Topic: New Official AMT Thread - page 107. (Read 149454 times)

Did they also license the Scrypt based design?

What he said....+1

Overall questions though is timeline. We need to know this is going to be ready in a very fashion and if not what compensation is going to be offered for our trouble. But from Marto's post it looks like they are just using his reference design and assembling the hardware. Like you said might not be a bad thing if they don't deviate from the design and just get us working hardware ASAP. With the questions above answered and a realistic timeline on delivery we can at least weigh our options better.

That is not necessarily true (e.g, it could be great news, depending).  A company like Flexitron could probably do 1000 or many times that boards in a day without breaking a sweat.  Probably much faster, with better QA and at lower cost, than even a high quality specialized vendor could do.  The problem is in setting up the assembly line, and lead times.

So the questions to ask are:

(1) Is the techno bits licensed design fully developed, tested, and ready to go, or will a testing phase be needed?  IS the PCB design already in layout ready to be printed?

(2) Who is doing the PCB board manufacture and soldering?  Do they have all parts in hand or will there be a lead time and if so what is that?

(3) When will assembly start and how many boards can be finished and tested per day?

(4) What are specs on the techno bit boards?  Are they a design that is already being sold or a new AMT only design?

(5) Are there any post-board assembly tasks that will not be performed by the assembler?  E.g., will the hashing chips need to be mounted separately, or will AMT ship those out to be mounted too?  and how long before assembler has all parts in hadn and can start therefore?

(6) Has contract been signed with assembly shop or is it still under negotiation?  Are there hard and fast dates and volumes in the contract or is it a reasonable efforts type contract?

Thanks for the info Marto

This might do it. http://www.digikey.com/product-detail/en/ATS-52230P-C2-R0/ATS1134-ND/1284848 Certainly more surface area that the ones BM/AMT used, just be sure to check its base size to get one that covers the entire top of the chip. Again, due to gravity highly suggest having the boards horizontal.

Oh man.. not looking good!  

AMT entrusted to do the production and assembly!     I would have preferred that they ship units directly from technobit.   The delay to ship from Europe would likely less than the delays that AMT is known for.

AMT decided to do this only 2 weeks ago?  What the hell were they doing while our systems were known to be not working?   Now they take again the long route by saying that they'll assemble it themselves!

BTW, how many boards will be needed to reach 1.2 THs?

HI,all,

Actually AMT purchase a  non-exclusive licence from us about 2 weeks ago.
They did business with us in the past, so having troubles with their current design/manufacturing process , they decided to get an alternate design.
In any case all their boards will be produced in USA as  far as I know.
We will not be involved in the production and assembly except with know-how help if needed from distance.


Best Martin

This would work to cool the chip.  Will it fit when you include fittings (alas, I don't even have non-working parts to look at)?

http://www.frozencpu.com/products/13206/ex-blc-916/Alphacool_HF_14_Smart_Motion_Universal_Chipset_GPU_Block_-_Copper_Edition.html?tl=g30c89&id=wo6LsxuJ

The adjustable mounting bracket should make it possible to create some kind of jerry rigged mounting solution, though right now I can't visualize it without seeing the board and the existing backplane.  Also, even without the mounting brackets, you can use thermal GLUE.  I did this with my BFL 80 and just using thermal glue I got the chip temps down from 80c to 45c.  Of course that is not a reversible solution lol.

Also interested in the timelines, and what to do with our RMA's, please AMT include that in your posting as well.

At this point its a safe bet they are not using that old design any more. Seems like its more trouble than its worth. Seems like everyone with a miner has problems of some kind AMT already owned that. On the off chance there might be one lucky person with a working miner but likley harder to find than bigfoot. But at this point noone we know of has 100 percent working hardware. Might be easier to just focus now on getting the technobit solutions (or whatever option you go with). I am reading the forum on this https://bitcointalksearch.org/topic/ann-technobit-hex8a1-240-ghs-coincraft-a1-board-finished-production-376351  And it seems like a solid solution. If we are getting the stock design then great....

They were supposed to send out an email to us or some option on their site to make this all official today........AMT any updates on the site change or email?
Also Timelines and process for the delivery of new hardware? Thanks

Technobit might actually spend money on a real box too.

The most relevant question now that we know what we will be getting for option 2, is how long? This situation is just a mess and we are being forced to wait even longer. So a clear cut commitment from AMT on how long this will now take (and something we can hold to account). Now seeing as this is being sourced to a 3rd party vendor Maybe they can chime in and provide some input on how long this will take. Being that its Technobit they can give some timeline.

 Honestly I would rather just have the hardware shipped directly from them to me than go from technobit to AMT to me. Seems like it just adds more unnecessary time to the whole process. If Technobit just ships us the miners directly it will save insane amounts of time. Like I said before this way the support matter now gets passed to technobit to address should there be any issues. Considering the shipping issues people have experienced it seems like a more viable option for us to just get it directly. At this point it would be a paperwork issue with AMT and Technobit confirming orders fulfilled and checking with customers. Frees AMT up from alot of added hassles.

Solid copper spacers to raise ASIC level above other components.
However the small SMT components may then overheat due to lack of airflow.
Immersion is the only cureall for this design of boards top level.

Holddowns for the small sinks as suggested world take a board revision to accomplish.
I think since that chip is isolated (other chips show no signs of overheating), was a short from dislocated backplane or other damage occurred during ship.

Come on guys, get serious. This is the heatsink for my first $300 10Mh/s Bitcoin board. What do you think you should use on a 1.2 Th/s system? Don't go cheap on a $6,000 miner.

I guess one could customize it, cut the corners off... (not suggesting you per say).

5 square inches?  The problem is that there are capacitors in the way of the chip, so you can't mount a big one.

Here is photo:



See the four orange capacitors that's in the way of mounting a sufficiently large heat sink?  Anyway around this?

Yea, not that size, but it should have a bar like that or screws on the corners to stay secure. You would never use JUST thermal tape on something that would smoke and burn if it fell off. You might use a tiny heat sink on something like a Pi board that just needed a little help, but could run fine without the heat sink.

At least would stay on but needs at least 3x dissipation area than that support chip sink has. In free air I'd start with a guess of 5 sq. inches of surface PER-CHIP.

Can anyone point to one that will work?

I guess a this time it is not advisable to run AMT design without having a sufficiently large heat sink to transfer the heat.

Retarded.

You can blame the companies AMT hired to do work on these all you want, but eventually the blame falls squarely on AMT for their incompetence. They took everyones money and picked the manufacturers to build these machines. They are 100% responsible.

Anyone that is hardware savvy knows that heatsinks need to be clamped down tightly. The fact that they thought they could use thermal tape to stick the heatsinks on these chips shows complete ignorance of technical skill.

I'm sick of reading all these excuses both Opieum and Fuzzy want to make for AMT. They are responsible for this mess, regardless of who dropped the ball on the design and/or execution.