Total Hashrate Forecast Q1, Q2 2014 (Community work)

[crazydownloaded]

Is the front loading accidental or based on real hard data ?

I didn't understand this part BTW, sorry.

[1714871468]

1714871468

[1714871468]

1714871468

[1714871468]

1714871468

[1714871468]

1714871468

[1714871468]

1714871468

[BuildTheFuture]

I think he was asking why it seems so much new hash rate is hitting in March/April, then tapering off a bit.

[crazydownloaded]

I think he was asking why it seems so much new hash rate is hitting in March/April, then tapering off a bit.

Ok, then I think I already answered ^^ I simply put the estimated numbers in the columns I think the quantities will be shipped, not artificially smoothing them by looking at the totals

[NicholasB54]

KnC all delivered in March ?

[NicholasB54]

Is the front loading accidental or based on real hard data ?

I didn't understand this part BTW, sorry.

Just meant were you more concerned about the overall End of June figure and less concerned about the actual staged release of product - which you already answered :-)

I need to do more research before offering alternative dates so only able to query your dates at this point

Reading beyond the headline in this http://forum.kncminer.com/forum/off-topic/miners-cafe/28830-seems-like-good-news-for-neptune
seems to me it is unlikely to be KnC product they're making

Will post something more constructive when I've done more checking

[BuildTheFuture]

ASICMiner has provided a news update: https://bitcointalk.org/index.php?topic=438359

Sounds like they are maybe 2 or 3 weeks behind schedule, but otherwise on track for deploying a whole lot of cheap, high-efficiency chips. Since there is currently no other chip manufacturer announcing anything close to those J/GH, I assume ASICMiner will be supplying these heavily on into the summer.

[crazydownloaded]

Sounds like they are maybe 2 or 3 weeks behind schedule

- Which schedule? When do you think they will start shipping?
- They speak about an option for "immediate delivery" when "Chips will be available for ordering once the functionality and yield rate are confirmed with testing on actual chips.", seems pretty close, isn't?

[BuildTheFuture]

I'm just going by their previously announced tapeout date of around Jan 20th. Turned out to be more in early Feb they are saying now. So 2-3 weeks.

As for massive amounts of their hashrate hitting the network, those chips have to still be transported and someone has to have boards made to put them on. I'm not a manufacturing expert, but if the chips just start to be available in March, it might not be until April that you see any 3rd party boards for sale using ASICMiner chips. Now, ASICMiner says they will also be selling their own "devices", so maybe some of theirs will start to hit in late March? Just guessing.

[BuildTheFuture]

Black Arrow just pushed back first deliveries to at least May 1st: https://bitcointalk.org/index.php?topic=294197.msg4832034#msg4832034

[crazydownloaded]

Black Arrow just pushed back first deliveries to at least May 1st: https://bitcointalk.org/index.php?topic=294197.msg4832034#msg4832034

That's breaking news! I updated the spreadsheet with that info.

[kano]

I think he was asking why it seems so much new hash rate is hitting in March/April, then tapering off a bit.

Ok, then I think I already answered ^^ I simply put the estimated numbers in the columns I think the quantities will be shipped, not artificially smoothing them by looking at the totals

So you are expecting them to stop production once they deliver a batch?

(which I think might be the point of the original question?)

[crazydownloaded]

So you are expecting them to stop production once they deliver a batch? (which I think might be the point of the original question?)

Nope, but I can't/don't want to speculate on this. We don't know yet if they will deliver the full batch, continue producing it, release a new chip, etc.
Let me know what you'd like, in which cells, and we can discuss about this (possibly adding them with a different color, to differentiate announcements and speculation)

[MANofthePEOPLE]

Do you think we'll see 40-46% jumps like in october when cointerra et all ships? Or can we expect it being around 20-25% jumps from now on?

[BuildTheFuture]

It's possible to see some bigger jumps in March and April particularly I think. Did you look at crazy's spreadsheet? It shows a couple of spikes over 40%. But that is if all the vendors shipped exactly on time, and managed to get out those full quantities. It seems to me historically that most vendors are late and/or take longer to deliver the whole shipment. So I kinda lean towards the spreadsheet being a more conservative/worst case. But then you look at the current network hashrate it is jumping up, clearly there are significant sources of hashrate this spreadsheet isn't taking into account...

[World]

It looks like the beginning of April will be interesting.
http://www.cryptocoinsnews.com/wp-content/uploads/2014/02/log-rise-times.png

[BuildTheFuture]

I think some of the hashfast stuff is starting to ship earlier than in the spreadsheet. I saw one of their customer mention today they got theirs, and they are batch #2. On the other hand, the MPP stuff has not shipped and the company hasn't released any info on when it will.

[BuildTheFuture]

So today's new is KNC. As best I can tell:

1. The Neptune is running quite a bit late. They still haven't wrapped up their chip design. I'm guessing May or later for these?

2. They are beginning construction on their own cloudhashing type farm, using at first their older 28nm chips/Jupiters. The timing of when this will be ready is unknown. But it might very well begin coming online in Q2, as some people speculate they have already been producing/hoarding 28nm chips for it so they may be ready to go. So here's a source of hashpower not in the spreadsheet.

[de_ixie]

We have a lot of new information to crunch our numbers. I will sit down on weekend and publish an update latest sunday.

Thanks again for all the inputs

[baddw]

Thank you for providing the info in this thread.  I will be following it closely.

[BuildTheFuture]

This article has a little more info on the KNC farm:

http://www.datacenterknowledge.com/archives/2014/02/06/bitcoin-miners-building-10-megawatt-data-center-sweden/

So, 10 megawatts, using very rough numbers of 1000 watts per KNC Jupiter, then something like 600 GH/s per Jupiter, this is in the ballpark of 6 PH/s? Someone check my math.