I'll probably talk to the same people that I have to test Compacs when the Amita and TypeZero are available for testing. There shouldn't be any "private deals".

Things aren't really getting close, if you follow the discussion. We're close on a one-chip stick miner. We're close to prototyping two chips talking to each other without issues. On the previous page of the thread, addressing a pretty similar question to this one, I mention I don't expect to even have TypeZero boards prototyped until the end of May, let alone have a plan for manufacture. And depending on what happens with folks releasing new chips, y'all might decide it's not even worth buying by the time you'd get 'em, or we might not be able to buy chips, or any number of other things could get in the way.

So, I'm not going to guess.

Right, but that's you. How many people for whom the "home miner" label is appropriate know anything past 120V 15A wall sockets? This thread has already debated what "home miner" basically means, and a 9KW box is pretty universally not included.

Cool, so then the people that don't have 'em can help you pay for yours.

Except 12U 16KW machines.

Gentlemen, this discussion has been repeated verbatim how many times in how many threads in the past year? Please stay on task just a bit.

I'd be interested to see what Avalon comes up with as far as new chips. If the efficiency and price don't suck, we'll have to look pretty seriously into working with them. SFARDS really seem proactive (so far) on assisting independent development, which is great, except high-current BGA and dual-algo are both things I wish to avoid. More so the package, really.

Yeah, small miners below several hundred GH/s would be useful mostly as learning tools to get noobs in and understanding what the whole system is about, and how much fun it is to tinker with stuff. Not a lot of intrinsic profitability, but nonintrinsics abound when you consider how much knowledge can be gleaned from a sub-$100 hardware purchase when the hardware is designed to be flexible.

What DC output voltage are we talking about? It might be interesting to hack power supplies to bypass active PFC and feed high-voltage DC directly into the main switchnodes. That'd give a decent conversion efficiency boost over using AC, especially 120V into KW-grade PSUs.

I forgot the start time and my block erupter was at the shop. If I'd thought about it I'd have set it going before leaving work last night (8:45 local, 6:45PST) and been okay. Maybe I should mess with y'all and point a few TH of old gear at it for the weekend.

Oh well, no reason to turn it off now. It's on the same machine I've had the GPU running 30Mkey worth of humorous vanitygen addresses (1buttcheek, etc) for about the last five months just because. Woo, add another 2.5W to it.

As for modding, I'm not sure I'll have the time to do what I'd really want to for one. Novak's walked on his sweet idea, we got too much other stuff going on at the shop these days. I guess I could hijack one as the USB/UART for my Compac prototype and add that to the competition but I think I'd rather do nothing than do something halfass.

What about waiting 8 hours?

Also, out of interest, who's using what versions of cgminer? I'm on 3.1.1 with the -S flag in play.

Count me in on this. I just forgot to say something sooner.

Course no electrical test. It's a cheap quick-turn one-off. We ordered a few more than we needed just in case.

All the pictures I've taken have been posted in this thread.

The BE300 was what made us think seriously about developing a miner. We were waiting for news of it, and giving up on that is why we moved to the BM1384. Bitmain treated us favorably so we started designing, but the concept was initially built around the BE300.

I have what I hope is the completed Compac PCB, but I won't have boards in hand until after the 4th - more like the 8th I'd guess. We submitted the gerbs on Friday but there was a software bug causing an issue in one layer that we didn't get worked out with the fab house until Tuesday, which burned off basically the whole expected fab time. I'm glad they caught the problem instead of just making it as it was, because stuff would have caught on fire and I'd have had no idea why. I think four days lost is worth the more than four days I'd have spent finding the issue and the much more than four days spent getting new boards made. What "updated media" are you looking for?

The business I run is currently approximately built up around goods and services to support the bitcoin economy. The general skill set required relies on a knowledge of bitcoin stuff directly (as far as advice and consulting goes) but also includes a lot of general hardware knowledge, in electrical and electronic and programming and mechanical design and repair. It's very possible to have a livelihood built around the bitcoin community that is never boring and also not built around single-purpose skills. It's sorta dumb to do otherwise for exactly the reasons you mentioned.

Yeah that's what I'm hoping too. SFARDS looks like they're open to cooperation, but I don't care about scrypt so I don't really feel like paying dual-algo prices for chips and I also don't like high-power BGA. If someone could get BE300 into the marketplace I'd be all over that.

Hammer, not Rockerbox. It's 6.8 SP10s worth if I'm thinking right, but yeah that is a lot of chips.

To address a potential point of confusion, you're currently discussing two different products. What we're currently working on is what became stage 1, turning a single-chip test into a single-chip USB stick miner. That's what I have prototypes on the way for and what I'll have a better price estimate (though I've said we're shooting for $20 about five times and people keep asking) for soon.

The thing mounting to an S1 chassis is stage 3. That I won't even have prototyped until probably around the end of May.

Yes, we're looking at writing a cgminer driver for the TypeZero boards (the not-stick-miners) which will include command-line flags for clock speed and core voltage. That goal has not changed, and very likely will not. The goal stated in the first post is still exactly the goal and all of its details are still relevant. Barring technical impossibilities, the design goal will not change.

I like to think that our lights are the sexiest stick miner lights, but then again I'm biased. Novak also suggested building a fan onto the Amita but I'm not sure if that's... cost-effective? Yeah let's go with that.

Sucks to say, but after about eleven years of sleeping next to a desktop, and after a year and a half sleeping next to miners, I now have trouble bedding down without the presence of excessive case fan noise. It's getting too warm to keep the Dragon going much longer and I'm not sure what to replace it with.

As long as I have resources available to keep making for the small miners, I'm gonna. When solving blocks is the sole domain of rich guys I might just walk away from bitcoin entirely, but not until it becomes impossible. I've already had someone with money asking me (a month or so ago) about the feasibility of making their own chips and building miners for their own farm; said someone is interested in our miner designs and might end up contributing to the overall project. If it turns into a relationship able to sponsor chip fab (unlikely but not impossible), I could redirect our take from that work to use some of those chips for affordable and non-crappy miners to sell to whoever I want. That'd be alright.

Yes, it's not a problem to discuss the future of hardware in the hardware section. But it does seem off-topic to talk about stuff not pertaining to miner sellers trustworthiness in this particular thread.