I didn't know either, but apparently the have a switch on 'em to go scrypt-only or both and these have the switch replaced by jumpers. It would not be hard to shift the jumpers to the other position.

Raskul, I'd said if I got to the first of the year and nobody'd turned up some old Garden blades I'd let you know. Here it is, and I still have none, so... whatcha selling?

Eneloop, I do actually have some dualminers but they're the scrypt-only version, if that matters. I've got six but don't need that many if you're interested in picking up one or two.

Hey raskul, you still got those old Garden blades?

I'm a bit interested in the stuff that's not S1.

Nope. They like 240V, but they'll run on anything 100V and up.

I'm here not necessarily because I trust anyone to do anything ever, but because I do like the idea of building an ASIC miner and have looked into it a bit. So if that's the present discussion, I'd like to participate.

BlackArrow was several months behind on shipping their miners. Basic things supposed to have shipped in February were unavailable in July. They offered to their customers a trade which amounted to about six Minion ASICs for every ASIC in a piece of hardware (so the X-1, a 1-chip miner, would become six loose chips) and they could do whatever they wanted with the chips. Technobit, capitalizing on that, announced a four-ASIC board built for the Minion ASIC and an offer to take in chips in trade for assembly. A four-chip board would cost eight chips - four to install on the board, and four taken as payment. They were supposed to be shipping out these boards in August. Minersource worked out a big "group buy" which should have been delivered just a hair over four months ago and, as far as I know, very few people have gotten theirs yet. Boards were trickling out at a few a day for a while, and riddled with problems. The thread hasn't been updated with progress in a couple months and I think a lot of people have just given up. So it's possible that Technobit has a crate full of Minions hanging out in the back room not doing anything - that's not saying they should, since some of those probably belong on customer boards, but some of them were taken in as payment for things which might have been delivered. If they haven't used those chips yet, they might be willing to sell them. Or stick them on boards for you?

In any case, the BlackArrow Minion is probably not a good chip to build around. If it had met its target efficiency (about 0.55W/GH) instead of blowing it pretty badly, it'd be very competetive for a while. If I'm thinking right, the $500K/25k chips comes in about $200/TH which is a good price but ASICs are only going to be half the cost of a miner - even less if PSU are considered. And BlackArrow has been really good about blowing deadlines and ripping people off, so who knows if you'd ever even get the chips.

If you want Blackarrow chips, I bet Technobit has a bunch they'd be willing to unload since they haven't put them on boards for customers due in August yet.

Novak and I are toying with the idea of a configurable USB-connected single-board miner for BE300 chips, something that could run off a DC brick (or better PSU, it'd have a barrel and screw terminals) with full under/overclocking options from the command line. Basically the Jalapeno market sector, except with tomorrow's power efficiency. Drop me a line if you think there's merit.

You need to weigh the expected savings of producing miners vs purchasing exising miners in bulk from a large manufacturer, against the opportunity cost of mined coins during the time between which you'd have hardware in hand if you bought existing versus the expected start of mining from in-house-production hardware. If you can use less-efficient gear and have somewhere to put it, talk to Bitmain about some of the petahashes of S3s they're unloading on the cheap. Buy a truckload or three, run them until the daily operating cost equals the daily average yield, then start working them down with undervolts and underclocks.

I don't really know. Who's selling chips currently? Anything built now is going to have to compete with a generation of 0.2-0.3W/GH stuff in a few months, which is going to hurt longevity. Also as a rule I ignore altcoins so can't (or won't, either way) give advice there.

If you can get a barrel of BE200 for cheap and underclock the crap out of 'em might be okay. Avalon's new chips are available in quantity and have decent (for extant hardware) power efficiency, not sure about pricing though. I'm looking forward to BE300 availability but information is sparse and the chips themselves are two months away.

I've got several NF6 in the museum, not sure if they're Rev2 chips or not (USB miners is more novak's project). Swimmer, you built miners? Nice. We should talk sometime.

Or build a board around someone else's chips, which can do pretty well if you do it right (Technobit, Peppermining etc).

Depending how many you want to string, it might not be too bad. But stringing works best if you can divide the rail voltage, which for 5V means more chips than he's talking, so you'd need to build a VRM anyway. At that point it's much easier to just kick the voltage down to single-chip level and run several chips in parallel, standard topology.

Pretty sure the NF6 is strung chips, but it's also (as noted above) a very high current requirement. Not caring about standard USB port current limits changes the ballgame.



Also, he beat me to the punch but MrTeal is pretty much always right.

PCB and heatsink aren't much more difficult if you put a bit of work into it. Heatsink on top of chips is not really any harder to accomplish than heatsink on top of board. Just stick your ASICs on the backside, control hardware on the top side; grease your chips, sink the backside and screw to it through the board.

Looks like shipping would be $25, and $42 on the board and cables.

International shipping prices depend on whether you want just the board and cables, or a full kit including a PSU. Full kits ship up to three in one flat-rate box for $65; non-kit orders will have to be bid individually.

Full kits are $55; interface boards are $28 apiece ($26 for more than 5) and our 18" cables are still $3.50 apiece while 36" cables are $4.50

Yes, yes we are.

That's certainly how I'd do it. It's not hard to find some convincing-looking PCBs to strap big CPU coolers on, drop them in a somewhat generic 3U/4U rack case (but make sure airflow directions are logical) with a hefty PSU and cable it to an internal Pi for connectivity and control. Make sure you understand the mathematics behind cgminer stats before faking your screenshots and video so ckolivas doesn't call you out for having inconsistent data, point the appropriate hashrate at a posted Eligius address to look like a real test unit is flying solid and let it run for a few weeks. If possible, post a picture of a pallet of your (probably empty) generic 3U/4U cases, along with 3 or 4 on a slick bench somewhere all cabled up with lights flashing and the one "display model" open with a few innards visible. Make sure there's some miscellaneous lab equipment on the bench (solder station, scope, decent power supply or two and a scattering of random ICs and resistors strewn in the back) so it looks more legit.

Why do these guys want to do everything with photoshop? So lazy.

Two weeks and a thousand bucks in materials, you'll make it back inside of day one.

True, nominal efficiency ratings are important. A topped-out 1200W server PSU is probably more efficient than a bronze-rated PSU at its best possible operating point.