cgminer does display each device's hash rate. What are you seeing?

I've had this build running on my sketchiest tester with the sketchiest sticks (it's been a billboard example of the kind of instability and low hashrate folks are complaining about), without AB enabled, for the last 8 hours and I'm seeing 99% target hashrate and no zombies. Try it out.

Right now (literally right now) I'm working on getting a single chip talking. Once that's done, it won't be difficult to make two, ten or a hundred work together. The R808 failed mostly because of manufacturing issues but it also wasn't that good because of chip efficiency.

If I build anything larger than an improved Terminus, it'll be new boards for an S1/3/5 chassis. With improved heatsinking (like some S3 had) and adjustable voltage and clocking, they should be able to run pretty quietly.

Talking about jstefanop's Apollo? I've been meaning to follow up with that guy for a while. He does good work.

Stickminers are all I got right now, and I'm just about the only one in the world still building them. In the last year or so I've seen two other outfits come out with stickminers, but they were both Chinese clones of my designs.

My Terminus R808 from the first part of this year was the first sub-100W miner in a long time but it wasn't very good so was discontinued. I'm currently working with Bitmain's BM1387 on the stickminer but that chip is EOL so something new must be had for a close-to-viable updated design. But I'll let you know, it's already in the works.

I run some old Dell Optiplex 755 with Debian 6 and VH's latest build. I've got three Plugable 7-port USB2 hubs per machine each running 4 sticks for 12 sticks per tester.

The latest few hundred sticks are not as stable as the first few hundred, even on my testers. The only change made was the chip supplier. I will not be buying from them again, just in case.

Technically every miner will stop hashing when temperature becomes an issue. Whether that's a destructive process depends on the miner. These have no temperature-measuring mechanism, or it'd be visible in cgminer.

Give yourself some credit. It's been about twelve days since the price was at best barely tapping "within 5% of 4500" margins.

With how Bitmain distributes work to its chips, timing is very particular to keep the buffers filled and it only gets worse with higher hashrates and more devices. Different chips have different internal work buffering mechanisms, and the particulars of ASICBoost also change that dynamic, so non-AB Bitmain with multiple chips is pretty much a worst-case scenario for timing requirements. Barring unforseen disaster, I likely won't be using their chips again in the future. Leastways not anytime soon.

If you can get ASICBoost working on your NewPac setup, the timing is relaxed by about 3/4 on them. That means if you can handle two R808 on a Pi, you should be able to handle about three times the hashrate worth of NewPacs. Should be able, I haven't fully tested that exact setup yet.

(Thereby lending support to Google, the largest advertising agency and data profiteer on the planet, further reducing the credibility of your worries)

Maybe Bitmain can set up a payout address that gets periodically cashed out and donated to UNICEF or something pretty much everyone likes. Put those pennies to work.

Okay that's pretty balls. I built the darn thing and I've never got one to run 300MHz. Nice.

By the way - the barrel jack on the R808 is only rated for 5A, so a peak of 60W reliably. I'm looking at an 8A-rated part for a different project and if I like it we'll probably use that exclusively on anything needing a barrel in the future. Might be more sturdy.

I'm gonna guess it's the same problem everyone has with this step, and that you're getting a gcc error instead of script completion on the autogen because you're typing dash zero two instead of dash capital oh two on CFLAGS.

Looks like it hasn't been mentioned in this thread yet, but I lost count of how many times in the 2Pac support.

If that's not the case, you'll need to let us know what the error was that caused autogen to fail.

It's pretty much the same places.

If you see the two bare pads below the bottom yellow thing, near the geometric center of the stick, the left one (to ground, which can be pulled from the USB jack) measures the total voltage (stock is 830mV) and the right one measures the midpoint voltage between the two chips (ideally about 415mV).

The best part about the tariffs is, we're now punishing Americans for letting their parents let China take over global electronics manufacturing. The retaliatory tariffs China put on us are for stuff they can buy from 20 other places but we're taxing ourselves for spending 30 years telling them they should do the work instead of us. You know MAGA hats and Trump 2020 signs are all made in China.

Of course the BTC economy is being controlled by China now. We gave it to them. We paid them hand over fist to take it.

Has it ever been on a downward trend this long though? It's been pretty much dropping steadily the entire year. The last major dumps I can think of were the bottom falling out of media hype bubbles, and the MtGox fiasco which revealed a whole lot of market manipulation and outright lies.

MFB, I gave y'all some money for the chips sale y'all brokered but that's the only "account" I might have. As for financial anything else, I abhor debt so I won't borrow except in emergencies and have never endorsed trading for profit. Call me an optimist but the bulk of my savings for the last 15 months or so have been in BTC. Needless to say I've been taking a beating lately.

But hey, if diff keeps dropping maybe I'll get some good heater-mining in. The apartment and shop are all electric heat.

Phil - not really, no. I've got an S9 I cobbled together from salvage parts and something like a dozen S7 I'll be lighting up for heat soon but that's about it.

MOTHERF$@&

Would be a freakin' marvellous time to buy if I didn't need to *increase* liquid capital right now.

I haven't put AB releases on the test stations yet, only using it on my own bench. It worked perfectly for me, pushing six sticks to 145GH each. I've been meaning to update the test stations so maybe that'll reveal an issue VH can track down?

Admittedly the lightweight fan was barely enough cooling for those six sticks; it only worked as well as it did because my shop's ambient temperature is like 60 degrees right now. Really should have had more for long-term safety.

That's a good rule of thumb. When we had a temp sensor on some early test sticks, VH noted issues starting when the [cooler than the chips and heatsink] PCB reached about 120F.

With these sticks, a little bit of cooling goes a long long way. Note that the uncooled 100MHz power draw and the slightly-cooled 200MHz power draw are pretty much the same. My six 650MHz sticks (drawing 14W each) were collectively cooled by a single 12V/200mA fan.

Also note that these guys can be sensitive to extended overheating. Bake the chips too much and they may never work right again.