You'd need to figure out a way to mount the blocks, as well as still provide some airflow over the board to cool the power supplies etc. We've not released dynamically clocking bitstreams at this point, but it is in the works. So far I've not been able to synthesize anything that'll run over 200Mhz in Xilinx's ISE. You could push it even if the ISE tells you it might not work, but there's no saying if it'll work properly. If that changes I'll post an update. I've yet to run all the possible instances under ISE, I'm only dedicating 2 of my PC's possible 4 cores for running this at the moment because I still game on it.  |
|
|
|
|
Please be very very careful with the 200Mhash bitstreams in fact anything above 166 is serious business. You'd probably need to remove the stock tape and thermal epoxy the heatsinks on or use a heftier heatsink.
Whatever you do, you still need to maintain some airflow over the board as the Vregs are getting warm too. So even if you so decide to watercool the FPGAs you'd need a low speed fan to push some air over the Vregs.
|
|
|
|
O_shovah: you sure you use the thermal imager just for PCBs? xD hehehe But yeah it'd be interesting to see. Fizzi: told ya thermal epoxy makes a world of difference. I've been successful taking mine off by popping it off with a wide plastic spatula between the FPGA and the heatsink when it's warm. This way you don't put stress on the balls. You might crush the FPGA itself with too much force though. So it's still not a 100% solution especially if the epoxy is strong. I was using this http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htmI mixed up a small amount on a disposable spoon. Applied it to the center of my heatsink and put a 1KG weight on top of the heatsink overnight. |
|
|
|
340.93 MH/s | 0: 22569/123/121 0.5%/0.5% | 1: 22708/161/125 0.7%/0.5% | 1w29m56s | AH00WOWI Anyone (besides fizzisist) successfully running the 180Mhz bitstream? Care to comment on your speed and what cooling you're using? Given how stable the 166Mhz bitstream has been for me, not sure if the upgrade is worth it. I'm running a 190Mhash bitstream per FPGA. Which gives me 380 total. My heatsink isn't anything special, however it's attached on with thermal epoxy instead of tape and has a high performance fan blowing on them. I see alot of posts requesting for mounting holes, rest assured they will come in future board revisions. I'm thinking of working out some cheap hardware that'll enable users to clamp various heatsinks onto the current board, until then thermal epoxy might be the next best solution. |
|
|
|
|
I've noticed the Mhash display in software sometimes goes crazy and displays an insanely high value, only to freeze there for a moment before dropping back down. The timer still works all the time though, not sure what"s going on there.
|
|
|
|
|
If you're still running the boards passive, it'll probably be a good idea to put on a fan to get adequate airflow over the heatsinks. If the board feels warm under the FPGA, it's probably not cooled enough.
I'm getting about 4% rejects on 190Mhash.
|
|
|
|
|
You'd probably need to run smartxplorer to try and automate the process. It is running fairly close to the max as it stands.
|
|
|
|
|
Pre-fix was getting 20% avg rejects. Now I'm down to under 3%.
|
|
|
|
|
Might be a good idea to just use a lower rated fan instead of trying to use a resistor, it won't work all that well because the fan is not a fixed load, it requires more initial power at turn on to start it spinning.
|
|
|
|
|
I have found that forcing the getwork interval back to 20 secs (it's now default at 30 I believe), decreases my % of stales. Not sure if it applies to everyone, but you might wanna give it try.
|
|
|
|
|
Hi all,
Apologies for taking so long to come up with a response, I've been real busy with work these few months.
I've managed to synthesize a bitstream that should net you 120Mhash. It'll require further testing, but it should work. I'm working with Fizzi to get it hosted on his site, but meanwhile until that is done I could email you guys individually if you'd PM me with your email address.
|
|
|
|
|
Do you have the D2xx drivers and the right version of pyusb installed for your version of python?
|
|
|
|
No code improvements for higher hashrate yet?
Sorry for the very late reply. Unfortunately, no improvements yet. Once these boards are finally on their way out to customers, we expect to be able to devote a lot more time to that work. When updating software, can you keep in mind your first customers that have x5000? Yes, absolutely. Unfortunately, I really don't know much about the X5000, because it was before my time. We need fpgaminer's help, but he's mostly MIA lately because he is swamped with work and moving to a new apartment right now. Hopefully he will be able to work on this stuff soon. I do know that if you could switch from the Xilinx platform cable to an FT232R breakout board, the hardware and software interface would be identical to the X6500, so all code would be completely compatible. This would also eliminate the need for ISE. Actually, this cable soldered to the correct connector would probably be perfect. If you were willing to pay for the cable, I'd be happy to solder it up and send it to you. At that point, I think you would just plug it in and run the X6500 software (with the X6500 bitstream) and you'd be off and running. fpgaminer, can you confirm that my idea makes sense? Do you have other ideas for how to maintain support for the X5000? fizzisist, I would happily pay for the cable if that method works. I hate that "xilinx platform cable" anyway. You would have to ship it to germany, but it should probably fit in a letter, right? If shipping is prohibitively expensive, I can solder myself given you tell me what goes where. fpgaminer, would that work? Hm, doesn't seem fpgaminer is reading this... Hi Molecular, yes it does seem like this is the easiest way to maintain support for the X5000, although you'd have to spend a little bit for more hardware. |
|
|
|
Just so it's clear, these are the first 5 PCBs, which need to be stress tested to prove that our design has no problems. After that's done and all looks good (knock on wood), we will be placing large orders. We'll keep you updated! Do you know when you will allow pre-orders again? Timeline for 2nd wave? We're still concurrently working on the logistics of shipping the orders. Look forward to the news soon. I believe that sorting it out properly will result in better pricing and availability. |
|
|
|
Received the parts, populated the board and now onto testing. Looks good so far hardware-wise, software side of things it looks like we need to iron out some small issues and we should be set.  |
|
|
|
|
Electric: why would you need the jtag connector? The board has jtag for debug purposes only, everything should be handled via USB during real world usage.
Not too sure if you can mount 2 boards in a bay without needing a high CFM fan to push air to cool the FPGAs. Maybe it'd be possible if we can find some custom or weird heatsink that has 2 flat sides and then mount the top board upside down onto the same heatsink, and then put 2 fans to blow air across, if you get what I'm thinking.
|
|
|
|
|
Hi there,
Yeah it's 12V input to the board and 5v to the fan for the X5000. Anywhere between 6V and 19v should be fine.
Without the fan the heatsink does get quite warm, warm enough to warrant the fan.
|
|
|
|
Quick suggestion - it may be worthwhile to offer some boards with SATA power connectors.
I've thought about it, however at that time I couldn't find on digikey any male sata power connector alone (they were either the combined data + power, or data alone), might just be me using the wrong search strings. I have also not seen any sata splitters around locally, which led me to use a molex, the next best solution would be to see if we can get Cablesaurus to stock SATA to molex converters. oh btw are there plans for cases yet?
I've not put alot of thought into this yet, I don't personally feel any1 will appreciate a case, and if I can't somehow pickout a already mass produced case and go with something custom, that's gonna add alot of cost to the final product. You're free to do whatever you want to your board though =P. |
|
|
|
Here's a render of the board done in POV ray, I haven't got models of all the parts, so some of them are unpopulated, but this should give us a good idea of how the final product will look like.  Personally I haven't been posting all that much on the forums, due to me working in the background on the board routing, parts and logistics of getting the boards out. Hopefully I'll be able to hang around more after most of that is done with. I should have the 1st few boards within 2 weeks, and then a couple of days to populate and test that out, will update this thread again. |
|
|
|
|
Show Posts
