍 Would be cool with a video showing how to create some basic fpga 'programs' Tip: Add some black electrical tape when you try to measure the temperature of shiny surfaces with IR

The only gripe I have about this board is the JTAG connector. They could have included micro JTAG connectors like the one on Ox64. Spending 40$ on a Tag-Connect is insane.

isn't `ifconfig` deprecated now?

ha, you are younger than me, get with the 'ip' tools. I know ifconfig is strong with the muscle memory but ip is much more logical once you get used to it.

I really wonder how big and powerfull that FPGA is. while mostly overlooked by most people, a decent FPGA actually should be added in basically every computer which isn't low end. since with a FPGA you essentially have optimal hardware encoding and decoding and acceleration for basically anything, and even if you customize the settings(which normal dedicated hardware can't do). while technically seen pure dedicated hardware is slightly more efficient and fast and so it remains a use for some very heavily and commonly used things which you also use while also using other optimized hardware stuff. a FPGA should be added. for example a FPGA can run AI on levels which a GPU would try to dream of if the AI is also optimized for running on a FPGA, which means that the AI is optimized for working well and as it should rather than designing it for speciffic hardware, the FPGA will just take on the form of the hardware which is optimal for the AI, and while a FPGA can also take on the form of cuda cores and such, the truth is, a AI designed out of the concept that dedicated hardware will be made for it completely and then also using that dedicated hardware can easily be more than 10 time and sometimes even hunderds of times faster than a similar AI designed to run on existing hardware such as cuda cores. next to that a fpga can also be optimized for whatever combination of incstructions is most used, so could greatly reduce the load on a CPU, GPU, and even ram in some ways, essentially it can perform all of those instructions, and in the case of CPU and GPU, it can actually do them better than those, just still would also have a GPU and CPU since doing everything from only a FPGA would require a big FPGA, and so the FPGA would just do the things which could be accelerated the most.

no GNU screen, ok.. tmux?

I wonder if the Beagle will run Klipper?

perth boy?

Nice board. Please stop banging the hardware around, i.e. even when in the box or bag. Place it gently down...

hdmi？

I'm hoping Beaglebone/Microchip don't drop the ball on the IDE for the FPGA. Intel and Xilinx's IDE have been bloated licence ridden behemoths for decades now. I'm hoping microchips IDE is as open as the risc-v architecture.

Just buy a mini PC.

milk-v has worst sdk I ever seen

It took 6 minutes to get to the content. That is a huge waste of your viewer’s time, you should probably be be more respectful of their time. As for me, I’m out and I’ll find another review elsewhere. Thumbs down and a do not recommend your channel.

@PlatimaTinkers You wanna colab?