So, I was thinking that the prop is ideal for software interface to many peripherals, like sensors and actuators and preprocessing many fast data stream; and the RPi is't;
and the RPi is ideal for providing a GUI, communications, hardware float, and a nice camera interface; and the prop isn't;
but both can have a very nice compatible high speed communications channel, so they will fit together in for example a spider bot with a vision system.
I was thinking the GA144 would ROCK in such a system, since (maybe) we can run the video through it as fast as we can without it breaking a sweat, and use it as a vision accelerator.
So would the Parallela be maybe for the AI engine? Or would this be for a neural network? Just to get the conversation on the design started....
No idea how many logic blocks you get there, must check the data sheets, but I did read somewhere that it is configurable using the same ISE tool used for other Xilinx FPGA's. I guess that does not run on the ARM though. However the idea seems to be that the ARM boot's and downloads the logic block config file to the logic blocks and away you go.
The Zynq 7010 have 28k logic cells, (about 430k gates). This is quite a lot. In some projects, I used a 10k logic cell FPGA, it was enough to embed a 32 bit processor and a few peripherals (servo and dc motor driver, sensor conditionner etc...).
This device will be able to run linux with full ubuntu distro, you should also be able to reprogram the fpga on the fly to create custom peripherals.
I pledged 99$, its really a deal compared to other Zynq board like the Zedboard which is about 300$.
28K logic blocks. Hmm... So I can put a ZPU in there.
What I'm wondering is how well the Zynq runs things like Ubuntu. I see no mention of a GPU to speed up X Windows and web browsing. The Raspberry Pi runs these things well but will be even better when they have the acceleration of the GPU working with them.
Anyone with hands on experience of a Zynq board care to comment?
Seems you need to use the logic blocks to drive the HDMI output and then you need some more to drive the Epiphany chip so at least some of that space is reserved.
That does not say so much or give a feeling for how well a web browser, say, might run there.
I'm sure I read somewhere that the Parallella board will boot up Linux and from there download the logic config file to the logic blocks, no need for any JTAG or USB connection. Od course I can't find that paragraph now, perhaps it was in a video.
I still wonder if the ISE tool runs on an ARM though, and would that be fast enough to be usable.
One interesting thing on that page is that they have a Zynq Linux image runnable under QEMU on the PC. I'll be giving that a try out in the next few days.
I missed the deadline, but reserved a board on the newly parallella forums/site. Let's see what happens. I'd like to see a couple of high res photos of the thing... we all love/need hw pr0n :-)
Comments
and the RPi is ideal for providing a GUI, communications, hardware float, and a nice camera interface; and the prop isn't;
but both can have a very nice compatible high speed communications channel, so they will fit together in for example a spider bot with a vision system.
I was thinking the GA144 would ROCK in such a system, since (maybe) we can run the video through it as fast as we can without it breaking a sweat, and use it as a vision accelerator.
So would the Parallela be maybe for the AI engine? Or would this be for a neural network? Just to get the conversation on the design started....
The Zynq 7010 have 28k logic cells, (about 430k gates). This is quite a lot. In some projects, I used a 10k logic cell FPGA, it was enough to embed a 32 bit processor and a few peripherals (servo and dc motor driver, sensor conditionner etc...).
This device will be able to run linux with full ubuntu distro, you should also be able to reprogram the fpga on the fly to create custom peripherals.
I pledged 99$, its really a deal compared to other Zynq board like the Zedboard which is about 300$.
Adapteva has a list of ideas for us: http://www.adapteva.com/white-papers/104-parallel-computing-projects-for-next-summer/ many of those seem applicable to the world of gadgets and robots that the Propeller lives in.
FredBlais,
28K logic blocks. Hmm... So I can put a ZPU in there.
What I'm wondering is how well the Zynq runs things like Ubuntu. I see no mention of a GPU to speed up X Windows and web browsing. The Raspberry Pi runs these things well but will be even better when they have the acceleration of the GPU working with them.
Anyone with hands on experience of a Zynq board care to comment?
Seems you need to use the logic blocks to drive the HDMI output and then you need some more to drive the Epiphany chip so at least some of that space is reserved.
http://www.xilinx.com/products/zynq-7000/linux.htm
I think that the FPGA is programmed the same way as other Xilinx FPGAs, via JTAG and a USB connection.
That does not say so much or give a feeling for how well a web browser, say, might run there.
I'm sure I read somewhere that the Parallella board will boot up Linux and from there download the logic config file to the logic blocks, no need for any JTAG or USB connection. Od course I can't find that paragraph now, perhaps it was in a video.
I still wonder if the ISE tool runs on an ARM though, and would that be fast enough to be usable.
One interesting thing on that page is that they have a Zynq Linux image runnable under QEMU on the PC. I'll be giving that a try out in the next few days.
The Xynq ARM cores do have NEON extensions to help with multimedia demands. But no GPU.
I am already beginning to regret not buying one.
Now what?