Where can I find a datasheet/specs (prelim or otherwise) on the A3?
I guess you'd have to ask them - beyond their claim of tape-out and "over 60% power saving over [A1]", I don't recall seeing any solid public info from them. Presuming it is the chip LK Group would be using, they lay claim to a future 5Th/s, 1kW miner. BW.com claims 0.26W/Gh. Either would fall under preliminary at best. 
Topic: GekkoScience BM1384 Project Development Discussion - page 35. (Read 146665 times)
Probably having a senior senility (brain fart) moment, but, what hash chip(s) does Innosilicon make that are appropriate to Bitcoin hashing?
They made the A1 Booster, aka Bitmine's CoinCraft A1 - and the A3 is likely to be what the LK Group will be using as their 14nm solution.(you can read about the drama between those two+ companies, if you're bored; later posts turns into a mess about mostly unrelated things)
Where can I find a datasheet/specs (prelim or otherwise) on the A3?
Probably having a senior senility (brain fart) moment, but, what hash chip(s) does Innosilicon make that are appropriate to Bitcoin hashing?
They made the A1 Booster, aka Bitmine's CoinCraft A1 - and the A3 is likely to be what the LK Group will be using as their 14nm solution.(you can read about the drama between those two+ companies, if you're bored; later posts turns into a mess about mostly unrelated things)
Innosilicon doesn't seem inclined to go into the farming business.
There is still hope!
There is still hope!
Probably having a senior senility (brain fart) moment, but, what hash chip(s) does Innosilicon make that are appropriate to Bitcoin hashing?
Innosilicon doesn't seem inclined to go into the farming business.
There is still hope!
There is still hope!
I wouldn't expect to be selling boards based around the coop chip until next year for sure, it'd be Bitmain or Avalon or something.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
I for one am Tired of the lack of gear available from them.
I'm predicting that'll get much worse, to the point where no component parts (i.e. SHA-256 hash chips), boards, and complete miners are available (to the general public).
The inherent design of the Bitcoin system (reward decrease in July of 2016) and the enlargement of the block size to enormous (20Mb) will force everyone not at the extremes (extremely small and extremely large) out of BTC mining. The big four are the large extreme and the hobbyist is the small extreme. Even the large multi-megawatt farms are in jeopardy because they, to a large extent, rely on the big 4 for their gear.
The big 4 (i.e. the chip makers) are gearing up for, promoting, and shifting their respective business models for this eventuality. Some of the big 4 are more transparent about this than others, Bitfury most notably:
"BitFury Group is the leading, fully integrated Bitcoin Blockchain infrastructure provider and transaction processing company."
"We build ASICs, PCB boards, servers and data centers – everything in-house and everything full custom which allows us to be the most efficient Blockchain transaction processor."
"We run one of the largest operations in the world with specially designed data centers located in Iceland and the Republic of Georgia."
To this end I'll quote lyrics from the X Ambassador's song Renegades (of Jeep fame) which embodies my belief of a possible solution:
"Long live the poineers
Rebels and Mutineers
Go forth and have no fear
Come close and lend an ear"
"All hail the underdogs
All hail the new kids
All hail the outlaws
Spielbergs n' Kubricks
It's our time to make a move
It's our time to make amends
It's our time to break the rules
Let's begin ..."
Isn't this sort of how the Avalon and Prism projects got started?
8-)
8-)
Or when the only thing available is gear that's been run for months and then sold as "new" for more than it's worth.
I wouldn't expect to be selling boards based around the coop chip until next year for sure, it'd be Bitmain or Avalon or something.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
I for one am Tired of the lack of gear available from them.
I wouldn't expect to be selling boards based around the coop chip until next year for sure, it'd be Bitmain or Avalon or something.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
I also have no problem with the coop not doing boards, but GS certainly intends to. It'd be pretty cool to see more designs pop up like when ASICMiner opened up BE200 specs.
Hopefully we get some good heads on board for silicon design. All I can really help with is what I'd want to see for packaging; at this point I don't really have well-reasoned opinions on protocol and no real knowledge of actual architecture. But I really like the idea of a community-centered initiative that'll put pressure on the Big 4 to start behaving a little better.
If whatever cooperative project PlanetCrypto is heading comes to manufacture and GekkoScience gets the contract, we'll be the manufacturer. I'd certainly be up for it. I'm sure Novak and I will probably both be in the coop helping out with hardware and software dev anyway. The chip dev coop will be a separate business entity from both PC and GS, but members of both businesses (and to some extent the businesses themselves) would also be members or participants in the chip dev coop.
So I'm wearing my CEO of PlanetCrypto hat, PlanetCrypto's preference for all things board level is and has been GekkoScience.
If the share holders of the chip dev coop decide to get into boards (which may not be the case and I would counsel against), GekkoScience has my votes as a contractual partner for design and production.
There is a valid case to be made that the chip dev coop should concentrate on being a chip manufacturer and not branch off into areas where it may not have core competencies (customer fulfillment, call center, warranty processing, etc . . . . . ).
It does not make business sense for the chip dev coop to become a board competitor, as that field already has a plethora of players. This is the kind of crap the big 4 would do. Compete with customers who are component customers.
While there is a nice synergistic set intersection of GS, the chip dev coop, and PC regarding Bitcoin mining hardware it would not make sense to merge the two/three entities as, speaking exclusively from PC's standpoint, PC's other business units have little to do in that arena.
In the next few days I'm gonna start sniffing around for info on new-gen chips from existing manufacturers. Hopefully once Novak gets back from vacation and I finish up a huge PSU order we can get back to miner dev. I'm really hoping we can pull off having an actual machine (of one kind or another, whether it's just TypeZero boards or a complete unit) shipping in time for Christmas. That's kind of a long goal but not impossible.
An aggressive tentative schedule for the new chip would be tapeout by Dec/Jan. Proto boards by Feb. Begin production on a finalized board in Mar. Sales of the board could commence mid Mar.
So any boards GS might have available by the Christmas Holidays will likely not use the new chip from the chip dev coop.
One thing that seriously hampers the big Bitcoin mining ASIC vendors is that all of them (thus far) are "pure plays,"
Not technically true for Innosilicon, but seems to be true of all the others.
Might be why Innosilicon seems to be about the only Bitcoin or Litecoin ASIC folks that never did preorders.
I'm in the shop, taking a brief respite from mass soldering of 750W boards to refresh emails. The wallet and all documentation for miner stuff is in the office, where I'll be in a few hours and I'll take care of order confirmations and updating then. Right now the shop is only about 85F so I'm out here until it gets super hot.
Can you check pm Sidehak. Its for Canada Bulk just need a confirmation 
If whatever cooperative project PlanetCrypto is heading comes to manufacture and GekkoScience gets the contract, we'll be the manufacturer. I'd certainly be up for it. I'm sure Novak and I will probably both be in the coop helping out with hardware and software dev anyway. The chip dev coop will be a separate business entity from both PC and GS, but members of both businesses (and to some extent the businesses themselves) would also be members or participants in the chip dev coop.
In the next few days I'm gonna start sniffing around for info on new-gen chips from existing manufacturers. Hopefully once Novak gets back from vacation and I finish up a huge PSU order we can get back to miner dev. I'm really hoping we can pull off having an actual machine (of one kind or another, whether it's just TypeZero boards or a complete unit) shipping in time for Christmas. That's kind of a long goal but not impossible.
In the next few days I'm gonna start sniffing around for info on new-gen chips from existing manufacturers. Hopefully once Novak gets back from vacation and I finish up a huge PSU order we can get back to miner dev. I'm really hoping we can pull off having an actual machine (of one kind or another, whether it's just TypeZero boards or a complete unit) shipping in time for Christmas. That's kind of a long goal but not impossible.
My bad, i thought that the 2 companies were going to become one!
Basically PlanetCrypto is the client of a separate project where Gekko is the manufacturer.
Best luck to both so...
Basically PlanetCrypto is the client of a separate project where Gekko is the manufacturer.
Best luck to both so...
Independent from what PlanetCrypto is working on, we've been thinking about building a 4U machine made for swappable boards and such. With BM1384 it'd put up 5TH, so next-gen chips would be even better. The plan was to make the layout fairly flexible, internal connections a generic USB bus, and publish the specs in case anyone else wanted to build boards for the same form-factor. Basically, an intention to define an open standard. Once I get back to hardware dev we'll see what happens.
I've looked into this as well. Any thoughts on using something based on/derived from Eurocard (aka the IEEE 1101) family of standards?-a[g
Nope.
There's some pre-order sale threads over in Group Buys, and I'll be starting to bring in parts before long. PCBs and heatsinks should be in hand in about two weeks. Novak's on vacation and I'm tied up with assembly tasks for a huge PSU boards order so there probably won't be any dev work done on either hardware or code until next week. I think we're going to skip Amita design and go right into a simple multi-chip board with full software control of clock, voltage and fan. From there we can extend it to 18- and 30-chip designs.
Independent from what PlanetCrypto is working on, we've been thinking about building a 4U machine made for swappable boards and such. With BM1384 it'd put up 5TH, so next-gen chips would be even better. The plan was to make the layout fairly flexible, internal connections a generic USB bus, and publish the specs in case anyone else wanted to build boards for the same form-factor. Basically, an intention to define an open standard. Once I get back to hardware dev we'll see what happens.
Independent from what PlanetCrypto is working on, we've been thinking about building a 4U machine made for swappable boards and such. With BM1384 it'd put up 5TH, so next-gen chips would be even better. The plan was to make the layout fairly flexible, internal connections a generic USB bus, and publish the specs in case anyone else wanted to build boards for the same form-factor. Basically, an intention to define an open standard. Once I get back to hardware dev we'll see what happens.
To clarify:
Gekkoscience is a standalone entity as is Planetcrypto.
I can unequivocally say that any interaction between the proposed entity and PC will be contractual in nature as it has been between Gekkoscience and PC.
We, as a company, may elect to contribute, gratis, resources/assets to any company or individual (time on one of our super computers, for instance, for chip simulation) at our discretion.
PC will probably fund the proposed entity (subject to our share holder approval) through the share purchase vehicle as I will personally.
Currently, I'm leaning towards a corporate organizational model that files for an IRS 503(c)(6) tax status which is commonly known as a Mutual Benefit Nonprofit Corporation.
Briefly described here: https://en.wikipedia.org/wiki/Mutual-benefit_nonprofit_corporation
This will make investing in the organization less attractive to "whales", i.e. individuals looking to make huge ROI's in the short term.
But is, I think, more philosophically inline with what the Bitcoin mining community needs.
To quote from the Wiki referene:
"Mutual benefit corporations are formed for common gain purposes such as ..." "... promoting the social or economic welfare of member individuals or organizations (for example through trade groups, professional organizations or business districts)." In this instance the the proposed organization would be loosely defined as Bitcoin mining hardware manufacturers.
Stated succinctly, The proposed Mutual Benefit Nonprofit Corporation will promote the social and economic welfare (common gain) of member individuals and organizations by getting to tapeout a state-of-the-art Bitcoin SHA-256 hashing chip.
And yeah, PC will be interested in purchasing some of those 5TH machines as well.
maybe a good idea would be to contact an accountant, as nonprofit would have to dispose of any possible profit that might accrue.
As far as whales are concerned-the only "whales" I can think of possibly investing in a nonprofit venture like this might be exchanges or wallet providers (with or without some marketing endorsements).
A MBNC is allowed to retain earnings and roll that over into a new fiscal year, after the taxes are paid of course.
My degree background includes way too much accounting and for a number of years was an Enrolled Agent with the IRS (I let my certs lapse back in the late 90's).
Starting 1987 I wrote tax software (amongst other tings), and to have the S/W accepted by the IRS the coding team needed Enrolled Agent status.
Even 501(c)(3)'s are allowed to retain funds from year to year (operating expenses, salary expense, etc.)
A 503(c)(6) only has partial tax exempt status so contributions, i.e. share purchases would not be considered charitable contributions on an individual or corporate tax return.
heh, everyone pipes up once you mentioned 5TH unit.. to rich(probably more then what i can fish up in a day, and ~1.5KW to much for my electricity bill) for my usage though..
Since you said the card inside the system would be standard USB, sound like I could buy up 1 or 2 of them cards and run them standalone. 120mm by 120/240mm card would be nice, could just strap 2 large fans to it no problem. if its a 120mm2 board, could even encase it inside an old PSU chassis, or CD drive..
None the less, your projects, I'll wait and see how things turn out.
Since you said the card inside the system would be standard USB, sound like I could buy up 1 or 2 of them cards and run them standalone. 120mm by 120/240mm card would be nice, could just strap 2 large fans to it no problem. if its a 120mm2 board, could even encase it inside an old PSU chassis, or CD drive..
None the less, your projects, I'll wait and see how things turn out.
Independent from what PlanetCrypto is working on, we've been thinking about building a 4U machine made for swappable boards and such. With BM1384 it'd put up 5TH, so next-gen chips would be even better. The plan was to make the layout fairly flexible, internal connections a generic USB bus, and publish the specs in case anyone else wanted to build boards for the same form-factor. Basically, an intention to define an open standard. Once I get back to hardware dev we'll see what happens.
I've looked into this as well. Any thoughts on using something based on/derived from Eurocard (aka the IEEE 1101) family of standards?-a[g
Jump to: