Pages:
Author

Topic: [ANN] CoinLab Protected Pool - page 52. (Read 97867 times)

-ck
legendary
Activity: 4088
Merit: 1631
Ruu \o/
August 15, 2012, 04:55:47 PM
#42
Also, won't we need to raise our mem clocks on our gpus for this kind of work? It will raise our electricity costs at least 10-20% depending on the cards.

Would memory overclocking also be helpful ?
The work requirements at this stage are entirely unknown, but we're talking (potentially) about a world where mining bitcoin on GPUs is completely unprofitable.
hero member
Activity: 535
Merit: 500
August 15, 2012, 10:29:51 AM
#41
Also, won't we need to raise our mem clocks on our gpus for this kind of work? It will raise our electricity costs at least 10-20% depending on the cards.

Would memory overclocking also be helpful ?

legendary
Activity: 1027
Merit: 1005
August 14, 2012, 08:28:51 PM
#40

So is the rate at $5.10 for PPS the minimum you will make at a specific difficulty per 1Gh/s/day? But possibly more? How is 'more' calculated? how much is 'more'? I would assume it is based on how much a company is willing to pay for the service but I could be wrong.

$5.10 was the expected dollar value of 95%PPS a couple days ago (using price of $11.7 and difficulty 2190865.9701).  With that price and difficulty, you would earn at least $5.10 if we used your computer for scientific compute work.

We are still in the sales process so it is too early to discuss specifics: all we can promise right now is that you will never earn less than you would from mining Bitcoins when we use your computer for other compute work (1+ months away).  We will have a period of less-lucrative HPC testing before we can actually generate revenue from it, and this 'test time' will be coming out of our own pocket.  We may give a bonus to users who help us test, but we don't want to obligate ourselves to do so at this time.

Once we are making money, we plan to pass some of this revenue down to the miners.  But seeing as we're still learning how big the pie will be, we can't say how we'll slice it yet. It will be "more" though  Wink.

PS: Looks like you quoted the wrong post.

That makes sense. Thanks for the info. And yeah it looks like I quoted the wrong post, I really do not like the way this forum works...
hero member
Activity: 504
Merit: 502
August 14, 2012, 07:13:20 PM
#39
Depending on the work and price, it might be time to boot up some teslas over at amazon for these jobs even if it gets ahead with just a couple of cents or perhaps $1 ;0
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 14, 2012, 07:07:13 PM
#38

So is the rate at $5.10 for PPS the minimum you will make at a specific difficulty per 1Gh/s/day? But possibly more? How is 'more' calculated? how much is 'more'? I would assume it is based on how much a company is willing to pay for the service but I could be wrong.

$5.10 was the expected dollar value of 95%PPS a couple days ago (using price of $11.7 and difficulty 2190865.9701).  With that price and difficulty, you would earn at least $5.10 if we used your computer for scientific compute work.

We are still in the sales process so it is too early to discuss specifics: all we can promise right now is that you will never earn less than you would from mining Bitcoins when we use your computer for other compute work (1+ months away).  We will have a period of less-lucrative HPC testing before we can actually generate revenue from it, and this 'test time' will be coming out of our own pocket.  We may give a bonus to users who help us test, but we don't want to obligate ourselves to do so at this time.

Once we are making money, we plan to pass some of this revenue down to the miners.  But seeing as we're still learning how big the pie will be, we can't say how we'll slice it yet. It will be "more" though  Wink.

PS: Looks like you quoted the wrong post.
legendary
Activity: 1027
Merit: 1005
August 14, 2012, 06:16:32 PM
#37
If you earn more than $600 worth of BTC in one calendar year and are in the US, we may need to send you a form 1099 to report our payments to you to the IRS (we will need your name, address, and US tax id). Miners in this group also get the option of  receiving US dollar payouts, or a mixture of USD and BTC (you can take your electricity costs in USD and keep your profits in BTC).

Does that mean this is only available to those in the U.S.?   


We’re evaluating a means of waiving this requirement, so we’ll keep you posted.

Waiving that income would be reported or waiving the requirement for a U.S. tax id?

Our pool is available internationally, excluding countries and persons on the OFAC list.  You can find these lists here:
http://www.treasury.gov/resource-center/sanctions/SDN-List/Pages/default.aspx

We're evaluating a means of waiving the reporting requirement altogether.  We're still researching but it appears we may not have to 1099 foreigners.

So is the rate at $5.10 for PPS the minimum you will make at a specific difficulty per 1Gh/s/day? But possibly more? How is 'more' calculated? how much is 'more'? I would assume it is based on how much a company is willing to pay for the service but I could be wrong.
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 14, 2012, 06:06:33 PM
#36
If you earn more than $600 worth of BTC in one calendar year and are in the US, we may need to send you a form 1099 to report our payments to you to the IRS (we will need your name, address, and US tax id). Miners in this group also get the option of  receiving US dollar payouts, or a mixture of USD and BTC (you can take your electricity costs in USD and keep your profits in BTC).

Does that mean this is only available to those in the U.S.?   


We’re evaluating a means of waiving this requirement, so we’ll keep you posted.

Waiving that income would be reported or waiving the requirement for a U.S. tax id?

Our pool is available internationally, excluding countries and persons on the OFAC list.  You can find these lists here:
http://www.treasury.gov/resource-center/sanctions/SDN-List/Pages/default.aspx

We're evaluating a means of waiving the reporting requirement altogether.  We're still researching but it appears we may not have to 1099 foreigners.
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 14, 2012, 05:46:48 PM
#35
Quick question:

How will you handle the sorts of problems in scientific computing that favour different sorts of GPUs?

For example, I do a lot of work with various CFD flow solvers - there are some out that use CUDA, and some of the newer ones support OpenCL (an exciting time!) - the issue is that whilst a high end Nvidia card will be rubbish for BTC mining, often they'll perform better than the equivalent ATI card for some given task.

Therefore it's questionable how much of a correlation there will be between bitcoin hashrate and the rate at which these cards will perform tasks in scientific computing. I don't honestly think you'll be able to offer a fixed $/Gh rate, as the measure of Gigahashes/second will almost certainly be irrelevant, unless you're trying to crack passwords  Wink

We plan to benchmark work units across many different hardware setups to determine which cards (or other factors) solve which of our work units most efficiently.  We'll prioritize the most efficient cards for different jobs. Then, there are also demand side factors, where some HPC customers may be willing to pay for our whole cluster's capacity for shorter bursts (or all CUDA).

We'll always pay out at least the expected earnings your particular hardware would have earned Bitcoin mining for the same period of time, if not more. 

There is currently a much greater demand for NVIDIAs than AMDs for HPC, so NVIDIAs will probably be the first cards that are able to earn a rate greater than Bitcoin mining.  Watch for an NVIDIA-only mining pool offer from CoinLab in the coming weeks.
sr. member
Activity: 336
Merit: 250
August 14, 2012, 05:24:26 PM
#34
Quick question:

How will you handle the sorts of problems in scientific computing that favour different sorts of GPUs?

For example, I do a lot of work with various CFD flow solvers - there are some out that use CUDA, and some of the newer ones support OpenCL (an exciting time!) - the issue is that whilst a high end Nvidia card will be rubbish for BTC mining, often they'll perform better than the equivalent ATI card for some given task.

Therefore it's questionable how much of a correlation there will be between bitcoin hashrate and the rate at which these cards will perform tasks in scientific computing. I don't honestly think you'll be able to offer a fixed $/Gh rate, as the measure of Gigahashes/second will almost certainly be irrelevant, unless you're trying to crack passwords  Wink
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 14, 2012, 04:55:18 PM
#33
Quote
The current expected earnings from a 1GH/s for 24 hours is $5.37 (using price of $11.7 and difficulty 2190865.9701).

Based on my calculations, $5.37 is what I can mine with 1Gh/s in 24 hours at current difficulty and $11.7 price. is that what your saying here or is that what we could earn with the 95%? 95% of $5.37 is $5.10. Just trying to understand the 95% PPS here.

You are correct, $5.37 is what you earn at 100%. $5.10 is what you will earn on our pool at 95%PPS.  Thanks, updated OP.
legendary
Activity: 1027
Merit: 1005
August 14, 2012, 03:15:20 PM
#32
Quote
The current expected earnings from a 1GH/s for 24 hours is $5.37 (using price of $11.7 and difficulty 2190865.9701).

Based on my calculations, $5.37 is what I can mine with 1Gh/s in 24 hours at current difficulty and $11.7 price. is that what your saying here or is that what we could earn with the 95%? 95% of $5.37 is $5.10. Just trying to understand the 95% PPS here.
legendary
Activity: 1330
Merit: 1026
Mining since 2010 & Hosting since 2012
August 14, 2012, 01:01:01 PM
#31
I really hope this isnt the best things to come from coinlab using their investor money.




This is real forward thinking on their part.  I applaud them.
hero member
Activity: 535
Merit: 500
August 13, 2012, 11:37:12 PM
#30
Well someone has to be able to make use of 10-12 TH of GPU's run by hardware enthusiasts who have been wrestling with them for months to years.

I would love to grow my farm and keep playing with GPU's especially some nvidia ones(better drivers), but to ask us to assume your risk in exchange for future earnings is tough when right now my farm nets over $2500 per month and rising.

I love the future hedge, but what happens if you are out of business ? Bitcoin miners have already had to assume great risk in building farms in the face of uncertainty. I think if you guys can simplify and reduce our risk, you would have 4-5 TH on no time.

I have 16-17 GH right now and could double that in a month without issue. I would go to 100 GH within 6 months if I had the right situation(steady guaranteed income)

I'm okay with 95% PPS as long as we get guaranteed payout for the 5% if you fail to deliver down the road.

I would also be okay if you gave us miners an equity position based on supporting your project early with our farms. Something like a profit-sharing situation down the road which includes a bonus for joining and supporting this early.

Anyway, simplify and I'm probably in.
legendary
Activity: 2506
Merit: 1010
August 13, 2012, 11:25:18 PM
#29
If you earn more than $600 worth of BTC in one calendar year and are in the US, we may need to send you a form 1099 to report our payments to you to the IRS (we will need your name, address, and US tax id). Miners in this group also get the option of  receiving US dollar payouts, or a mixture of USD and BTC (you can take your electricity costs in USD and keep your profits in BTC).

Does that mean this is only available to those in the U.S.?   


We’re evaluating a means of waiving this requirement, so we’ll keep you posted.

Waiving that income would be reported or waiving the requirement for a U.S. tax id?
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 13, 2012, 12:55:40 PM
#28
Announcement: We've extended the expiration time on Loyalty Points to 2 years after the price drop. 
legendary
Activity: 2576
Merit: 1186
August 13, 2012, 10:43:50 AM
#27
Note that there's an average of about 5% orphan blocks, so anything over 95% is really unsustainable in a fair system.
full member
Activity: 141
Merit: 100
August 13, 2012, 10:25:00 AM
#26

In our HPC sales process, we are often finding that the first question we are asked is, "How big is your cluster?".  We want our pool to grow fast so we can get the most lucrative compute jobs online as soon as possible.  Because HPC pricing is time sensitive, 100 miners mining for an hour is worth more than 1 miner mining for 100 hours.


OK I see how having a big cluster is good for marketing.  But the weird terms of use severely limit your potential pool size.

You aren't going to find a more advanced user base of early adapters than here on BTCTalk.  So what does is mean when when the whole "future earnings insurance" idea goes over like a lead balloon, confusing some very smart people and even generating accusations of OMGPonziScam?  It means keep it simple sweetheart.  

If you want to get everyone to consider using your pool, you must offer a more competitive rate for the same quality of service as OzCoin, DeepBit, or BTCGuild.

Look at how well CoinSauce did by initially offering 100% PPS.  All you need to do is beat the market rate of 95% and viola, instant user base of up to 200Petaflops!

Don't try to make your users share the burden of future risk, it's now obvious that they are not interested.  Bitcoins talk, BS vaporware walks.  /tough love

Focus on your core competency of linking *coin miners' supply of processing power to the HPC market's demand.  The drama and risks of running a pool dilute and threaten the CoinLab brand.  

Have you thought about partnering with existing pools instead of running your own?  The tax implications alone are sufficient reason to deal with a small number of pool operators instead of a large number of pool users.

Hey iCEBREAKER, thanks for these thoughts. I'm sold; we'll work with our team to launch a 100% PPS pool as well. It's possible we're just thinking farther ahead than most, but I would be stunned to find that $/GH don't drop by a factor of up to 10x in the next year, so I think we're offering real value to miners right now with this offer.

That said, it's complicated, I grant. I love simple. We'll work on it.

legendary
Activity: 2156
Merit: 1072
Crypto is the separation of Power and State.
August 12, 2012, 05:11:20 PM
#25
Very good idea for a unique PPS pool/monetized distributed computation hybrid.

As regular reader of gpgpu.org, I've been waiting for something like to come along.  Can't bother doing it myself, so cheers to the LazyNet.

Bitcoin establishing a price floor for the market is an incredible catalyst for suchlike.

You're going to need lots of $$$ to be competitive in this space, doing Big Science and whatnot.  Are you talking to VCs yet?

I suggest starting by supporting a couple of well chosen proprietary products, such as Schrödinger Suite 2012.

http://www.schrodinger.com/news/63

Run it cheaper than EC2 and profit!

There's also plenty of open source platforms to host, like GROMACS and BLAST.

CloudBroker was moving towards creating an open market for this stuff, but seems to be down ATM.

http://webcache.googleusercontent.com/search?q=cache:XLHUSI0BwDUJ:www.cloudbroker.com/&hl=en&prmd=imvns&strip=1

Be careful.  I've seen many companies enter, fail, and leave this space.  But since Bitcoin, conditions seem ripe for revolution...


Thanks for the support. Great leads too.  I'm passing your post along to Biz Dev. If you (or anyone else) has suggestions of companies that you think we should contact, we'd love to hear your input.

I've been monitoring the emerging nexus of distributed computing and online markets since before buzzwords like crowdsourcing and cloud computing existed.

10 years ago, I made my first bit of internet money by running Porivo Peer for a few days.  Sure, it only earned a nickle ($.05USD) but it was proof of concept.

So I could go on about this kind of stuff for days.  Send a PM if you want a consultant with tons of ideas and a deep background in this space.  Plus I know some very key people.

We could be the world's first trillionaires by playing our cards right...  Wink

legendary
Activity: 2156
Merit: 1072
Crypto is the separation of Power and State.
August 12, 2012, 04:46:51 PM
#24

In our HPC sales process, we are often finding that the first question we are asked is, "How big is your cluster?".  We want our pool to grow fast so we can get the most lucrative compute jobs online as soon as possible.  Because HPC pricing is time sensitive, 100 miners mining for an hour is worth more than 1 miner mining for 100 hours.


OK I see how having a big cluster is good for marketing.  But the weird terms of use severely limit your potential pool size.

You aren't going to find a more advanced user base of early adapters than here on BTCTalk.  So what does is mean when when the whole "future earnings insurance" idea goes over like a lead balloon, confusing some very smart people and even generating accusations of OMGPonziScam?  It means keep it simple sweetheart.  

If you want to get everyone to consider using your pool, you must offer a more competitive rate for the same quality of service as OzCoin, DeepBit, or BTCGuild.

Look at how well CoinSauce did by initially offering 100% PPS.  All you need to do is beat the market rate of 95% and viola, instant user base of up to 200Petaflops!

Don't try to make your users share the burden of future risk, it's now obvious that they are not interested.  Bitcoins talk, BS vaporware walks.  /tough love

Focus on your core competency of linking *coin miners' supply of processing power to the HPC market's demand.  The drama and risks of running a pool dilute and threaten the CoinLab brand.  

Have you thought about partnering with existing pools instead of running your own?  The tax implications alone are sufficient reason to deal with a small number of pool operators instead of a large number of pool users.
sr. member
Activity: 270
Merit: 250
1CoinLabF5Avpp5kor41ngn7prTFMMHFVc
August 10, 2012, 03:12:25 PM
#23
Very good idea for a unique PPS pool/monetized distributed computation hybrid.

As regular reader of gpgpu.org, I've been waiting for something like to come along.  Can't bother doing it myself, so cheers to the LazyNet.

Bitcoin establishing a price floor for the market is an incredible catalyst for suchlike.

You're going to need lots of $$$ to be competitive in this space, doing Big Science and whatnot.  Are you talking to VCs yet?

I suggest starting by supporting a couple of well chosen proprietary products, such as Schrödinger Suite 2012.

http://www.schrodinger.com/news/63

Run it cheaper than EC2 and profit!

There's also plenty of open source platforms to host, like GROMACS and BLAST.

CloudBroker was moving towards creating an open market for this stuff, but seems to be down ATM.

http://webcache.googleusercontent.com/search?q=cache:XLHUSI0BwDUJ:www.cloudbroker.com/&hl=en&prmd=imvns&strip=1

Be careful.  I've seen many companies enter, fail, and leave this space.  But since Bitcoin, conditions seem ripe for revolution...


Thanks for the support. Great leads too.  I'm passing your post along to Biz Dev. If you (or anyone else) has suggestions of companies that you think we should contact, we'd love to hear your input.
Pages:
Jump to: