CoinTerra announces its first ASIC - Hash-Rate greater than 500 GH/s

[VolanicEruptor]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

[plato14]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

Nothing we can do really, just wait on the next update, cointerra like all other pre-order mining companies has us by the balls

[testerx]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

I don't think there were as many orders with cointerra as some of the other manufacturers. They've only sold about 4000 units so far and a lot probably went to the same people. 

[de_ixie]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

I don't think there were as many orders with cointerra as some of the other manufacturers. They've only sold about 4000 units so far and a lot probably went to the same people. 

Hey, may I ask you from where you got that information -> 4000 units? Any source?

[JoseSan]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

I don't think there were as many orders with cointerra as some of the other manufacturers. They've only sold about 4000 units so far and a lot probably went to the same people. 

Hey, may I ask you from where you got that information -> 4000 units? Any source?

http://www.coindesk.com/cointerra-cuts-price-of-terraminer-iv-bitcoin-mining-rig/
This article from coindesk mentioned the following:

CoinTerra claims it intends to increase the current bitcoin network power by 2 Petahashes per second, which is over four times the current capacity of the whole bitcoin network (487.50 TH/s according to Bitcoin Charts).

Going by that number means they sold 1000 2TH December units. This also roughly correlates to the order numbers I saw when they said they had sold out (they roughly stopped at an enumeration of ~1000). You can estimate that every month they sold another 1000 units, thus December, January, February, March and April orders would amount to 5000 units.

[testerx]

why is it so quiet in here?  You would think there would be a lot of action in these threads.
Is everyone still alive?

I don't think there were as many orders with cointerra as some of the other manufacturers. They've only sold about 4000 units so far and a lot probably went to the same people.  

Hey, may I ask you from where you got that information -> 4000 units? Any source?

You can make estimates based on info they gave in the past as well as in recent articles.  The December batch was much smaller than the other batches and the other batches run approximately 1000 units each and they've sold out four.  The best indicator that my estimate is correct are their revenue figures.  The source is my own research, I would suggest you do your own as should anybody else committing money

[VolanicEruptor]

A Friday update would be nice..
How's the testing going??

[Easy2Mine]

No new update from Cointerra?

[lucazane]

They are talking about 132 GH per die (http://cointerra.com/engineering-update-asic-bringup-diagnostic-testing/)
And here we can see each ASIC has 4 dies (http://cointerra.com/engineering-update-goldstrike-packaged-chips-system-board-assembly/)

Can we expect to have 528 GH per ASIC and so 2112 GH per TerraMiner ? What are you thinking ?

[plato14]

Mine @ p2pool

I think that the cointerra customers are going to put a lot of hashing power on the network here in the next few months. It would be tragic if we were to put most of that hashing power on the big pools i.e. ghash.io, btcguild, or elegius. If we brought all of this hashing power to p2pool we that would be awesome and would help keep the network decentralized. There are no fees and the pool is hardly ever down. Its really not that hard to setup and setting up a node is really not that bad if you have a little extra room on your hard drive for the client, and if you don't you can always use someone else's node. Here is a link that makes it really easy to setup, i also hear that some people are working on a GUI to make it even easier to use.

http://p2pool.in

Edit :  http://www.reddit.com/r/Bitcoin/comments/1uz4g0/p2pgui_a_gui_for_mining_using_p2poolorg/

Edit:  http://p2pool.org

[ImI]

Mine @ p2pool

I think that the cointerra customers are going to put a lot of hashing power on the network here in the next few months. It would be tragic if we were to put most of that hashing power on the big pools i.e. ghash.io, btcguild, or elegius. If we brought all of this hashing power to p2pool we that would be awesome and would help keep the network decentralized. There are no fees and the pool is hardly ever down. Its really not that hard to setup and setting up a node is really not that bad if you have a little extra room on your hard drive for the client, and if you don't you can always use someone else's node. Here is a link that makes it really easy to setup, i also hear that some people are working on a GUI to make it even easier to use.

http://p2pool.in

i like p2p and think its longterm the way to go

but, with knc jupiters there were massive performance declines on p2p and alot of miners switched.

i hope that those problems will be sorted out sometime.

[aerobatic]

They are talking about 132 GH per die (http://cointerra.com/engineering-update-asic-bringup-diagnostic-testing/)
And here we can see each ASIC has 4 dies (http://cointerra.com/engineering-update-goldstrike-packaged-chips-system-board-assembly/)

Can we expect to have 528 GH per ASIC and so 2112 GH per TerraMiner ? What are you thinking ?

while that does look possible in theory.. itd be an unsafe assumption to simply take the announced single die performance and multiply by four to conclude that when all four dies are operating that you will get a simple multiplying of the result...

...  as we dont know that when all four dies are running at the same time, whether the chip will hit the max power or thermal limits of the package (or the cooling system, or the dc/dc converters... or the power supply.. etc).   theres plenty of parts of the system that could place limits on its overall performance so lets wait til they tell us what it can do... as its been designed to operate at faster speeds and presumably more watts (and thus hotter) than any chip we've seen so far..

.. and since they said it was gonna be 500 GH per chip, i think it would be optimistic to speculate it goes any faster than that (... unless they tell us that)

[matthewh3]

Mine @ p2pool

I think that the cointerra customers are going to put a lot of hashing power on the network here in the next few months. It would be tragic if we were to put most of that hashing power on the big pools i.e. ghash.io, btcguild, or elegius. If we brought all of this hashing power to p2pool we that would be awesome and would help keep the network decentralized. There are no fees and the pool is hardly ever down. Its really not that hard to setup and setting up a node is really not that bad if you have a little extra room on your hard drive for the client, and if you don't you can always use someone else's node. Here is a link that makes it really easy to setup, i also hear that some people are working on a GUI to make it even easier to use.

http://p2pool.in

i like p2p and think its longterm the way to go

but, with knc jupiters there were massive performance declines on p2p and alot of miners switched.

i hope that those problems will be sorted out sometime.

KNC Jupiter's have been working well on P2Pool for a while now.  Although yes CT should work with P2Pool dev's to make sure their machines work well with the pool from the start.  Also I can recommend the pool Bitparking.  Which is a very small pool (to help decentralisation) where you can also merge-mine up to five other alt-coins.  Merge-ming alt-coins is basically money for nothing and if BTC went to >$10K.  Then who know's where the alt-coin prices would go.  With a rising tide lifting all boats IMHO.

[VolanicEruptor]

If Cointerra decides not to sell their soul to devil and deliberately take a month to test our chips, I think they will be KNC's ultimate competitor for the coming years.   6 months down the road they will be the only 2 left standing I think, like microsoft vs apple.  This is if Cointerra actually delivers in a reasonable time frame..

[Tegija]

Mine @ p2pool

I think that the cointerra customers are going to put a lot of hashing power on the network here in the next few months. It would be tragic if we were to put most of that hashing power on the big pools i.e. ghash.io, btcguild, or elegius. If we brought all of this hashing power to p2pool we that would be awesome and would help keep the network decentralized. There are no fees and the pool is hardly ever down. Its really not that hard to setup and setting up a node is really not that bad if you have a little extra room on your hard drive for the client, and if you don't you can always use someone else's node. Here is a link that makes it really easy to setup, i also hear that some people are working on a GUI to make it even easier to use.

http://p2pool.in

Edit :  http://www.reddit.com/r/Bitcoin/comments/1uz4g0/p2pgui_a_gui_for_mining_using_p2poolorg/

Edit:  http://p2pool.org

in some points you are deffinently right: it is very important, that not all the hashing power goes to big pools and it is time that small and good pools start to grow.
in some points i have to say that your given informations are not really correct or they are correct only in some points.
1. please notice that p2pool.info is the mining pool and p2pool.org not. p2pool.org is practically only one node of p2pool. same as elizium.name or any other node.
2. yeah! you are right, p2pool is hardly down! but... many nodes are very unstabile!!! it happens pretty often, that even bigger nodes are down (like elizium.namen in last months). if miners want to have stability, they should have own node or they have to try to find node witch they really can trust. second problem is, that sometimes nodes make mistakes and lots of hashing power just gets lost somehow... (i had 60GH running nonstop in p2pool and did not get anything. i was in holiday and could not change anything.) and if it gets lost, you are not able to ask from anybody why it disappeared...
3. Making node is easy? yeah, right!!! did you greate one yourself? it´s really easy to greate the node, but if you want that it works stabile and good and if you think that there are some more people who would like to use your node, then it is not really such easy. all parameters have to be good to be able to mine really effective. most nodes have latency (and that is only one parameter out auf many) of 0,5 and that means that effective mining is actually not possible... at least not when it´s not your own node.
4. its not such important in this case, but don't forget, that most of new miner hardware  don't need any computer, so if you want to make your own node, then you  will have one electricity-eating-noise-and-heat-producing machine more to run.

Not everything is such easy and effective as it looks in the first moment...

I like p2pool.info very much, but i see those big problems and not many people are trying to solve them. i would not have anything against one node who takes some kind of low fee (0,2-1,0%) and puts up node witch is stabile and user can get some kind of support. so if user has a problem, they can get at least some kind of information about existing problem.

to get people to p2pool several things have to change. i have ordered 4 terra miner and when they get shipped i would like to mine in p2pool, but so as it is now i am not sure i will do it. making own node, for that i don't have enough skills/patience and as i said already: enough computer/miner running already anyway, no need for one more.

regs,
Tegija

[Carlton Banks]

p2pool does not go down.

Individual nodes do. But that's nothing to do with p2pool, it's attributable to loss of internet access local to the node.


ckolivas has confirmed that all the hardware and driver design for the Terraminer IV does/will conform to what is required for efficient operation on p2pool (long polling, stratum). That's not to say that there won't be possible improvements once the real hardware is available, but it appears that everything is being done to make the Cointerra units work on p2pool from the outset.

[Tegija]

p2pool does not go down.

Individual nodes do. But that's nothing to do with p2pool, it's attributable to loss of internet access local to the node.


ckolivas has confirmed that all the hardware and driver design for the Terraminer IV does/will conform to what is required for efficient operation on p2pool (long polling, stratum). That's not to say that there won't be possible improvements once the real hardware is available, but it appears that everything is being done to make the Cointerra units work on p2pool from the outset.

that is actually what i said, exept infos from ckolivas.
individual/public nodes are pretty unstabile. of course its not a problem of p2pool, more likely problem of nodes and their setup (hardware, software etc.)

driver design of terra miner can be conform of p2pool and in general i have no doubts that terraminer can run efficient in p2pool, only problem is that the miner hast to be connected to one node and is not able to have node in itself. that means, that with terraminer you will need one stabile node in p2pool network. using public nodes is now not really a good alternative, because most of nodes are pretty unstabile. if you would have some miners like BFL (plugged in one computer witch can function as node) it would not be a big problem to run your terraminer in this node to. still you would eventually have problems with latency (if it is not good enough) because terraminers are not directly connected to your node (in this case BFL miners would be directly connected and latency would be almost unimportant), so you have to "push-up" your node: memory, stabilization of internet access, etc).
i am sure that every terraminer can run efficient in p2pool and they already did everything that it is possible, but they can't make node for miners, that we have to make ourselves.

[Carlton Banks]

i am sure that every terraminer can run efficient in p2pool and they already did everything that it is possible, but they can't make node for miners, that we have to make ourselves.

That's right, you need bitcoind and p2pool clients running on the node, this will almost certainly be too demanding a burden on the controlling device running cgminer.

What controller is Cointerra using, do we know the answer yet? Anyone?

[JoseSan]

i am sure that every terraminer can run efficient in p2pool and they already did everything that it is possible, but they can't make node for miners, that we have to make ourselves.

That's right, you need bitcoind and p2pool clients running on the node, this will almost certainly be too demanding a burden on the controlling device running cgminer.

What controller is Cointerra using, do we know the answer yet? Anyone?

Cointerra is using a BeagleBone Black controller board, same as KNC, and using a Texas Instruments TM4C1233 80MHz 32-bit microcontroller for intermediate firmware (not sure what for, but it can be seen in the board photos).

[Ytterbium]

There are limits on how many units can be made per unit of time

Theoretically, yes.  But practically, we are so incredibly far removed from that, it doesnt really matter. Where exactly do you see a long term bottleneck?
If you think its assembly; globally we are currently producing ~50M less PC's than last year. Imagine even if only that (assumed idle) assembly capacity is brought to bitcoin miners, then the industry could assemble ~10 EH worth for Cointerra rigs per month based on idle capacity alone.

In that case 10EH/month would be the bottleneck. If the bitcoin difficulty were to grow at the rate it's been growing, we'd have to surpass 10EH/mo in less then 10 months.  Sooner then the end of 2014. 

And of course, people need to make the investments needed to make these things happen.  The faster the growth, the larger the upfront investment will need to be.