somestranger
|
|
January 06, 2014, 08:47:49 PM |
|
When you didn't have functioning chips in October as originally forecast,
Source? This isn't "months and months of no show". We were supposed to start getting chips (low volume production?) at the end of November. It has been less than 1.5 months since then
We were supposed to be in full production in end of Nov (before first delay). Source?
|
|
|
|
stenkross
|
|
January 06, 2014, 08:51:03 PM |
|
Source?
This thread, around mid-nov.
|
|
|
|
somestranger
|
|
January 06, 2014, 08:51:35 PM |
|
Source?
This thread, around mid-nov. Show me. I bet you can't.
|
|
|
|
stenkross
|
|
January 06, 2014, 08:53:27 PM |
|
Show me. I bet you can't.
Bet all you want, it's here if Ken has not removed those posts.
|
|
|
|
Vigil
|
|
January 06, 2014, 08:57:58 PM |
|
I don't get it. Sure, you might be a little late. But months and months of no show or nonfunctional product?
This isn't "months and months of no show". We were supposed to start getting chips (low volume production?) at the end of November. It has been less than 1.5 months since then and Ken implied that full production of the Intellihash(tm) chip has started or is starting soon based on his weekly update from Christmas. I'm not just talking about ActM, I am talking about all the ASIC ventures. There have been so many that just failed to materialize when we are really building glorified microcontroller boards that you see on the DIY projects at on MakerMagazine. These aren't Ati GPUs or Asus Motherboards or anything.
|
|
|
|
Vigil
|
|
January 06, 2014, 09:00:17 PM |
|
Assuming the production schedule started on Sep. 5th (the press release day) we are right at the beginning of week 18. It was 16-18 weeks till full production. I don't know how much of the orders were to be fulfilled by the low volume production but we haven't even passed the full-scale production timeline yet. Obviously, the Oct/Nov delivery estimate was for whatever orders were filled by the sample chips and low volume production. If there were sample chips delivered and protoboards then it is possible that some of the early buyers (maybe those involved int eh project and ken) have something hashing.
|
|
|
|
Vigil
|
|
January 06, 2014, 09:23:40 PM |
|
High power and high clock rate boards in volume isn't easy.
Sure you can build a microcontroller at home and get the hardware and firmware debugged in a month or 2 of tinkering, but delivering in volume is a different kettle of fish.
This whole market has moved from a hobby to serious engineering enterprises in 9 months. Lots of folks are going to fall as the stakes increase over the next year.
Ken has missed his window. It was at best 2-3 months wide to begin with.
We are basically going to have 3 ASIC manufacturers. How much faster can the technology get though? Aren't we coming to a roadblock?
|
|
|
|
Bargraphics
|
|
January 06, 2014, 09:42:16 PM |
|
High power and high clock rate boards in volume isn't easy.
Sure you can build a microcontroller at home and get the hardware and firmware debugged in a month or 2 of tinkering, but delivering in volume is a different kettle of fish.
This whole market has moved from a hobby to serious engineering enterprises in 9 months. Lots of folks are going to fall as the stakes increase over the next year.
Ken has missed his window. It was at best 2-3 months wide to begin with.
We are basically going to have 3 ASIC manufacturers. How much faster can the technology get though? Aren't we coming to a roadblock? There are still 2 process nodes - 22/20 nm and 14 nm to go. On top of that there is room for design optimizations. How much can actually be squeezed out is tough to say with any accuracy. There are a lot more than 3 suppliers already. Let me list With known functional hardware: ASICMINER Avalon BFL Bitfury KNC AMT Shipping this month (high confidence, and proof shown) Hashfast CoinTerra Expected soon: Bitmine.ch Black Arrow Activemining (?) Rumored 21e6 I expect there are other groups running development projects targeting 20 or 14 nm nodes in stealth mode funded by VC dollars. I also wouldn't rule out AMD offering a hashing engine for license at 14 nm. The market is getting that large. AMT is using bitfury chips and is going to be using bitmine.ch's chips Bitmine.ch has proof of chips - http://bitmine.ch/?p=4543You forgot Bitmain's AntMiner's are currently shipping.
|
|
|
|
mannybitcoins
|
|
January 06, 2014, 09:46:26 PM |
|
Guys, Ken isn't dumb. He obviously has miners ready to go but will be using them to mine Coinye West coins instead. TO THE MOON!!!!
|
|
|
|
Subcide
Member
Offline
Activity: 70
Merit: 10
|
|
January 06, 2014, 10:45:20 PM |
|
Now that bitmine.ch blog post is what our updates SHOULD be. Ah well. Sit and wait for wednesday's "Soon" update.
|
|
|
|
shadallion
|
|
January 07, 2014, 12:07:30 AM |
|
Perfectly willing for Ken to stop all other core business activities for a day if he would just list all the shares on Crypto-Trade so that the doubting Thomases can all divest.
|
|
|
|
Flashman
|
|
January 07, 2014, 12:11:22 AM |
|
Post up Ken or bring umbrellas for us shareholders, it's a troll bukkake fest in here.
|
TL;DR See Spot run. Run Spot run. .... .... Freelance interweb comedian, for teh lulz >>> 1MqAAR4XkJWfDt367hVTv5SstPZ54Fwse6
Bitcoin Custodian: Keeping BTC away from weak heads since Feb '13, adopter of homeless bitcoins.
|
|
|
wasubii
|
|
January 07, 2014, 01:46:15 AM |
|
I thought i would just pop my head back up over the parapet to try and steer this back on topic. it would be great if some of the thinkers in this thread could weigh in on this:
If we do not get hashing soon, there will become a point where we will not be able to keep up with difficulty increases.
We will have a fixed capacity for production in the short term, and we will also run into efficiency issues with our 20GH/s chips in that heat, power, physical space for thousands of small, low power chips will be a lot more costly and difficult to manage than the same hash rate of powerful chips (like Cointerrra's 500GH/s Goldstrike chip).
I base this particularly on the very recent example of ASICminer, who at one time had 40% of global hash rate due to being quick to market with their small, inefficient ASICs. Despite the huge amount of BTC rolling in during this period, they faced issues over supplying power to so many boards, cooling so many boards and actually the physical space needed for so many boards. Now they don't even show on the pie chart of hashing distributions. So quick and so fast their fall from grace has been.
This is because efficiency of ASIC mining is not ONLY about power consumption of chips.
It is about efficiency of deployment
There are two ways i can see ActM will fail to keep up with the global hash rate.
1) We begin mining but we do not mine enough coins to pay for the equipment we need to keep up with the difficulty increases.
2) We begin mining and make enough coin, but our production capacity is too small to keep up with the difficulty increases (this will happen eventually if we do not expand capacity, but the way things are going it may kill us before we even begin)
BTC ASIC companies that have failed to deliver on time (BFL, Hashfast, Cointerra etc etc) are usually the ones who do the assembly themselves.
Contrast that with KnC or ASICminer (in the early days) who asked one company to design and produce the chip, and another company to assemble the chips onto boards. They were on time because they dealt with reputable partners with proven track records and lots of experience. They were left to orchestrate and market their products while the nitty gritty of production was left to professionals.
What worries me with ActM is that we have a very small family team of non-engineers that are presumably going to assemble 'The biggest BTC mine on the planet' from their 5 benches in a garage (I'm not trolling, I'm pointing out facts - please correct me if I'm wrong)
So my question to you all is, at what point are we too late? At what point are our piddly little 20GH/s chips as useful as ASICminers old 0.3Gh/s chips?
Remember, AM could have continued pumping out millions of tiny hash rate chips but they ran into huge problems in their efficiency of deployment
I welcome any sensible comments. I would love to hear your estimates on how many 25GH/s platinum modules could be made in one month based on our own assembly (assume the chips roll in with no problem).
|
|
|
|
kleeck
|
|
January 07, 2014, 01:54:48 AM |
|
I thought i would just pop my head back up over the parapet to try and steer this back on topic. it would be great if some of the thinkers in this thread could weigh in on this:
If we do not get hashing soon, there will become a point where we will not be able to keep up with difficulty increases.
We will have a fixed capacity for production in the short term, and we will also run into efficiency issues with our 20GH/s chips in that heat, power, physical space for thousands of small, low power chips will be a lot more costly and difficult to manage than the same hash rate of powerful chips (like Cointerrra's 500GH/s Goldstrike chip).
Let us not forget the very recent example of ASICminer, who at one time had 40% of global hash rate due to being quick to market with their small, inefficient ASICs. Despite the huge amount of BTC rolling in during this period, they faced issues over supplying power to so many boards, cooling so many boards and actually the physical space needed for so many boards. Now they don't even show on the pie chart of hashing distributions. So quick and so fast has been their fall from grace.
This is because efficiency of ASIC chips is not ONLY about power consumption.
It is about efficiency of deployment
There are two ways i can see ActM will fail to keep up with the global hash rate.
1) We begin mining but we do not mine enough coins to pay for the equipment we need to keep up with the difficulty increases.
2) We begin mining and make enough coin, but our production capacity is too small to keep up with the difficulty increases (this will happen eventually, but the way things are going it may kill us before we even begin)
BTC ASIC companies that have failed to deliver on time (BFL, Hashfast, Cointerra etc etc) are usually the ones who do the assembly themselves.
Contrast that with KnC or ASICminer (in the early days) who asked one company to design and produce the chip, and another company to assemble the chips onto boards. They were on time because they dealt with reputable partners with proven track records and lots of experience.
What worries me with ActM is that we have a very small family team of non-engineers that are presumably going to assemble 'The biggest BTC mine on the planet' from their 5 benches in a garage (I'm not trolling, I'm pointing out facts - please correct me if I'm wrong)
So my question to you all is, at what point are we too late? At what point are our piddly little 20GH/s chips as useful as ASICminers old 0.3Gh/s chips?
Remember, AM could have continued pumping out millions of tiny hash rate chips but they ran into huge problems in their efficiency of deployment
I welcome any sensible comments
I appreciate your effort to steer this thread in the proper direction. I believe your concerns are appropriate. Much of the future of ActM lies in the efficiency and power of the Intellihash chips along with a properly architected reinvestment strategy. The later is probably the most important, in terms of the AMC Mining Operation. What worries me with ActM is that we have a very small family team of non-engineers that are presumably going to assemble 'The biggest BTC mine on the planet' from their 5 benches in a garage (I'm not trolling, I'm pointing out facts - please correct me if I'm wrong) This does not appear to be the case at all. The assembly area is not for engineering boards or even placing chips on boards. The board development and the pick and place process are being handled by a third party engineering firm that has been spoken highly of. I'll try to track down the name. Ken specifically told me that the pick and place process would be handled outsourced to the engineering firm. It would seem that what the assembly area is actually doing is placing pre-made/placed boards into the cases, firing them up, burning them in, boxing them and shipping them out. I'll think about assembly time/volume for enough Platinums to compete. I think we're mostly limited by money in this regard though. It all comes down to the reinvestment strategy. I agree with the heart of your post: "We need to be hashing."
|
|
|
|
wasubii
|
|
January 07, 2014, 02:08:49 AM |
|
I thought i would just pop my head back up over the parapet to try and steer this back on topic. it would be great if some of the thinkers in this thread could weigh in on this:
-snip-
I welcome any sensible comments
I appreciate your effort to steer this thread in the proper direction. I believe your concerns are appropriate. Much of the future of ActM lies in the efficiency and power of the Intellihash chips along with a properly architected reinvestment strategy. The later is probably the most important, in terms of the AMC Mining Operation. What worries me with ActM is that we have a very small family team of non-engineers that are presumably going to assemble 'The biggest BTC mine on the planet' from their 5 benches in a garage (I'm not trolling, I'm pointing out facts - please correct me if I'm wrong) This does not appear to be the case at all. The assembly area is not for engineering boards or even placing chips on boards. The board development and the pick and place process are being handled by a third party engineering firm that has been spoken highly of. I'll try to track down the name. Ken specifically told me that the pick and place process would be handled outsourced to the engineering firm. It would seem that what the assembly area is actually doing is placing pre-made/placed boards into the cases, firing them up, burning them in, boxing them and shipping them out. I'll think about assembly time/volume for enough Platinums to compete. I think we're mostly limited by money in this regard though. It all comes down to the reinvestment strategy. I agree with the heart of your post: "We need to be hashing." Good to hear from you again Kleeck Thanks for pointing out the pick and place will be outsourced (as it surely must be, being a highly mechanised process). But i was actually talking about the assembly of the individual PCI modules. Surely they need to be paired with fans at the very least, then tested before they can go into the cases. Or will the pick and place company deliver us a final finished PCI module ready to slot in? That is why i was hoping someone with more knowledge than me could estimate how long it would take to assemble each PCI module before it gets assembled into the larger cases? If we are being delivered a finished PCI product ready to slot in then i agree, it will come down to the money we need to buy components and the lead times for chips and finished PCI boards.
|
|
|
|
Vigil
|
|
January 07, 2014, 03:46:02 AM Last edit: January 07, 2014, 04:00:35 AM by Vigil |
|
This is why I was trying to contact StilWell Baker... to see if they are involved in the board assembly at all. Ken isn't going to have his family flowing eASIC BGA onto the boards. or soldering transistors and resistors. They are just going to plug stuff in and use electric drills.
|
|
|
|
drawingthesun
Legendary
Offline
Activity: 1176
Merit: 1015
|
|
January 07, 2014, 04:13:45 AM |
|
This is why I was trying to contact StilWell Baker... to see if they are involved in the board assembly at all. Ken isn't going to have his family flowing eASIC BGA onto the boards. or soldering transistors and resistors. They are just going to plug stuff in and use electric drills.
Well if we are using StilWell Baker that is very cool, I would expect them to be able to handle everything except building the ASIC. If we are using StilWell Baker they could easily manufacture the entire card and deliver it to Ken as a product.
|
|
|
|
Vigil
|
|
January 07, 2014, 04:28:06 AM |
|
This is why I was trying to contact StilWell Baker... to see if they are involved in the board assembly at all. Ken isn't going to have his family flowing eASIC BGA onto the boards. or soldering transistors and resistors. They are just going to plug stuff in and use electric drills.
Well if we are using StilWell Baker that is very cool, I would expect them to be able to handle everything except building the ASIC. If we are using StilWell Baker they could easily manufacture the entire card and deliver it to Ken as a product. That is what I was thinking but I haven't heard back from them yet. Maybe we could try calling or contacting Ted or whatever his name is.
|
|
|
|
wasubii
|
|
January 07, 2014, 06:26:16 AM |
|
This is why I was trying to contact StilWell Baker... to see if they are involved in the board assembly at all. Ken isn't going to have his family flowing eASIC BGA onto the boards. or soldering transistors and resistors. They are just going to plug stuff in and use electric drills.
Well if we are using StilWell Baker that is very cool, I would expect them to be able to handle everything except building the ASIC. If we are using StilWell Baker they could easily manufacture the entire card and deliver it to Ken as a product. I hope our chip delays haven't messed up our position in the StilWell production queue
|
|
|
|
CanadianGuy
|
|
January 07, 2014, 06:56:01 AM |
|
This is why I was trying to contact StilWell Baker... to see if they are involved in the board assembly at all. Ken isn't going to have his family flowing eASIC BGA onto the boards. or soldering transistors and resistors. They are just going to plug stuff in and use electric drills.
Well if we are using StilWell Baker that is very cool, I would expect them to be able to handle everything except building the ASIC. If we are using StilWell Baker they could easily manufacture the entire card and deliver it to Ken as a product. I hope our chip delays haven't messed up our position in the StilWell production queue You have absolutely NOTHING to worry about. Everything will be on schedule.
|
|
|
|
|