minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 07:02:35 PM |
|
Two points: KnC are not 'at 20nm'. KnC are working on a 20nm but that ASIC technology is still unproven. There is no guarantee that the reduced die size with limiting non-scalable components will allow for any significant increase in overall efficiencies over top-end 28nm. (Like the ACtM eASIC full custom 28nm). Infact, the increased cost of producing the KnC 20nm chip, plus the inevitable delays in actually designing and achieving a workable 20nm die (extremely difficult) could make KnC's Neptunes unable to ROI. That is a real possibility. Second point - foundry capacity. Let's see some figures about the ability of the foundries producing bitcoin ASIC's being able to keep up with 100% production increase per month. No matter what their capacity now they are unable to meet that future demand - fact. UMC currently has a huge capacity for 55nm but their total 28nm capacity seems to be only around 25% above that. So with a huge demand (100% increase per month) for 28nm, 28nm production limits will soon be met. http://www.umc.com/English/class_300/c.aspWhat that says is that very high demand for 28nm could reach a foundry back-log in the near future. That's why chip runs are indeed limited by the foundries - they do not have endless capacity when the world of electronics of all types runs off their production lines. 28nm is in high demand. 55nm is an established tech but with huge infrastructure supporting it and it does not have a 100% increase in market demand per month. 'Yau Kae Sheu, senior director of UMC's 12-inch Specialty Technology Development division, said, "UMC first taped-out customer 55nm SDDI products in late 2012. Achieving 15 million accumulated shipments and mature production yields in just one year is a significant milestone and highlights our engineering and manufacturing strength as a worldwide foundry specialty technology leader. With our 300mm Fab 12A in Tainan, Taiwan and Singapore's 300mm Fab 12i both providing ample capacity support and established economies of scale, we look forward to bringing additional 55nm products to volume production as many new products are scheduled for the design-in stage in early 2014'http://www.prnewswire.com/news-releases/umc-surpasses-15-million-shipments-for-customer-55nm-sddi-chips-239014911.htmlSo the addition to our portfolio of a custom, top-end, 55nm that has a very high volume production capacity from UMC could be a very very fortunate - or smart - development for VMC. If our competitors (X-farm mining) can only source 100k chips per quarter while we can source 300k chips per quarter and our chips cost 60% of the cost to fab and run in their life-time than X-farms chips, then we can grow our farm above and beyond the capacity of X-farm to grow theirs.
|
|
|
|
kleeck
|
|
January 27, 2014, 07:07:09 PM |
|
Two points: KnC are not 'at 20nm'. KnC are working on a 20nm but that ASIC technology is still unproven. There is no guarantee that the reduced die size with limiting non-scalable components will allow for any significant increase in overall efficiencies over top-end 28nm. (Like the ACtM eASIC full custom 28nm). Infact, the increased cost of producing the KnC 20nm chip, plus the inevitable delays in actually designing and achieving a workable 20nm die (extremely difficult) could make KnC's Neptunes unable to ROI. That is a real possibility. Second point - foundry capacity. Let's see some figures about the ability of the foundries producing bitcoin ASIC's being able to keep up with 100% production increase per month. No matter what their capacity now they are unable to meet that future demand - fact. UMC currently has a huge capacity for 55nm but their total 28nm capacity seems to be only around 25% above that. So with a huge demand (100% increase per month) for 28nm, 28nm production limits will soon be met. http://www.umc.com/English/class_300/c.aspWhat that says is that very high demand for 28nm could reach a foundry back-log in the near future. That's why chip runs are indeed limited by the foundries - they do not have endless capacity when the world of electronics of all types runs off their production lines. 28nm is in high demand. 55nm is an established tech but with huge infrastructure supporting it and it does not have a 100% increase in market demand per month. 'Yau Kae Sheu, senior director of UMC's 12-inch Specialty Technology Development division, said, "UMC first taped-out customer 55nm SDDI products in late 2012. Achieving 15 million accumulated shipments and mature production yields in just one year is a significant milestone and highlights our engineering and manufacturing strength as a worldwide foundry specialty technology leader. With our 300mm Fab 12A in Tainan, Taiwan and Singapore's 300mm Fab 12i both providing ample capacity support and established economies of scale, we look forward to bringing additional 55nm products to volume production as many new products are scheduled for the design-in stage in early 2014'http://www.prnewswire.com/news-releases/umc-surpasses-15-million-shipments-for-customer-55nm-sddi-chips-239014911.htmlSo the addition to our portfolio of a custom, top-end, 55nm that has a very high volume production capacity from UMC could be a very very fortunate - or smart - development for VMC. If out competitors (X-farm mining) can only source 100k chips per quarter while we can source 300k chips per quarter and our chips cost 60% of the cost to fab and run in their life-time than X-farms chips, then we can grow our farm above and beyond the capacity of X-farm to grow theirs. http://www.taipeitimes.com/News/biz/archives/2013/12/13/2003578882This also makes me want to invest in a foundry. Edit: I like your post.
|
|
|
|
somestranger
|
|
January 27, 2014, 07:15:04 PM |
|
So the addition to our portfolio of a custom, top-end, 55nm that has a very high volume production capacity from UMC could be a very very fortunate - or smart - development for VMC. If our competitors (X-farm mining) can only source 100k chips per quarter while we can source 300k chips per quarter and our chips cost 60% of the cost to fab and run in their life-time than X-farms chips, then we can grow our farm above and beyond the capacity of X-farm to grow theirs.
Yes the foundries may be able to pump out chips, but will the board manufacturers be able to keep up? I don't mean producing the PCBs -- that should be easily scalable, but actually mounting the chips could be labor intensive especially since the 55nm ones are only ~1.2 cm^2.
|
|
|
|
minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 07:21:11 PM |
|
somestranger- don't know about that aspect, aren't they pushed on enmass by machine and soldered in solder baths? I thought it was all machine done?
Great link kleeck:
'The company [TSMC], which commands more than 90 percent of the global 28nm chip market, said earlier this year that 28nm chips would be the biggest contributor to revenue this year, since production capacity and revenue are set to triple on an annual basis.'
So the global production of 28nm will rise by around 330% maximum in 2014. That's a fact.
EDIT - while the TSMC press release makes the future roll out of 20nm look promising for around March 2014, there is yet no guarantee of the chips efficiencies (and certainly not massive efficiencies) over top-end 28nm.
|
|
|
|
wolfbiter
Newbie
Offline
Activity: 11
Merit: 0
|
|
January 27, 2014, 07:34:11 PM |
|
You thought correct.
Ok thanks! Where can I find them? I tried sifting through the threads but didn't find anything except mentions of cryptostocks.com
|
|
|
|
wasubii
|
|
January 27, 2014, 07:40:52 PM Last edit: January 27, 2014, 08:08:16 PM by wasubii |
|
Oh my god - are you guys freakin serious?
You're talking about millions of dollars of mining equipment...chips, boards and such somehow becoming a bottleneck in a 1/4 trillion dollar semiconductor industry?
Seriously?
Lets play out the figures shall we
Say the wholesale cost price of chips, boards controllers etc are $2/GH for the finished product. Forget splitting it down into chips and boards for simplicity's sake.
A global combined investment of 1 billion dollars in 2014 would yield around 500,000,000 GH of mining power. Thats 500 PH compared to the current network hash rate of 17.5 PH
1 billion dollars of investment in mining equipment could easily be swallowed by the global electronic industry - it wouldn't even blink.
EDIT: corrected GH to PH - thanks PurpleTentacle
|
|
|
|
minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 07:40:56 PM |
|
Another worry about the Neptune is the power-draw. This guy thinks it will be over 4500Watts and so too load much for many domestic settings. One answer - to wire it into your electric cooker outlet would not be possible or acceptable for many. http://coinchomp.com/2014/01/06/the-knc-miner-neptune-should-consume-at-least-4500-watts/So the point here is ACtM are the only mining industry company with uncertainty over the viability of their product. I think a mass refund request on the KnC Neptune is a real possibility. We don't need that for our 55nm and shortly after our 28nm machines to succeed in the market but if we scooped up some of the KnC 28Million of orders we would be doing well. Even we don't get any of their 28Mill, their Neptunes have sold out and they sold out quickly. The market demand for any machine that will ROI is very very high.
|
|
|
|
minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 07:44:55 PM |
|
1 billion dollars of investment in mining equipment could easily be swallowed by the global electronic industry - it wouldn't even blink.
I don't think we are talking about the value of orders, we are talklng about foundry capacity.1) On the 28nm scale global capacity it will rise by no more than 350% in 2014 - fact. It doesn't matter how much money you have to make orders if the foundries will not take those orders. 2) KnC's order of 20nm tech has been on a strict limited batch for a reason - that most likely is foundry capacity. 3) The world runs on these foundries output - Bitcoin ASIC's are not the only chips fighting for space on these production lines.
|
|
|
|
zumzero
|
|
January 27, 2014, 07:45:45 PM |
|
You thought correct.
Ok thanks! Where can I find them? I tried sifting through the threads but didn't find anything except mentions of cryptostocks.com Hey wolfbiter, Edit: Our shares are not trading yet An account will be created for you on the trading platform Crypto-trade, not to be confused with Cryptostocks. This is expected to happen in only a matter of days but be sure to tune back in on Wednesday around 2345hrs GMT and look for Ken's 'Weekly Update' for more information. You will need access to the email address you used in Bitfunder (expect to be contacted through that) as well as the public address from Bitfunder that you would have given to receive dividends. To prove ownership of your shares you will need to be able to 'sign a message' from that public address when the time comes.
|
|
|
|
mainline
Newbie
Offline
Activity: 56
Merit: 0
|
|
January 27, 2014, 07:50:00 PM |
|
... So the point here is ACtM are the only mining industry company with uncertainty over the viability of their product...
If you mean to say "not the only," then you're right. Labcoin is having some issues... *you guys trading yet?
|
|
|
|
PurpleTentacle
|
|
January 27, 2014, 07:52:40 PM |
|
... A global combined investment of 1 billion dollars in 2014 would yield around 500,000,000 GH of mining power. Thats 500,000 PH compared to the current network hash rate of 17.5 PH ...
500,000,000 GH = 500,000 TH = 500 PH
|
|
|
|
MrTeal
Legendary
Offline
Activity: 1274
Merit: 1004
|
|
January 27, 2014, 07:55:32 PM |
|
Another worry about the Neptune is the power-draw. This guy thinks it will be over 4500Watts and so too load much for many domestic settings. One answer - to wire it into your electric cooker outlet would not be possible or acceptable for many. http://coinchomp.com/2014/01/06/the-knc-miner-neptune-should-consume-at-least-4500-watts/So the point here is ACtM are the only mining industry company with uncertainty over the viability of their product. I think a mass refund request on the KnC Neptune is a real possibility. We don't need that for our 55nm and shortly after our 28nm machines to succeed in the market but if we scooped up some of the KnC 28Million of orders we would be doing well. Even we don't get any of their 28Mill, their Neptunes have sold out and they sold out quickly. The market demand for any machine that will ROI is very very high. That guy has no idea what he's talking about, and is starting from completely incorrect assumptions. KnC recommends a 1200W PSU, that doesn't mean that the Jupiters use 1200W.
|
|
|
|
zumzero
|
|
January 27, 2014, 07:56:33 PM |
|
"Do you accept there is a difference between point zero, zero, two cents and point zero, zero, two dollars?" "No sir, I do not." ROFL
|
|
|
|
wolfbiter
Newbie
Offline
Activity: 11
Merit: 0
|
|
January 27, 2014, 07:57:38 PM |
|
You thought correct.
Ok thanks! Where can I find them? I tried sifting through the threads but didn't find anything except mentions of cryptostocks.com Hey wolfbiter, An account will be created for you on the trading platform Crypto-trade, not to be confused with Cryptostocks. This is expected to happen in only a matter of days but be sure to tune back in on Wednesday around 2345hrs GMT and look for Ken's 'Weekly Update' for more information. You will need access to the email address you used in Bitfunder (expect to be contacted through that) as well as the public address from Bitfunder that you would have given to receive dividends. To prove ownership of your shares you will need to be able to 'sign a message' from that public address when the time comes. Thanks a ton for the summary, and sorry for the seemingly stupid questions! It's nice to know that there are still helpful people despite all the FUD out there...
|
|
|
|
PurpleTentacle
|
|
January 27, 2014, 07:58:45 PM |
|
"Do you accept there is a difference between point zero, zero, two cents and point zero, zero, two dollars?" "No sir, I do not." ROFL 10 times worse than that, actually...
|
|
|
|
minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 08:10:06 PM Last edit: January 27, 2014, 08:30:01 PM by minerpart |
|
That guy has no idea what he's talking about, and is starting from completely incorrect assumptions. KnC recommends a 1200W PSU, that doesn't mean that the Jupiters use 1200W.
I don't know where you get that from. He didn't even mention power supply units in doing the math: ' The Neptune’s product details page did however cite a 30% reduction in watts per GHash. Jupiter’s power consumption was 2.18 Watts per GHash. Reducing this by 30% gives you 1.526 Watts per GHash. Multiplying this by the Neptune’s 3000 GHash (3Terrahash) gave me a power consumption of 4578 Watts. How I Got This Number: Jupiter’s Power Consumption (JPC) is 1200 Watts' EDIT - OK I see what he did above, but if they recomend 1200Watt PSU it must use atleast 80% of that? This means the Neptune will use 80% of his figure which is still close to 4000Watts of power? EDIT2 - looks like it's around 60% of his figure as a Jupiter draws about 700W. Still a draw of 2700Watts for the Neptune. That's like a three bar fire. The heat from this machine will be immense?
|
|
|
|
minerpart
Sr. Member
Offline
Activity: 280
Merit: 250
IIIIII====II====IIIIII
|
|
January 27, 2014, 09:11:31 PM |
|
A question for Vince for the weekly Q and A:
When will VMC start advertising and taking pre-orders for our 55nm-based mining machines and chips.
|
|
|
|
st4nl3y
Legendary
Offline
Activity: 1050
Merit: 1000
|
|
January 27, 2014, 09:29:39 PM |
|
Ken: "Anyone who thinks he's stupid may stand up!" *Nobody stands up* Ken: "Im sure there are some stupid investors over here!!" *investor x stands up* Ken: "Ohh, you think you're stupid?" investor x: "No... i just feel bad that you're standing alone..."
|
|
|
|
Bitdust
Newbie
Offline
Activity: 42
Merit: 0
|
|
January 27, 2014, 09:35:01 PM |
|
I think Ken should put up a bid wall at .005 before we go live to prevent panic sells. Anyone else think it's a good idea?
|
|
|
|
JoTheKhan
|
|
January 27, 2014, 09:36:53 PM |
|
I think Ken should put up a bid wall at .005 before we go live to prevent panic sells. Anyone else think it's a good idea?
I do not.
|
|
|
|
|