Ok I will try to answer a bunch of things.

Cabling is as much important as connectors and many of the wires used for the disk drives wiring strings are very thin. These have actually got worse over the years that I have been using ATX power supplies and part of the reason for that is disk drives in general now use a lot less current than 20+ years or so ago when this standard started to be used. I have seen a lot of variation in the thickness in PCIe connector wiring as well but at least you get 2 or 3 strands each of 12V and GND to share the load.

We have added an extra power connector now to the design that will be on Issue 1.1 which is what most people will get that are on preorder. It's overkill for the board but it allows us to do custom wiring easily or even maybe a PCB distributor. No recommendaton but an example is http://uk.farnell.com/xp-power/dnr120as12-i/psu-din-rail-120w-12v/dp/1372699 could be used to power 2 boards. Much bigger units can be bought and may be a better way than ATX power supplies. We also have a way to use multiple ATX power supplies in a larger system. The negative of some of the industrial power supplies is that they have exposed mains voltage terminals and need a proper encloseure when that is the case (no pun intended).

Moving on to software we are going to be some testing this week. Hopefully some on software size and we hope anything that works with Icarus will work will Cairnsmore. When we know more I will update you all on this thread. Same goes for bitstreams.

Pricing for the single and dual version of the board is coming. We are adding some daisy chaining features in the Issue 1.1 to allow a system build up and more on that later when we have that finalised. That may also allow a reduction in USB cables and maybe an Ethernet option eventually. It's likely we will offer the single and dual options for shipment in July onwards.

Programming - we are hoping to do one the following but all are subject to confirmation (1) ship with bitstreams already loaded or (2) ship with a programming cable (least likely as extra cost) or (3) ship with JTAG feature through USD data interface or (4) ship with a direct bitstream loader maybe with a falling back image feature.

To pre order send an email to bitcoin AT enterpoint DOT co DOT uk with address and contact details.

Yohan

To pre order send an email to bitcoin AT enterpoint DOT co DOT uk with address and contact details.

Yohan

Payment will only asked for when the unit is more or less ready to ship. In reality what we will probably do is look how many boards go through test in any one day and ask the next set of people for the money at that point with the idea of packing those units the day after. The only exception to that will be the bypass that we are going to try and give everyone 1 unit on an early shipment.

As the list of people is getting very long we will probably have to freeze the list for the early ship offer very soon.

As said we are looking at Bitcoins but initially we will use our traditional payment methods which are bank transfer, PayPal, credit or debit cards.

Yohan

Pre-order is still open but we may pause it for couple of days sometime this week whilst we take stock on a couple of the more minor components and whilst we see how many we can "grap" either from our other projects or find from the general components market. 

For existing pre-orders there isn't a problem unless we are let down by one of our suppliers. The highest risks at the moment in this respect are the heatsinks and fans which there are other option albeit possibly not so good options compared to those we have selected.

There 3 ways of puting 12V into this board. The first is a jack either 2.1mm or 2.5mm (TBD). There is hard disk drive connector (usually refered to as Molex) and there is also a PCIe (6way) graphics power which incidentially is also made by Molex.

It is also possible to use the disk drive connector as an output to power a second board when using the PCIe graphics as an input. I would not recommend more than 1 power chained this way except when they are the dual or single versions of Cairnsmore1.

Depending on who makes a disk drive connector, and the mating half, the contacts are rated between 5A and 10A. On 12V input that gives a range of 60-120W. We also have a possibility to run on a slightly higher voltage (to be confirmed) say 15V and power then scales 75-150W. It's the current that matters on these connectors.

On power supplies it may be worth considering a single rail industrial style power supply for racks. We are going to look at this and it may be a more cost efficient path for arrays of boards.

Yohan

We will look at something more complex with Ethernet and a few other ideas but we won't do much on this for few weeks or a couple of months. We have lots of non-Bitcoin projects to keep on target as well.

In the short term we are going the other way and going to offer Cairnsmore with 1 or 2 FPGAs only as lower cost entry level system options. It's not quite modular but an option some people might like. I add some info on these in the next couple of days.

The efficiency is much the same between the 2 approaches. But having 48A, say around a big regulator, in a localised area causes losses in the copper of the PCB and causes heat. In the structure we have 12V at say 4-5 amps max is the distribution and this is a relatively low loss. Similarly the max point at the output of the smaller regulator approach only has to deal with a max of 12A so less loss and heat. This is a simplistic analysis but reasonably accurate.

Distribution voltage also effects the efficiency of most regulator circuits. 12V going down to 1.2V tens to be less efficient than 5V going down to 1.2V. However this is offset by the higher distribution loss of 5V. So it's swings and roundabouts which way you go.

Efficiency of most regulators varies with output current. At low currents they are usually poor. At high currents they lose some efficiency. That's copper conduction losses mainly. The best place to operate tends to mid capacity e.g. about 6A in out 12A circuit case. In this there is an equation where operating your FPGA at maximum might not be the best return and you might be cheaper with more FPGAs doing the task. Remember FPGAs have metal leadframes and solder balls which also have a conduction loss.

We actually thought about the possibilities of a failure and that is why each FPGA has it's own 12A core voltage switching regulator. We have actually made it possible to shut these down independently for power saving when not in use. We can use that as a last resort in a major problem event. It also a lot better situation if a fault is detected at 12A rather than a combo 48A supply. If you don't "see" a fault until you hit 48A and that might be a burning board by then.

The controller runs off the USB feed and it can be operational for debug or any other useful purpose without powering the main array. This also lets us phase power up reducing big surges at switch on.

We have thought about modular systems and I can understand how some people sell them as a good idea and we might do one eventually. Depends what level you want to be modular at. The negative side of this is that it is more expensive to manufacture say a motherboard and 4 modules that's because you but 2 types of PCB, 5 actual PCBs, and the assembly cost and test costs are close to 5X that of a single board approach. You also add the cost of connector and the electrical losses in currents passing through them, Connectors also add another thing to fail and more thimgs to buy and assemble.

Thermal design of a modular system can be poorer as well if not done well that can increase the fail rate of boards as a result.

Update on our warranty policy. It has been extended to 1 year for our Cairnsmore1. We will review that further once we know a bit more about the board.

Yohan

It probably won't be cheap and you get into an equation that it's probably cheaper just have more boards and keep it simple. This sort of stuff can make a horrible mess if it leaks or spills.

It is 181.403mm x 126.492mm plus or minus a little tolerance.

Note the dual sets of heatsink holes on the latest CAD image. We have a fairly good northbridge type heatsink ordered for the front that will user the bigger pair. Optionally you can have a smaller standard one on the back if you like. We are hoping to offer some addition options for heatsinks e.g. like copper options and maybe a water coolers. We will also be looking at some other techniques as well that might work well but that's for us to experiment with here first.

Yohan

There is a new CAD image now on http://www.enterpoint.co.uk/cairnsmore/cairnsmore1.html. This is the finalised prototype which has gone to manufacture. We are more or less on schedule and the first prototypes, assuming no problems in PCB manufacture (problems do occur regularly in PCB manufacture) will go on our assembly line next week. If that all goes exactly to plan then we will show some pictures of a real board either at the end of next week or the beginning of the week after.

Those of that have ordered you will get an update in the next 2-3 weeks when we have a better idea on when parts are going to arrive with us and also depending on how we are going to load our line for coming weeks.

It's our intention to try and give everyone that is on the order list (as is today) at least 1 unit on the early end of the schedule and gather as much feedback on performance as we can for as broad a range of customers as is possible. This will offer a level of assurance to purchasers of multiple units to see what they are getting before they have to pay for their entire orders.

Yohan

Our board should not need JTAG very often with the on-board bitstream storage. It shoul only be needed when an update is required.

It will also be possible to update the bitstream stored in a non-JTAG way if build the necessary function into the design. We can also store a few different bitstreams per FPGA which can be selected using a multi-boot technique.

A little more product detail just gone on there.

There is one set of holes per FPGA for a heatsink already. It just doesn't render as well as the plated holes. We are looking to see if we can fit some slightly wider ones on the other corners for our preferred heatsink option.

Guys

CAD image is now here http://www.enterpoint.co.uk/cairnsmore/cairnsmore1.html. I will be adding some more detail to this page later today.

Yohan

CAD image should be up this afternoon. We are a few hours behind on our target schedule but nothing serious or a show stopper. We are just doing final tidy up and checking through the design. We are aiming to order a fast prototype PCB today so that is taking priority at the moment. If it goes to plan we will start assembling this prototype on Friday or Tuesday depending on what time we complete the design today.

As soon as we are happy on the design it will be there and I will post again.

You might well be right although in our standard products, i.e. ones you see on our website, that go into "normal" applications we have only ever had 2-5 boards returned, or reported, as faulty over 9 years of making boards. That's over an awful lot of boards and a fail rate way below industry norms. Resultantly we don't have to think about warranties very much. The Bitcoin market is different in that we know people are going to try and stress these boards to get the maximum out of them and will attempt to take them beyond sensible operation. I have even seen posts on the forum where people have blown boards by abuse and are being told to try and "cover up" to get a warranty repair.Now in my book that's not fair to the manufacturers either.

So in here there is a fairness thing between us and the customer and that is something we need to try and get right and we will do our best on that front. Most of our existing traditional customers have had an excellent service from us and keep buying again and again and we think we have done a good job there. We are hoping that the Bitcoin community with end up thinking in the same way and that is a matter of time whilst you all get to know us and what we do.

If you want to be accurate the EU law I believe that it applies only to consumer purchases. I'm not a lawyer but in my humble opinion as most miners aim to earn money by these purchases they would not count as consumers and hence they won't be covered by that law. That's probably a gray area so you can all argue about that.

There is a second part of this law as far as i understand it is that after 6 months it is up to the consumer to prove that they didn't abuse a product. In most things this isn't likely to be an issue but Bitcoining where everyone is trying to overclock that's going to be a hard one.

We are looking at our design in terms of the warranty and we have some features that might allow us to do 1yr and maybe 2 yr warranties at minimal risk. At the moment the decision simply isn't made so I won't say any more on this until we do.

There will be a warranty of at least 90 days which is our current standard for industrial customers. We may increase that to 1 year for this product. Whatever the warranty it won't cover outright abuse of these boards of course.

As to colour purple is unlikely in the first boards because it will impact the number of places we can get PCB manfactured. The cost isn't a problem. Maybe later we will consider something like this but building the first boards it is more important that we keep our timescales.

The board will be open in the sense that we will tell you pinouts and any other information you need to get it going. It will all become a little clearer this week when we show the CAD image of the board.

As I didn't say it the end of the year sounds about right for Artix. We do know more but are under NDA so there is a limit on what I can say.