mainline
Newbie
Offline
Activity: 56
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:20:37 PM Last edit: January 30, 2014, 07:31:06 PM by mainline |
|
this. all of this.
finflof you haven't replied to my pm about why you think my my earlier posts were so out of order. I find it concerning that you accept every word of this Mr Teal without question. His figures have not even been researched or cross-checked with the official figures as stated by Ken. Yet you agree with his every word? Every one? Because MrTeal knows what he's talking about.. and you like apples.
|
|
|
|
shaofis
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 106
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:21:40 PM |
|
Here is some back of the napkin numbers...
Ken estimates 1-2PH.
Current design indicates that the base unit holds 6 cards and expansion units hold 15 cards.
Current labels show a maximum build out of 1 base unit + 6 expansions for a total of 24.576TH.
Extrapolating that downward we can see that each card has 16 chips and at our new 55nm specs that equates to around 32GH per card. This means that your initial 4U produces 192GH and each expansion roughly 480GH. For a net total of 3TH per 28U of rackspace.
Now it's certainly possible that the boards are redesigned to hold more chips; simply can't say. Now there are some other legistics problems in a data center... how long are the interconnect cables? Can an expansion shelf be in a different rack than the head unit? Tossing all of that asside for a moment... at 3TH per 28U you are looking at 9334U for 1PH or 222 racks (@ 42U per rack) *not accounting for spread, network gear, etc.
Or if you want to assume worse case and go with 1 Set (1 head unit + 6 expansion) per rack you are at around 334 racks.
But all of that aside; the real challenge isn't getting power... it's going to be in the cooling at least in my opinion.
*** Assuming I did my maths right ***
|
|
|
|
finlof
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:22:22 PM |
|
this. all of this.
finflof you haven't replied to my pm about why you think my my earlier posts were so out of order. I find it concerning that you accept every word of this Mr Teal without question. His figures have not even been researched or cross-checked with the official figures as stated by Ken. Yet you agree with his every word? Every one? again you show your maturity by "cleverly" changing someone's name to something that tries to make it seem like they are less intelligent than you. keep doing it if you think it makes you look smarter. I did reply to you. some of us go to lunch and do actual work and dont keep pressing F5 like their lives depend on it. here is the reply so everyone can see it since he referenced it: EDIT - and yes I accept MrTeal's explanation and breakdown. much more so than I accept anything Ken says. Just because you (and some others) hope/need/want ActM to succeed and meet all projections by Ken doesnt make it true. And kleeck says it's feasible. I agree it is possible should all the planets align, but highly unlikely, especially given the way things have happened with this company so far. finlof,
I'm interested in what you think I did wrong on the last page? This guy said the 1-2 PH data centre could not be created from scratch (even by Q3) because of the power infrastructure needed to be built. I provided a link that shows you can hire a facility with all the power required for the machines. He then ignored this evidence and went on to say we can't get the electricity cheap enough.
So what is wrong in your opinion with me supplying evidence that proves this guy is completely wrong about the 1-2PH plan and the power infrastructure?
you provided a link that shows a 11700 sqft spot for lease. the entire building has 120 parking spaces, has 9 megawatt capacity, etc etc etc. this is for only 11700 sqft of a 400000 sqft building. so you obviously wouldnt have access to that amount of power, and whatever power you did have access to you would still have to build out the infrastructure of the rented space, including wiring it within the terms of the law (licensed electrician, etc). that lease price would certainly not include electricity/utilities, nor the cost for any renovations need to house 1 - 2 PH of mining equipment (quite pricy, time consuming and difficult, especially using 1.9gh/s chips). see his latest post in the forum for all the shit that would need to happen once chips were received. it really is next to impossible.
|
|
|
|
minerpart
Sr. Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:22:26 PM |
|
looks way better and more convincing than 3 empty workbenches in a garage
Apart from them having the chips in stock not at all. They have 6 assembly benches. We have three double sided benches = six assembly areas, so the same assembly capacity. Simply stating facts. EDIT - actually it looks like they have 5 benches so we have a larger 'assembly facility' than they do. Maybe we can put that on the website?
|
|
|
|
kleeck
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:23:28 PM |
|
Here is some back of the napkin numbers...
Ken estimates 1-2PH.
Current design indicates that the base unit holds 6 cards and expansion units hold 15 cards.
Current labels show a maximum build out of 1 base unit + 6 expansions for a total of 24.576TH.
Extrapolating that downward we can see that each card has 16 chips and at our new 55nm specs that equates to around 32GH per card. This means that your initial 4U produces 192GH and each expansion roughly 480GH. For a net total of 3TH per 28U of rackspace.
Now it's certainly possible that the boards are redesigned to hold more chips; simply can't say. Now there are some other legistics problems in a data center... how long are the interconnect cables? Can an expansion shelf be in a different rack than the head unit? Tossing all of that asside for a moment... at 3TH per 28U you are looking at 9334U for 1PH or 222 racks (@ 42U per rack) *not accounting for spread, network gear, etc.
Or if you want to assume worse case and go with 1 Set (1 head unit + 6 expansion) per rack you are at around 334 racks.
But all of that aside; the real challenge isn't getting power... it's going to be in the cooling at least in my opinion.
*** Assuming I did my maths right ***
I do think that the data center logistics are probably the most concerning element of the Q2 mining operation statement.
|
|
|
|
minerpart
Sr. Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:26:18 PM |
|
again you show your maturity by "cleverly" changing someone's name to something that tries to make it seem like they are less intelligent than you.
Sorry it was a typo, don't get so wound up. EDIT - and yes I accept MrTeal's explanation and breakdown. much more so than I accept anything Ken says.
OK so despite him getting basic figures out by 50% you believe him over Ken and his team. Well that's your right. I hope to be buying some shares off you soon then.
|
|
|
|
finlof
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:30:36 PM |
|
again you show your maturity by "cleverly" changing someone's name to something that tries to make it seem like they are less intelligent than you.
Sorry it was a typo, don't get so wound up. really, so all the times you have typed it that way and have done that to several other users on this forum have been nothing but typos? forgive me for thinking it was a deliberate debating tactic...
|
|
|
|
shaofis
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 106
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:31:07 PM |
|
I'd like to call something out as well.
So based on my math with the 55nm chip using 16 chips per card we come up around 3TH per "Set". Ken indicated 10.xxx before... but based on some quick math at 1.9GH per chip that's 5520 total chips or 57.5 chips per card. Packaged at 11x11 anyone care to guess the total size of a card to include 57.5 chips per card?
Am I missing something?
|
|
|
|
Minor Miner
Legendary
Offline
Activity: 2408
Merit: 1019
Be A Digital Miner
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:31:54 PM Last edit: January 30, 2014, 07:42:37 PM by Minor Miner |
|
minerpart/slaughter whoever you may be. You need to be quiet when people that know are speaking. There are few ways that the slaughters can avoid prosecution but one of them is to actual get something going. When people are posting information that ken needs to read and consider, you need to stay off the thread. What you have in the last two pages are MANY pieces of information that Ken will need to know (and likely does not) and it might just help him get something done. In addition to that, there is your pollution. the last thing bitcoin needs is more blowups with whiny little bitches like you complaining to the press that you lost your investment. The more you post about details you do not understand, the more people who do know things realize what a fool you are. My greatest fear is that you actually might be a slaughter because if you are, there is no chance this ends well. Arrogance combined with ignorance rarely breeds success. It is time Ken starts asking for help so he can get something done. There are tons of people here that have already done what he needs to do and many of them would be willing to help. Want to make some BTC Minerpart? Find me this 1.7 cents per kW/h electricity in the USA (all in cost for baseloaded power, that does not require an economic development review (ie. you do not have to hire xx number of people and start a factory since a data center does not do that) and I will give you 5 btc. If you are interested in spending your time doing that, tell me and I will define fully what you need to do (the hurdle will not be high, I will just need confirmable proof from the utility, or federal power authority so after you find it, you will need to budget a day getting me the package together).
|
|
|
|
MrTeal
Legendary
Offline
Activity: 1274
Merit: 1004
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:39:32 PM |
|
Let's play with some numbers, shall we? Please fill in the blanks.
A ridiculously long post, and it's filled with assumptions, inaccuracies and unanswerable speculative questions. How long is a piece of string? -100PH by when? Who determined that? Do you have that in writing and are you considering the possibility of a plateau in difficulty rises? So that is a guess with a large margin of error which you fail to note. Total hash could be 75PH by then which would reduce the required capacity of our farm by 25% - future hash rate it is an unknown figure. I didn't say anything about the difficulty or how profitable it would be. My post entirely focused on the feasibility of bringing 1PH/s online by the end of Q2. -The 4U expansion units are not 1TH each, there are 6 expansion cases and one master case for our full-sized 55nm-based 10.488TH/s rig giving a figure of 1.498TH/s per unit. You are such an expert (proved substantively wrong twice already today) but you couldn't work that out with available figures. So your guestimate on this figure is 50% out - because you didn't do the research.
As I said, it was a number picked to keep the math simple. I wouldn't be so quick to assume the 1498GH/s will actually get into a 4U chassis though, my simplified number might actually be closer to the truth. You're looking at 2kW just in the ASICs for that kind of hashrate and somewhere north of 2500W for the whole chassis. -So we knock 33% off your 16k boards figure to give 10,666 boards. Suddenly outlay and workload and investment in resources is reduced by a third - and that is fact not a guestimate.
Sure, if you want to. It was just a back of the hand number I picked to show scale. Your 10,666 boards isn't any more realistic though, unless Ken has confirmed the number of hashing boards in each 1.498TH/s unit. -Wafer numbers - the UMC foundry we are using has a capacity of 45,000 12inch wafers per month. I can't find details on how many chips per wafer UMC will be fabbing.
What does the capacity of UMC have to do with the time it takes to finish a couple hundred wafers. It's not like if it takes 10 days at a fab that can do 4,500 wafers/day that UMC could do it in 1. -The rest of your post is speculation. What are my time estimates on 10k boards? I don't have any, I'm not a PCB fabricator you would need to ask Ken's engineering team that one. Certainly your knowledge of the industry seem a bit sub-par so with respect I'll trust Ken's expert team on their estimates for this project over yours.
Of course the post is speculation. The company isn't transparent enough to make plans with anything resembling the full information. For the PCB time, for that quantity sub 1-week leads would probably be off the table at most vendors. I haven't quoted 10k pieces before (from the 3rd largest manufacturer in the US, which specializes in quicker turns), but I have gotten 1700pc quotes and while you might be able to shave a day or two off at that volume, be prepared to pay double or more for the privilege. Realistically you'd probably be looking at getting them made in China and taking a couple weeks, as doing an order that large with a quick turn in the US would probably cost you an extra couple hundred thousand dollars. But I digress, it seems you've entirely missed the point of my post. It had nothing at all to do with the feasibility of making a 1PH/s mine. That is entirely possible, and there will probably be a couple 1PH/s mines around by the end of the year. It was only discussing the possibility of getting a 1PH/s mine up and running by the end of Q2.
|
|
|
|
breadcrumbs
Newbie
Offline
Activity: 27
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:40:10 PM |
|
1) Whine whine whine
2) We all know who the real sock puppet monkeys are
3) You boys claim to be adults but then make big dramatic announcements about whom you are now ignoring
4) Crumbs under his current name still gets off on getting a rise out of you
5) What is really important is that the CEO of Active Mining has PROMISED to start verifying shares today
6) That means that soon trading will resume
7) That is a very good thing
8 ) The CEO of Active Mining has also assured us that he has some real and competent engineers and a good project manager working for us
9) We do not have to be #1.
10) We all will make money even if we are #4 in the mining game
11) All other speculation is just fodder for crumbs-mainline
12) We will all be rich or at least richer than we were before Mr. Slaughter met us
|
|
|
|
finlof
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:44:14 PM |
|
again you show your maturity by "cleverly" changing someone's name to something that tries to make it seem like they are less intelligent than you.
Sorry it was a typo, don't get so wound up. EDIT - and yes I accept MrTeal's explanation and breakdown. much more so than I accept anything Ken says.
OK so despite him getting basic figures out by 50% you believe him over Ken and his team. Well that's your right. I hope to be buying some shares off you soon then. as to his "getting basic figures out by 50%" - you are saying there will be an average of close to 800 of the 55nm chips in each 4u case? I'd really like to see how that works out. no really. that's a lot of power and cooling needs. 800 chips x 2.5 watts = 2000 watts just for the chips. you can have all of my 416 shares. if you do any trading on cryptostocks that is. what did you say activemining was worth again? you can buy them for that 0.01 price...
|
|
|
|
Subcide
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 70
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:44:26 PM |
|
1. Stop making numbered lists please, it's silly. 2. That is all.
|
|
|
|
minerpart
Sr. Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:45:43 PM |
|
really, so all the times you have typed it that way and have done that to several other users on this forum have been nothing but typos? forgive me for thinking it was a deliberate debating tactic...
That time it was a typo - I'm trying to be more sensible and just present the facts as reported to us by this company. If we don't actually know the stated facts first how can we question them? At the moment people are believing the lies or half-truths spread by new-comers to this thread and so-called experts who have all of a sudden become very very interested in ACtM. Then people question these half-truths as if they were facts. That leads you no-where.
|
|
|
|
JoTheKhan
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:46:14 PM |
|
Ad Hominem attacks are unnecessary on this thread.
Ken can you give us more insight on the share transfer process; anything we need to do and when you plan to have it finished it by?
|
|
|
|
finlof
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:49:17 PM |
|
really, so all the times you have typed it that way and have done that to several other users on this forum have been nothing but typos? forgive me for thinking it was a deliberate debating tactic...
That time it was a typo - I'm trying to be more sensible and just present the facts as reported to us by this company. If we don't actually know the stated facts first how can we question them? At the moment people are believing the lies or half-truths spread by new-comers to this thread and so-called experts who have all of a sudden become very very interested in ACtM. Then people question these half-truths as if they were facts. That leads you no-where. the fact they you take them as facts from Ken, but then when MrTeal (who has been around much longer and has some real street cred) posts his take on it you tell him he cant be right cause Ken said this - it's just sad really. You are coming off like a 9 year old boy saying my dad can beat your dad up. EDIT - tried to made the 1st sentence make more sense.
|
|
|
|
minerpart
Sr. Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 07:54:09 PM |
|
the fact they you take them as facts from Ken, and then when MrTeal (who has been around much longer and has some real street cred) posts his take on it you tell him he cant be right cause Ken said this. You are coming off like a 9 year old boy saying my dad can beat your dad up.
Ken has the facts and details at hand, he is a very experienced software engineer, (as well as a very experienced businessman) he has two life-time engineers on his team, together with their suppliers they have come up with these projections. Mr Teal does not have all the details about this company infront of him, he has not negotiated with ACtM's suppliers or know the agreed timelines, and neither is he as knowledgeable as a combination of 3 life-time engineers.
|
|
|
|
MrTeal
Legendary
Offline
Activity: 1274
Merit: 1004
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 08:07:09 PM |
|
I don't see anything in this that indicates that chips arriving in Q2 could not be mining within that same quarter, i.e. 3 months time.
This is not a pipe dream. We are not taming unicorns, here. I'm not saying that for absolute sure ActM will have a 1-2PH farm in Q2, but it is in the realm of possibility in this day and age.
The design is done. Tape-out indicates the final step of the design process, once all details have been worked out, (although certain things can come up that require re-spins). Once we have the first article we will know if any re-spins are necessary or not. Board design is done long before Q2 and might very well be done already. Pick and place is lightning fast and completely automated. Placed boards will be tested and be sent to be placed in the actual machines, which will be a second point of testing, ensuring that the boards work with the rest of the device and mining software.
This has been done so many times long before Bitcoin that it is comical how so many think of this as some unattainable or unreachable goal.
The design of the chip is done. The design of the board still seems to be underway, but it should definitely be done by the time chips roll off the line. As for the speed of the pick and place machines I'd agree, although it would probably still take a couple/few days to get all those boards through. There is still the setup for the lines though and there is kitting and other prep work to do, and even the waiting to get on a line. Any good assembly house runs at a pretty high capacity factor, and some order like this is not substantial enough that they're going to kick another client off a line in the middle of their run so that they can immediately start on yours. It's not impossible for AMC/VMC to get working hardware into people's hands by the end of Q2. Hell, if they want to really pay they could probably even get 1PH/s online by June 30th of they get the chips April 1st if everything goes extremely well and they're willing to take some possibly expensive risks, at the cost of really raising their average unit price, and absolutely nothing goes wrong. I would disagree that this happens so many times long before Bitcoin though. The insane standing starts we see here are due to the time sensitive nature of the Bitcoin world. If Ken were designing cable modems using a custom ASIC or something the idea of have 10,000 units in customers homes 3 months after the first sample chip rolls off the line would be crazy.
|
|
|
|
kleeck
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 08:20:28 PM |
|
I don't see anything in this that indicates that chips arriving in Q2 could not be mining within that same quarter, i.e. 3 months time.
This is not a pipe dream. We are not taming unicorns, here. I'm not saying that for absolute sure ActM will have a 1-2PH farm in Q2, but it is in the realm of possibility in this day and age.
The design is done. Tape-out indicates the final step of the design process, once all details have been worked out, (although certain things can come up that require re-spins). Once we have the first article we will know if any re-spins are necessary or not. Board design is done long before Q2 and might very well be done already. Pick and place is lightning fast and completely automated. Placed boards will be tested and be sent to be placed in the actual machines, which will be a second point of testing, ensuring that the boards work with the rest of the device and mining software.
This has been done so many times long before Bitcoin that it is comical how so many think of this as some unattainable or unreachable goal.
The design of the chip is done. The design of the board still seems to be underway, but it should definitely be done by the time chips roll off the line. As for the speed of the pick and place machines I'd agree, although it would probably still take a couple/few days to get all those boards through. There is still the setup for the lines though and there is kitting and other prep work to do, and even the waiting to get on a line. Any good assembly house runs at a pretty high capacity factor, and some order like this is not substantial enough that they're going to kick another client off a line in the middle of their run so that they can immediately start on yours. It's not impossible for AMC/VMC to get working hardware into people's hands by the end of Q2. Hell, if they want to really pay they could probably even get 1PH/s online by June 30th of they get the chips April 1st if everything goes extremely well and they're willing to take some possibly expensive risks, at the cost of really raising their average unit price, and absolutely nothing goes wrong. I would disagree that this happens so many times long before Bitcoin though. The insane standing starts we see here are due to the time sensitive nature of the Bitcoin world. If Ken were designing cable modems using a custom ASIC or something the idea of have 10,000 units in customers homes 3 months after the first sample chip rolls off the line would be crazy. You're right. It is not entirely true that "this" has been happening for long. These stages of the complete process have been around for a while, but the time short time frames are a new element that really change the game.
|
|
|
|
finlof
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
January 30, 2014, 08:28:39 PM |
|
the fact they you take them as facts from Ken, and then when MrTeal (who has been around much longer and has some real street cred) posts his take on it you tell him he cant be right cause Ken said this. You are coming off like a 9 year old boy saying my dad can beat your dad up.
Ken has the facts and details at hand, he is a very experienced software engineer, (as well as a very experienced businessman) he has two life-time engineers on his team, together with their suppliers they have come up with these projections. Mr Teal does not have all the details about this company infront of him, he has not negotiated with ACtM's suppliers or know the agreed timelines, and neither is he as knowledgeable as a combination of 3 life-time engineers. if you are going to state these as facts then I ask for you to provide proof to back up your statements of: 1. he is a very experienced software engineer 2. a very experienced businessman 3. he has two life-time engineers on his team 4. together with their suppliers they have come up with these projections EDIT - not saying any of these arent possibly true, just want this backed up with proof.
|
|
|
|
|