GekkoScience BM1384 Project Development Discussion

[AJRGale]

I must ask Side, how's the talks with Guy going/went?

[1713503667]

1713503667

[1713503667]

1713503667

[1713503667]

1713503667

[1713503667]

1713503667

[notlist3d]

I must ask Side, how's the talks with Guy going/went?

I have a feeling he would love to tell us all a bunch on this subject.  But I suspect there is a big NDA stopping him.   

And if you have read any of these NDA's from asic companies it stops A LOT of information from going out.  So I understand if we don't hear about this for a bit.    But I would be pumped if he does say he got a chip source.

[AJRGale]

I must ask Side, how's the talks with Guy going/went?

I have a feeling he would love to tell us all a bunch on this subject.  But I suspect there is a big NDA stopping him.   

And if you have read any of these NDA's from asic companies it stops A LOT of information from going out.  So I understand if we don't hear about this for a bit.    But I would be pumped if he does say he got a chip source.

he could go "not sayn' a thing" for things are going swimmingly, or "nup, not going there" if things fallen out.

you know, the same way everyone else uses the "I cannot comment at this time" when they meant to say yes.

but hey, could take the silents as they are still talking

[sidehack]

One board would provide one pod. Since one board gives chips for about three pods, and chips would be one third the parts cost/value of a pod, one board provides one pod. So if you give one board, you get one free pod. If you want three additional pods you have to pay for three additional pods or send in three additional boards.

The cost of the pod is not just parts, and not all of those parts are free. The cost of the pod also includes the time I took to design it and prototype it and manufacture it and the cost of keeping the lights on in the shop I used for that amount of time. The price will be low, because I don't like marking things up a lot, but let's be honest I have to eat sometime and food costs money.

I actually haven't heard anything from Guy in a bit. He gave me some initial info, nothing technical, and then went on vacation for a week. I know some stuff that I can't talk about neener neener but I don't know enough yet to actually do anything about it. When I do know enough to actually do anything about it I might not tell you anyways because NDA, but right now I'm not really there yet.

[notlist3d]

One board would provide one pod. Since one board gives chips for about three pods, and chips would be one third the parts cost/value of a pod, one board provides one pod. So if you give one board, you get one free pod. If you want three additional pods you have to pay for three additional pods or send in three additional boards.

The cost of the pod is not just parts, and not all of those parts are free. The cost of the pod also includes the time I took to design it and prototype it and manufacture it and the cost of keeping the lights on in the shop I used for that amount of time. The price will be low, because I don't like marking things up a lot, but let's be honest I have to eat sometime and food costs money.

I actually haven't heard anything from Guy in a bit. He gave me some initial info, nothing technical, and then went on vacation for a week. I know some stuff that I can't talk about neener neener but I don't know enough yet to actually do anything about it. When I do know enough to actually do anything about it I might not tell you anyways because NDA, but right now I'm not really there yet.

Yea the NDA's are not the funnest thing.   Basically make you watch your steps twice not to accidentally say to much.  I think it is a great thing though that SP is talking to us.  I met Guy once at a bitcoin conference but I'm sure he does not remember me lots were there.  He and his team seemed pretty nice.  I also have a sweet SP shirt somewhere from them.

And also ontop of the parts on the pod there is time from both sidehack and novak.   So really need to provide something above parts for them .  We want out developers to  be able to live.

I really hope the SP thing works out though.  Would be amazing if they picked up supporting hobby miners and supporting your projects.   

[sidehack]

Just to have some more options to play with, I won't give up Bitmain or InnoSilicon about chips either.

I should probably go to a bitcoin conference one of these days. Not like I haven't been immersing myself in the community and economy for over two years now or anything, just never had the budget to take time off and go to wherever those things happen. If I make any kind of decent miner I should prioritize it though.

[VirosaGITS]

Just to have some more options to play with, I won't give up Bitmain or InnoSilicon about chips either.

I should probably go to a bitcoin conference one of these days. Not like I haven't been immersing myself in the community and economy for over two years now or anything, just never had the budget to take time off and go to wherever those things happen. If I make any kind of decent miner I should prioritize it though.

Maybe you could find some bigger players that would be up to help you start up a home-miner oriented factory for pod and 20-40 chips miner board? You already started making good stuff, maybe that will attract some nice attention after you release the pods and the S1 upgrade board.. Zero Type was it?

[notlist3d]

Just to have some more options to play with, I won't give up Bitmain or InnoSilicon about chips either.

I should probably go to a bitcoin conference one of these days. Not like I haven't been immersing myself in the community and economy for over two years now or anything, just never had the budget to take time off and go to wherever those things happen. If I make any kind of decent miner I should prioritize it though.

It was a great time.  Only thing that kinda stinks is the number for miners has went down considerably.  But meeting so many of the miners was great back in the day.  Was during 2014 in Chicago one I went to, was like North American Bitcoin Conference.  There were lots of miner companies and a few resellers along with other booths.

But now it seems a lot of the conferences are focusing on the investment side instead of mining.  Which might not be a bad thing if you can bring your projects infront of some of those people might lead to an angel investor type thing.  

I got a few funny things out of the conference.  I got to meet minersource in person and ask them if they would take my pictures they stole of SF-28's off or start sending me money.  They gave me a card and ofcourse did not remove it after.

Also met BFL i almost felt bad for the employee stuck at that booth.  They had a prototype monarch at that point.  I got a little bit of fun asking about the status of it.   It was pretty obvious even then they were not on up and up.  

Bitmain, and SP both were much more better experiences and I was impressed with both in person.  

The best list on upcoming events is really on coindesk: http://www.coindesk.com/bitcoin-events/ .

Enough of me going off topic .   I look forward to the POD's no matter what you end up using in them.  I already know I will want one.

[sidehack]

It'll be nice when I have the S1-sized TypeZero tested and working. The pod dev will actually get us a ways closer; it won't be the final chip or the final power but it'll be awful close to the final controls and that's not unsubstantial.

Partnering with bigger players is going to depend on the terms. If it's an investment where they want partial ownership of my enterprise, that I won't allow. A loan to be repaid from sales revenue or as miners themselves could be arranged, but neither Novak nor I like the idea of letting someone pay for the right to make decisions about what we're going to do. If I needed to expand our manufacturing capability what I'd like to do is just buy more equipment and set it up in the facility I already have. Might require hiring on to manage it if production really took off, but being able to hire help is one of the reasons I won't outsource production to another country for any reason - I hire American or not at all. Folks here need jobs, so if I have jobs folks here will get them or they won't be gotten, and no big player's money is going to convince me to do otherwise. Folks who know me know I'm motivated by deeper interests and ethics than money.

[VirosaGITS]

It'll be nice when I have the S1-sized TypeZero tested and working. The pod dev will actually get us a ways closer; it won't be the final chip or the final power but it'll be awful close to the final controls and that's not unsubstantial.

Partnering with bigger players is going to depend on the terms. If it's an investment where they want partial ownership of my enterprise, that I won't allow. A loan to be repaid from sales revenue or as miners themselves could be arranged, but neither Novak nor I like the idea of letting someone pay for the right to make decisions about what we're going to do. If I needed to expand our manufacturing capability what I'd like to do is just buy more equipment and set it up in the facility I already have. Might require hiring on to manage it if production really took off, but being able to hire help is one of the reasons I won't outsource production to another country for any reason - I hire American or not at all. Folks here need jobs, so if I have jobs folks here will get them or they won't be gotten, and no big player's money is going to convince me to do otherwise. Folks who know me know I'm motivated by deeper interests and ethics than money.

That's reassuring of you to say. I think its important you hold the reign too, lest you end up like the other big BTC ASIC manufacturers out there.

Hopefully people will come to see what you can do, like me, but will have significant money to front. I see myself willing to put what i can toward your project, so if i had more money i would see myself supporting and making your future project better/bigger.

TLDR; If i had money i'd fly over there and see what we can do, but i don't, so hopefully someone who does can do that.

[AJRGale]

...

I know some stuff that I can't talk about neener neener but I don't know enough yet to actually do anything about it.
...

Oh he's good at that, about a year ago I have gotten info out of him and told me to Shush about it. I have respect for him because of that, and what he did for people back then, I haven't followed much of what happen this year though.. none the less, I'm not going to throw him under the bus, I still see him as a good guy Guy.

Just to have some more options to play with, I won't give up Bitmain or InnoSilicon about chips either.

I should probably go to a bitcoin conference one of these days. Not like I haven't been immersing myself in the community and economy for over two years now or anything, just never had the budget to take time off and go to wherever those things happen. If I make any kind of decent miner I should prioritize it though.

Any luck with them 2? PlanetCrypto is/was looking into InnoSilicon yes?

[sidehack]

The way I see it, Spondoolies engineering annoys me to no end (those guys seriously do everything exactly opposite of how Novak and I would do it) but as a business they have my respect. No other manufacturer has their quality of customer support or communication. I might not want to buy any of their miners for engineering reasons, but I'd work with them as a business because, well for one thing they know how to speak English and for another they know how to not screw people.

I haven't heard anything out of either IS or PC lately. The last communication between them that I was privy to, IS was not interested in sample chips but they were interested in six-figure preorders for the new chip they wouldn't sell samples of.

[notlist3d]

The way I see it, Spondoolies engineering annoys me to no end (those guys seriously do everything exactly opposite of how Novak and I would do it) but as a business they have my respect. No other manufacturer has their quality of customer support or communication. I might not want to buy any of their miners for engineering reasons, but I'd work with them as a business because, well for one thing they know how to speak English and for another they know how to not screw people.

I haven't heard anything out of either IS or PC lately. The last communication between them that I was privy to, IS was not interested in sample chips but they were interested in six-figure preorders for the new chip they wouldn't sell samples of.

Six-figure would get my attention aswell.  I'm surprised they would not do samples though.  Before I would ever drop that kinda money I would want a rock solid design.

That amount of money is where it really would to take a angel investor.  As I'm not sure a bank would do a business loan in most cases unless a lot of collateral on a high tech deal they likely don't understand.  I hope they lower that number after they have introduced it to market.  As 6 figures is a lot to commit.

[sidehack]

I wouldn't advise anyone to pay down on a batch of chips without verifying samples first, no matter who makes 'em. If nothing else it also gets you a head start on dev.

[VirosaGITS]

I wouldn't advise anyone to pay down on a batch of chips without verifying samples first, no matter who makes 'em. If nothing else it also gets you a head start on dev.

No but if i had 6 figures to invest, i could wave it in their face and ask them for a sample mini-batch. Considering that its quite a significant amount of money, that would carry more weight than a small time ASIC Fabricator asking for a sample batch to be upped conditionally.

[sidehack]

I think PlanetCrypto would be willing to put up that kind of money for a chip, but not on a preorder for an unknown batch with a full foundry lead time. Samples would probably help convince him, but I doubt they'll see any money from him at all until chips are in hand to ship out. Even then I wouldn't advise paying on a big order without samples. Having samples earlier would be super nice, like I said, because it gets you a head start on dev - by the time they had chips in hand from the factory we might already have a test board and know we can work with 'em, and then they're worth buying. Getting samples from the factory batch and then coming back a few weeks later for a bulk buy puts you that much farther behind ideal schedule.

[AJRGale]

I wouldn't advise anyone to pay down on a batch of chips without verifying samples first, no matter who makes 'em. If nothing else it also gets you a head start on dev.

Its mean main business with any other IC manufacturer, want 10 chips of the latest Arm CPUs? call them, they'll send you a sample! (i could be lying it could be all distros doing)
Hell, even going though distributors, they happily send you samples.

But hey, tell that to these head honchos ..

[sidehack]

So, I'll admit I cheated just a little bit with the 8-chip proof of concept for the pod miner. I actually hacked off a chunk of an S5 board and used it for the hashboard. Of course that didn't test anything but the power concept since it was Bitmain's clocking and level shifters and node caps and everything, but it got the idea up to speed quickly. So I spent the last couple days building and testing a 4-chip using test board stuff I already had. It's running off an S5 controller for now but that's only because we don't have multichip code. I'll be out tomorrow but Novak's looking at grabbing the board to dev some multi-chip code of our own and hopefully it'll init more smoothly than what's on the S5 because the S5 controller is good but it's not perfect.

Right now it's operating off about 16.5W to get 44GH. That's not great, but it's also not terrible. Once I have 8 chips the core voltage per chip will drop and 44GH will happen at a fair bit lower than 16.5W, that's the goal anyway. This power doesn't take the controller into account, but the controller doesn't need to be considered on a per-machine level because it's not part of the machine. When I had it set to 100MHz I was seeing 22GH off a little shy of 7W board-level, which I thought was pretty good. Anyways this tests my level shifters and node-level power and a few other things. It's also pretty squirrelly with a spaghetti of wires all over the place which is probably making the efficiency even worse.

[notlist3d]

Thanks for sharing!  I love seeing the dev stage.  Thanks for sharing it.

This is one thing I love about your dev instead of other companies that keep it all secret and don't share the makings of.   16.5W to get 44GH for a start is not bad at all.  Compare to U3 that gets "1Watt/GH/s".  You are already kicking it's butt.

[sidehack]

I had some issues with the second node's reset that took me until after lunch to figure out, but when it lit up I immediately did several pelvic thrusts and then ran into the office to excitedly request multiple high fives. Why wouldn't I want to share stuff that exciting with you guys?

I'd consider pushing this thing to 300MHz/66GH and see how much it beat the U3 by at its own hashrate, but the chips aren't heatsinked well enough. We'll have to wait until I have a better setup.