Klondike - 16 chip ASIC Open Source Board - Preliminary

[Bicknellski]

Please guys to save everyone time... read through the whole thread just BKKcoins posts if necessary to find the info. Took me 1 minute.

Will you provide an assembly service as well?

Now that you are hoping for a few different boards, can you give so updates for cost with and without parts? I am strongly considering going in on the funded ASIC group buy, and I am just trying to figure out how many I can afford. I am not looking you to lock into a price, maybe just a ballpark. Especially for the 64 chip board you mentioned.

If I had to ball park a kit price right now I'd throw out numbers like:

ALL QTY 1 --- SUBJECT TO CHANGE

(board + parts) (without ASIC, heat sink, ATX PSU, cables)
Klondike 1 - $16  (I did a design and parts list this evening, and board partly done)
Klondike 16 - $35
Klondike 64 - $100

(board only)
Klondike 1 - $6
Klondike 16 - $12
Klondike 64 - $25

PDF Assembly Manual included (hopefully). Shipping not included.

I'm not planning to get into full assembly myself except maybe for the Klondike 1 (dongle).
I will supply boards/kits at wholesale pricing to other assemblers.

If you are in Indonesia and have chips or are interested in finished boards I am working on using a Japanese company based here in Jakarta for assembly of chips to boards and I am very keen on working with Bkkcoins to get the whole kit and kabooddle done here in Indonesia minus the chips of course on more regular run where I will have stock on the shelves that people can order NO PRE-ORDERS NO BACK ORDERS only taking money for things in stock will be how I work this. However will be just for Indonesia. I would suggest to anyone doing it as a single board to a few boards to learn about reflowing chips to boards also mentioned in the thread here or finding a plant or shop that can help you out. Get your equipment and start practicing. I will actually use 2 to 4 boards at school and do this with my High School students this coming fall. We also plan to visit the fabrication plant hopefully we can actually see Klondikes being made.

https://www.sparkfun.com/tutorials/59
http://www.ti.com/lit/an/sloa122/sloa122.pdf
http://www.ti.com/lit/an/slua271a/slua271a.pdf

[Bicknellski]

Heatsinks?

I know you are still reworking the board for more important aspects but have you given more thought to the attachment of the heatsink Bkkcoins? Will we more likely to be bolting a single Aluminum heat sink to the back of the board? Where will the through holes be?

[BkkCoins]

Also, exactly how scaleable is the klondike system? I know you have the k64. Would each k64 just plug into a different USB port on the host machine, or is it possible to just continue daisy-chaining the cards indefinitely, say, for example, into a 320 chip system, or something. Is there a point where, A: we need to use a separate USB port (to avoid signal degradation, or whatever), or B: where we just need to dedicate a new host machine to the system?

Sorry if this has already been answered.

The K64 is just 4x K16s on one board so the two can interconnect as if the same. For now I'm considering the limit on chaining the boards to be 8 K16 boards. I think that's a dependable number. However, it could easily be more as some people report having 25x I2C LED boards on an bus. I'll do testing when some are built. Regardless, at some point I'll put together a tiny I2C boost jumper that slips over the pins joining boards and expands this number. Also, initially I'll stick with 8-bit addressing and on I2C I think this means just a bit under 128 boards could be chained in theory, off a single USB connection. I doubt anyone will do that, and it probably makes more sense to run about 16 boards off a USB, and then use a USB hub to expand further, seeing as how they are dirt cheap and let you use the higher bandwidth of USB vs. I2C.

My question is how do you return the data from the ASICs? This might be the big 'unknown' until we have specs, but it seems to me passing that data back through the PIC would not be feasible. Can the data dump be passed directly to the USB controller?  (I have about 15 other scenarios but would zzZZzz this thread so I will save them)

Also, I am in for some serious bulk PCB funding (cash in the Ks USD or BTC), but I would need to see a working prototype (even if its on a test board) before committing.  

The result data gets back from the ASIC to the PIC via a secondary route as serial data. It is only a 4 byte nonce (I believe), and so is a very small amount to return back to the USB host.

I don't want to even talk about selling boards until I have a working prototype but everyone keeps asking so we have a lot of discussion about sales - that day will come and deals will be made. I will try to do what I can to get boards into assemblers hands so that miners can just buy boards and hook'em up.

Heatsinks?

I know you are still reworking the board for more important aspects but have you given more thought to the attachment of the heatsink Bkkcoins? Will we more likely to be bolting a single Aluminum heat sink to the back of the board? Where will the through holes be?

I'll try to put a drill file on github (but design could change if there is a better way to do it).

There is 4 corner holes and 4 spaced between the ASICs with the intention that the heat sink can be clamped down above each set of 4 ASICs. I think probably the 4 between ASICs are better suited for attaching the heat sink, and the corner ones for stacking with spacers or bolting to a case. But it could work out that bolting down the heat sink at the corners is feasible too. If the fins are cut back a bit there then spacers could be mounted on together with heat sink.

When I have a prototype I'll attach some test heat sinks and see what works best, and test with the new IR thermometer gun that some kind member here bought me.

BTW I'd welcome any sample/test heat sinks users want to send me so I can try them out and measure temperature with different air flows, etc.

[bassclef]

When I have a prototype I'll attach some test heat sinks and see what works best, and test with the new IR thermometer gun that some kind member here bought me.

I'd like to add... if there's anything you need, please let us know! I'm sure it's the least we can do for the time you're putting in.

[Bicknellski]

When I have a prototype I'll attach some test heat sinks and see what works best, and test with the new IR thermometer gun that some kind member here bought me.

I'd like to add... if there's anything you need, please let us know! I'm sure it's the least we can do for the time you're putting in.

+1  I think anyone with a larger order of chips should see if they can shunt the "test" chips BKKcoins way if at all possible. I am trying to determine what if any Avalon test chips I might be getting from the group buy. That would be a priority I think as the sooner he gets the testing chips even if he is not ready for it the better.

[steamboat]

When I have a prototype I'll attach some test heat sinks and see what works best, and test with the new IR thermometer gun that some kind member here bought me.

I'd like to add... if there's anything you need, please let us know! I'm sure it's the least we can do for the time you're putting in.

+1  I think anyone with a larger order of chips should see if they can shunt the "test" chips BKKcoins way if at all possible. I am trying to determine what if any Avalon test chips I might be getting from the group buy. That would be a priority I think as the sooner he gets the testing chips even if he is not ready for it the better.

BkkCoins is scheduled to get sample chips from batch 1 from my purchase.

[Bicknellski]

Nice one steamboat +1. Thanks from all of us!

Oooh getting so excited. Wish I could help more.

I am going to see if I can get some of those VGA style heat sinks aluminum extruded to you BKKcoins. Just need to track down a few of the manufacturers here in Indonesia. Maybe another aliexpress shipment from China headed your way if I can swing it.

[BkkCoins]

Thank You, Bicknellski and steamboat.

Updated Github.

Added new items to parts list for LEDs, Thermistor, resistors and ASIC VDD ref. supply.
Changed capacitors to match Avalon specs and quantity. May have to adj. again soon.

Added incomplete Gerbers to k16/plots directory. These are Gerbers for the real board but expect changes as I know at least a few items missing and a few I'll be moving. The drill file is there so  you can use that to locate accurate position info, which may change so don't treat as final. And the layers are there - but incomplete as I've hardly done any data signal tracks on the inner layers, and power supply still needs several changes. But this gives people some pretty pix to look at in better details than 3D renders.

Enjoy.

[hashblue]

The result data gets back from the ASIC to the PIC via a secondary route as serial data. It is only a 4 byte nonce (I believe), and so is a very small amount to return back to the USB host.

It looks like the hash chip sends back the report data asynchronously. Is a PIC controller capable of concurrently clocking in results from multiple chips?

[quiksilver911]

The Klondike has 16 ASICs but I have split them into 2 banks of 8 each. This allows pushing the data in twice as fast, and also means if one ASIC is damaged then only 8 cannot function, instead of 16.

I hope there is a mistake in what you have said, because it scares me a lot. We don't know what the yields are on Avalon chips, and I would seriously prefer a solution where only the damaged ASIC is the one that does not function. Losing a bank of 8 is significant, imho.

Also, what happens if a bank is not full, e.g. 6 chips instead of 8? I imagine there are plenty of people who don't who's total number of chips don't divide nicely by 8, and if we are shipped untested chips then yields could make that worse.

[Amph]

same thing
tell me when this is ready

[BkkCoins]

The result data gets back from the ASIC to the PIC via a secondary route as serial data. It is only a 4 byte nonce (I believe), and so is a very small amount to return back to the USB host.

It looks like the hash chip sends back the report data asynchronously. Is a PIC controller capable of concurrently clocking in results from multiple chips?

It's a wired OR bus, all outputs join together into one PIC input. The PIC does have a UART and amazingly I just hooked REPORT_P to that pin. What's even better is it has IOC (interrupt on change) so I don't have to poll it.

The Klondike has 16 ASICs but I have split them into 2 banks of 8 each. This allows pushing the data in twice as fast, and also means if one ASIC is damaged then only 8 cannot function, instead of 16.

I hope there is a mistake in what you have said, because it scares me a lot. We don't know what the yields are on Avalon chips, and I would seriously prefer a solution where only the damaged ASIC is the one that does not function. Losing a bank of 8 is significant, imho.

Also, what happens if a bank is not full, e.g. 6 chips instead of 8? I imagine there are plenty of people who don't who's total number of chips don't divide nicely by 8, and if we are shipped untested chips then yields could make that worse.

Just like the Avalon, the Klondike will work with how the ASIC was designed to be used. Instead of 10 chips/bank I'll have 8. For partial fills I haven't decided yet what I'll do - either I'll have to divide the range by N chips, or I'll just ignore that and divide by the next higher binary multiple, but only shift N sets of data. While it won't check every single hash in the range the probability of finding one doesn't change. You just have a lower hash rate due to less chips.

[quiksilver911]

Just like the Avalon, the Klondike will work with how the ASIC was designed to be used. Instead of 10 chips/bank I'll have 8.

Sorry I haven't followed Avalon's design closely. Could you clarify that Avalon would also lose all 10 chips in a bank if any one of them failed or was damaged?

[BkkCoins]

Just like the Avalon, the Klondike will work with how the ASIC was designed to be used. Instead of 10 chips/bank I'll have 8.

Sorry I haven't followed Avalon's design closely. Could you clarify that Avalon would also lose all 10 chips in a bank if any one of them failed or was damaged?

Yes, to the same extent, as I use the chips in exactly the same way. This wouldn't happen if the hash engine was damaged but would if the data path from input to bypass output was broken somehow.

edit: Note that technically it would only lose those after the dead chip in the chain, so some random number from 1 to N.

The chips are supposedly 100% tested at the factory. I think TSMC is trustworthy that way, seeing as how they're a huge tech company supplying most of the worlds other huge tech companies. Presumably a board assembler would also detect this during testing.

OTOH, doing DIY reflow at home is more likely to experience problems as getting the reflow profile perfected to assure consistent results may be more than most can manage.

[hashblue]

It looks like the hash chip sends back the report data asynchronously. Is a PIC controller capable of concurrently clocking in results from multiple chips?

It's a wired OR bus, all outputs join together into one PIC input. The PIC does have a UART and amazingly I just hooked REPORT_P to that pin. What's even better is it has IOC (interrupt on change) so I don't have to poll it.

Are you assuming only one chip will report data at a time (i.e. data is never being sent concurrently by more than one chip)? I have a feeling all chips will report some kind of status not just when a good nounce was found. But I guess we won't know for sure until the communication protocol is released.

[BkkCoins]

It looks like the hash chip sends back the report data asynchronously. Is a PIC controller capable of concurrently clocking in results from multiple chips?

It's a wired OR bus, all outputs join together into one PIC input. The PIC does have a UART and amazingly I just hooked REPORT_P to that pin. What's even better is it has IOC (interrupt on change) so I don't have to poll it.

Are you assuming only one chip will report data at a time (i.e. data is never being sent concurrently by more than one chip)? I have a feeling all chips will report some kind of status not just when a good nounce was found. But I guess we won't know for sure until the communication protocol is released.

At this time I believe it reports only found nonces and the chances of collision are slim. I'm basing this off how Icarus worked. Looking at the driver it's likely it also reports back the full midstate and data, which I can ignore since I have that in RAM already. If the driver was smart it wouldn't need that info back either. ie. it stores that info for each known miner module. I think we have to assume this since Avalon has all 10 ASIC chips wired this way. They all connect to a 470 ohm resistor tied high and a buffer to go off board to the FPGA. If the ASIC is "smart" it can detect traffic by watching before using the bus, but I don't know if it does that. It doesn't seem very efficient either to send the midstate and data all back again but if the FPGA cannot store it for each module then it would have to since the Avalon driver expects to read it back.

[uncaer9]

Would I be able to use one OCZ ModXStream 600W psu both for one Klondlike and my gaming PC(with or without OC)?
Intel i5 2500k 4.2GHz @ 1.25v
GIGABYTE GTX 670 OC - currently at stock but I'd like to able to OC it
Asrock P67 Pro3
Xonar D1
and Seagate 7200 1TB HDD

[Bicknellski]

Would I be able to use one OCZ ModXStream 600W psu both for one Klondlike and my gaming PC(with or without OC)?
Intel i5 2500k 4.2GHz @ 1.25v
GIGABYTE GTX 670 OC - currently at stock but I'd like to able to OC it
Asrock P67 Pro3
Xonar D1
and Seagate 7200 1TB HDD

Summary Preliminary Specs:

10cm x 10cm board
16 chips/board
24A 1.2V, 1A 3.3V supply on board (~32W total)
PCI Express Power connector for use with ATX PSU
USB mini type B connector, no power draw
PIC USB micro controller on board with bootloader for USB firmware upgrades
I2C board-to-board, daisy chain connector
Fan 3 pin connector (on board temp sensor and PWM speed control)

http://extreme.outervision.com/psucalculatorlite.jsp


Input the numbers and let us know if you can run one board. Just add ~32W total to your total and see if you are under your 600W with some to spare. Do you have any Avalon chips?

[uncaer9]

My PSU has two power rails and the second one is already heavily used. Guess I'll have to buy another PSU to be sure that my computer is stable, It's not a big deal if I buy used one or make one myself.

[tom_o]

My PSU has two power rails and the second one is already heavily used. Guess I'll have to buy another PSU to be sure that my computer is stable, It's not a big deal if I buy used one or make one myself.

You could probably just use a 12v laptop charger they're normally very consistent voltage wise and around the right amperage range for a board or two. Would I be correct in this?