I think NotFuzzyWarm's in y'alls decade neighborhood. That guy's been an engineer for at least 10 years longer than I've been alive.

I'd be impressed if you smoked an ASIC on and S1 before you smoked a regulator.

There is no pot position to achieve those speeds. He's hacked the buck to get higher voltages.

Ballin'

I might have implied that 99% would unplug 'em, but not that 99% of it would stop running. How much of the 0.3J/GH gear will be bought up by folks who can still run it profitably? I expect the first drop will be pretty good, but there's a lot of places that probably wouldn't mind getting their hands on cut-rate S7 and run 'em for another couple months on 2-3c power. That 1% is probably itching for it, and I bet anyone who paid, well, what Bitmain was charging for an S7 will probably sell to whoever he can in order to minimize losses since those things were pretty much overpriced from the get-go. Hopefully a lot of that hashrate disappears, but it's no guarantee.

And hopefully a socket that's actually rated for that kind of current, and blades on the fuse that are rated for it also, not like the AM Cube. The AM Blade had socketed fuses on some of 'em, which was handy since fuses blew on those more often than there was good hardware-failure-related reason for it (and the soldered-on ones were difficult to replace). But I wouldn't have run 8-10A through one of those little sockets. My board needs to be good for at least 20A, probably 25A just in case of crazy people. Automotive fuses make perfect sense for a 12V 25A load. Hopefully a 30A-rated board-mount socket won't be too hard to find.

Yeah, there's not even implied support for these sticks on Windows 10, because to test it I'd have to have an internet-connected Windows 10 machine and that's something I have zero interest in setting up. I'm sure that alienates some of the customer base, but to be honest I don't really care.

Also, if you're getting that reset error and it's not submitting shares and the white light isn't flashing, something is pobably wrong. Check the voltages to make sure everything's peachy. If the 0.9V is reading more like 1.0V something's wrong. If VCore is reading outside the range of 0.55-0.80V something's wrong. If Reset is reading outside 1.5-1.8V something's wrong.

If something's wrong, contact me for warranty info. But if you contact me for warranty info, I want to know everything you've checked so far. I want to know if the heatsink gets hot while it's running, I want to know what the frequency and voltage settings you've used are, I want to know if you're using cgminer stock or cgminer-gekko or BFG, I want to know what the voltage measurements are, and I want to know how cgminer and the stick's LED behave when it's plugged in and when cgminer tries to initialize and when it tries to mine.

Well, mining is self-regulating. The halving is going to hurt a lot, but it's not like bitcoin's going to disappear. What becomes unprofitable to run will no longer be run, coin prices will go up, fees will go up, and the system will keep on running. If anything has a chance to be profitable, it'll be the most efficient chip available at the time. That's dependent a lot on initial cost, but the market is adjustable as well - if BitFury can't sell enough of them at whatever price, they'll either use 'em themselves or lower the price. There are way too many variables to make any sort of accurate prediction, but the fact that pretty soon 99% of the world won't be able to turn a buck with 0.3J/GH gear helps the assumed viability of 0.1J/GH gear.

The buck chip I'm looking at, and the main FETs, are all rated for much higher than 15.5VDC. I'd just have to make sure the input caps and any accessory LDOs and such are rated for it too. Shouldn't be a problem. The main buck chip also has a UVLO input I can set to any voltage I want. So that's probably all doable, and fits in line with capabilities Novak and I were wanting in there anyways. I was gonna guarantee at least 14.4V operation, so 15.5V isn't too much farther. Fused input is a decent idea also, but I'd probably put a fuse such that the maximum expected buck output at the minimum expected input voltage doesn't blow it. Could prevent a fire or other damage in case something goes wrong, and I think a lot of miner hardware could benefit from that sort of protection.

I've tasted some of the pretty weak heavily-milked-and-chocolated stuff and if I can taste any coffee in it I don't like it. Not even the smell of coffee. If I have to dilute something so much I can't taste it in order to make it taste good, why even bother putting it in there in the first place? Water tastes better, slakes my thirst more efficiently and is free (or close to it) in most places.

And no, in these still early days of the internet there are still definitely small towns because the internet isn't in use everywhere. And it definitely wasn't in use in farms outside my small town (1900 people, 20 miles from anywhere bigger) 15 years ago. And yeah, I do tend to follow that "do unto others" - leave me alone, and I won't bother you. Try to figure it out yourself first and then ask for help when you've exhausted your own resources, and I won't ask you for help until I'm darn sure I need it. Don't be selfish, and I'll treat you with respect. Stuff like that. Add to it a disrespect for purchased authority and any other sort of entitlement that is bought or inherited rather than earned by labor or industry, and an overarching enmity for people who expect and demand rewards disproportionate to effort (and especially the people whose sole occupation is drawing profits from buying and selling things which don't exist at all or exist solely on paper and are never actually possessed). Like I said, unable to relate to most people these days.

But discussion like that belongs in a different place.

No, actually that's mostly explained by a combination of small-town upbringing (focus on Christian ethics, frugality and self-reliance) and 98th-percentile intelligence, which results in a low tolerance for idiots or lazy people and a general inability to relate to most folks even within my own generation. The facts that coffee tastes like burnt dirt and caffeine has no real effect on me anyway are mostly irrelevant to my attitude in general.

Things specific for my design should be addressed in the Community Miner Discussion thread, rather than a general Bitfury thread.

Also, I don't drink coffee.

It's not the case with Bitfury chips now that they have the separate comm multiplexer (nor BE100/BE200), but most mining chips from the last couple years (everything Bitmain and Avalon at least) had chained comms so one chip acted as a relay for the next. In that case, even with a regular VRM design you could still kill a whole board with the loss of one part. You could still have partial functionality on, say, an S1 if you lost a chip (or regulator) in the middle, as the ones before it would still hash, but your efficiency would be worse since the rest of the chips are actually still drawing power. I really do like string design but the lack of single-fault-tolerance has always been annoying.

The Prisma had a separate FET on each node as a current shunt, because if one chip dropped out the rest of the board was unaffected as far as signal goes - so the FET would kick on and draw one ASIC's worth of current, keeping all the nodes balanced for power and continue operating. Good idea in theory, but I would guess that's what caused some of their fires.

Good ol' BE300, December 2014. Those were the chips that made me and Novak decide we should design our own miner. Too bad ASICMiner imploded right around then, because their chips tested out better numbers than pretty much anything until the BM1385 was released. Would have trashed the S5 and SP20.

I assumed he meant that many chips wide, where the current does multiply. The wider a string, the more stable it's going to be as any changes in current demand (which then affect voltage) in one chip will be buffered out by the remaining chips in the node, so you really don't want to run a single-wide string if you don't have to.

I probably have enough for sticks. But if you want to trade it for one of the S5-controller-compatible pods that'd be alright because I won't have enough chips for everything.

Stock setting on an S1 is 350MHz/180GH but a typical rig will push to 400MHz/205GH pretty well. The extra 13% hashrate more than makes up for the extra 1-2% errors. So your bosses are okay with adding $50 a month to the electric bill so you can mine $5?

Depending on how near, look at kilo's "Community miner discussion" thread. That's basically taking over my TypeZero project and the plan is to build a USB-connected board that'll fit on an S1 chassis with BitFury chips.

This question has come up repeatedly in the past, but it's been a while so I will go ahead and answer it.

No. In no uncertain terms, no. No gridseed pods, no U3, no R-Box. U3 is perhaps the most likely candidate, as it's probably the least uselessly small and dense. But in general, no.

Well don't tell anyone I told you this, but I did today send out for prototype PCBs for a version that'll mount a BM1385, and approximately doubled the current rating of the buck's inductor. And my next round of protos will probably feature an Avalon A3218 version with software voltage control and temp sensing. The board's already drawn, I just want to work on the firmware on a testbench a bit before committing to the layout because I might end up changing some pins around on the micro. Unfortunately those would both be pulled-chip designs (unless Avalon starts selling) but maybe it'll be upgraded to Bitfury sometime if I get access to those (which is in the works).