Cryptocoin Mining?

[slashbinslashbash]

Has anybody gotten set up with Counterparty and/or Mastercoin? Both seem like a PITA. I've bought some but just left them on exchanges so far, despite my extreme misgivings about doing so. The clients both require a full Bitcoin client running alongside, and neither of them have Mac clients, and I don't really care to download the full Bitcoin blockchain into my Parallels installation.

Just wondering if there are any particular "gotchas" or if the process is relatively smooth if you follow their instructions.

 

[thilanliyan]

None of the above. ScryptGuild.

Does Scryptguild automatically sell coins and pay out in BTC?

 

[taserbro]

Does Scryptguild automatically sell coins and pay out in BTC?

Yes.
It also lets you set the payout threshold to anything you want with no further penalty. Their fee is taken when you let the system auto-convert for you and you can set it to not convert one particular type of coin and pay it out to your wallet if you want.

 

[thilanliyan]

Yes.
It also lets you set the payout threshold to anything you want with no further penalty. Their fee is taken when you let the system auto-convert for you and you can set it to not convert one particular type of coin and pay it out to your wallet if you want.

You've been using it?

What has been your typical BTC/MH/s?

 

[njdevilsfan87]

For those interested in trying something new, BlackCoin's multipool is going to have a beta test this weekend before going live. It's not a traditional multipool that pays in BTC. It pays in BC instead by first converting to BTC and then using the BTC to purchase BC off the exchanges. Now as to why this is so interesting, imo, is because BC is now a PoS only coin generating 1% per year. It's practically a fixed supply. This multipool brings back the option to "mine BC" so that it can be distributed among the masses much more easily, but does so without generating any new coins. You are going to be mining coins directly off the exchanges instead. This means that the first time ever, your GPU hash power (or the entire community for that manner) can truly support a position or investment you hold within the crypto market. There's no "mine and dump" that have had coins like DOGE hit hard before. "Mine and dump" would cancel itself out completely in the case of BC. The pool itself becomes buy support. And if BC becomes heavily desired, and more miners start jumping in trying to "mine BC", it's going to be a self pumping machine that can get to who knows where.

bcmultipool.com

And yes I am now holding some BC. I'm in with hardware paid off I don't need to mine and sell anything for quite some time.

 

[Binky]

Screw multipools. I just want maxcoin to stay alive! It was profitable enough until a few days ago. Now the difference is too much to bother mining it (but difficulty is falling fast). It so much more pleasant to mine it too - Scrypt is so much hotter and louder to mine. Hell, you can even play games pretty well while mining maxcoin.

 

[taserbro]

You've been using it?

What has been your typical BTC/MH/s?

Not for long enough to have any relevant numbers for anyone to make a decision on; it does feel significantly better than the other multipools in my usage case though.

I typically shut off my rig between 1pm to 7pm on weekdays to adhere to the power company's metering schedule so 24 hour figures would be kinda funky to compute since the returns aren't the same for all times of the day and I might be off from people who let their rigs run 24/7. That said, I still get an approximate 0.02 to 0.029btc with my combined 2mh/s +/- 200kh/s every 24hours of uninterrupted uptime.

YMMV

 

[thilanliyan]

Not for long enough to have any relevant numbers for anyone to make a decision on; it does feel significantly better than the other multipools in my usage case though.

I typically shut off my rig between 1pm to 7pm on weekdays to adhere to the power company's metering schedule so 24 hour figures would be kinda funky to compute since the returns aren't the same for all times of the day and I might be off from people who let their rigs run 24/7. That said, I still get an approximate 0.02 to 0.029btc with my combined 2mh/s +/- 200kh/s every 24hours of uninterrupted uptime.

YMMV

Thanks for the info. I'll give Clevermining 2 days then try Scryptguild.

 

[Attic]

Not for long enough to have any relevant numbers for anyone to make a decision on; it does feel significantly better than the other multipools in my usage case though.

I typically shut off my rig between 1pm to 7pm on weekdays to adhere to the power company's metering schedule so 24 hour figures would be kinda funky to compute since the returns aren't the same for all times of the day and I might be off from people who let their rigs run 24/7. That said, I still get an approximate 0.02 to 0.029btc with my combined 2mh/s +/- 200kh/s every 24hours of uninterrupted uptime.

YMMV

I'm seeing ~0.006btc/mh per 24hr period. Most the profit pools are pretty close these days. Waffle has lowest rejects, scryptguild is pretty bad for rejects/stales unless you run a low intensity with --gpu-threads 2. So 7970/280x users are ok, but otherwise the cards you run at 19/20 intesity are going to do 5-8% rejects.

I get the sense that massive hashpower is continuing to come online. Was hoping scrypt N would seperate itself from asic doomed scrypt by this time.

If crypto gets hit with another wave consisting of 50% drop in BTC value and 50% drop in rate of ability to miner btc per MH then things will get intersting. It's still pretty profitable if you get a 24/7 ~5mh+ setup going.

 

[IEC]

I've got the hash power in my sig (7.5 MH/s) and on a good day these days I get around 0.05 BTC. After power costs I'm looking at between $25-$30 a day which is profitable, but just enough to make it worth it (as long as all I do is just monitor scryptguild). Not enough profit to make it worth chasing new coins any more.

If I didn't have all my hardware paid off by now I'd probably have sold hardware a few weeks back.

eBay is already being flooded with $500 290X and $400 290s.

 

[njdevilsfan87]

I'm just going to eat electricity costs if I have to. I don't plan to sell any GPUs. If I had done that last time, I would have made about 50% more than I did when that bubble occurred.

Screw multipools. I just want maxcoin to stay alive! It was profitable enough until a few days ago. Now the difference is too much to bother mining it (but difficulty is falling fast). It so much more pleasant to mine it too - Scrypt is so much hotter and louder to mine. Hell, you can even play games pretty well while mining maxcoin.

You misunderstand this multipool. Read a little more closely and then you'll see it's not about maximizing BTC profit. (at least not in the short-term)

(but I do agree I did like SHA3 mining better than Scrypt)

 

[Mark R]

I'm starting to downsize my mining operation.

I've got 5 MH/s which generates about 0.035 BTC but with electricity costs @ about $13 per day that's only like $7 profit.

Not worth it, given the heat, noise and space. My half-dead 7950s are fetching prices close to 3 months profit on ebay, so it's a no brainer to sell them.

(heck, even the completely dead ones - the one listed on ebay as "broken PCB, suspect open circuit RAM data line - corrupted display, unable to boot to desktop. Also has seized fan." fetched nearly $50 LOL)

The rate of new scrypt hash power has been quite phenomenal. When I started scrypt mining in November, there was about 40 GH/s of hashpower split between LTC and clones. Now, there is about 300 GH/s. I suspect a lot of that new hash power is in places with cheaper power, and more space than me.

 

[Dr. Zaus]

I'm starting to downsize my mining operation.

I've got 5 MH/s which generates about 0.035 BTC but with electricity costs @ about $13 per day that's only like $7 profit.

Not worth it, given the heat, noise and space. My half-dead 7950s are fetching prices close to 3 months profit on ebay, so it's a no brainer to sell them.

(heck, even the completely dead ones - the one listed on ebay as "broken PCB, suspect open circuit RAM data line - corrupted display, unable to boot to desktop. Also has seized fan." fetched nearly $50 LOL)

The rate of new scrypt hash power has been quite phenomenal. When I started scrypt mining in November, there was about 40 GH/s of hashpower split between LTC and clones. Now, there is about 300 GH/s. I suspect a lot of that new hash power is in places with cheaper power, and more space than me.

Dude in his private office is going to out pace you every time.

 

[KompuKare]

The rate of new scrypt hash power has been quite phenomenal. When I started scrypt mining in November, there was about 40 GH/s of hashpower split between LTC and clones. Now, there is about 300 GH/s. I suspect a lot of that new hash power is in places with cheaper power, and more space than me.

300GH/s is 300,000,000KH/s which divided by 700KH/s would be the equivalent of 428,000 R9-280X cards. Whereas 40GH/s was only 57,000 R9-280X equivalents. So that's an increase in hashpower equal to 370,000 R9-280X cards.

So, are there any secret scrypt-ASICs deployed in quantity somewhere or is most of that hashpower deployed GPUs? If it's only GPUs, then that's an awful lot. Obviously, they would be spread out amongst various types from 7790s to 290X plus a few 780Ti owners mining with cudaminer.

The other thing: if we assume each 280X draws around 200W, those 430,000 draw over 85MW! What a waste of power!

 

[Mark R]

So, are there any secret scrypt-ASICs deployed in quantity somewhere or is most of that hashpower deployed GPUs? If it's only GPUs, then that's an awful lot. Obviously, they would be spread out amongst various types from 7790s to 290X plus a few 780Ti owners mining with cudaminer.

Maybe there are some secret scrypt-ASICs. However, there are also some public ones which have been shipping since mid February. Supply is very tight, and most of the ASICs are going to mining farms, but a few sellers on the bitcointalk forum have been selling mining rigs or kits.

However, the ones currently on sale are not impressive. Pricing for a typical kit is about $0.75-$1/kH/s, with kits running between 300 kH/s - 6 MH/s. Energy cost is a big benefit, apparently about 30 W per MH/s.

You save a lot on power, but the capital cost is crazy. Compared to my 7950s (600 kH/s @ 240 W @ wall), the payback time would be over 1 year. If you're prepared to take a bet than scrypt mining is good for another 18 months, it would be worth buying these.

There are a few other companies who have scrypt ASICs in design phase - e.g. alpha-technology. However, they've been hitting big snags with anticipated die size and power. They were promising a much lower price ($8k for 25 MH/s) but considering that they haven't even worked out what lithography they're going to be using, and are having to redesign their layout due to power constraints, I wouldn't expect shipping units before Q4.

The Scrypt family of algorithms was specifically designed to be result in low performance and high cost when built into ASICs. It can be done, but as we are finding out, it's only an incremental change over GPUs, and not the total game changer that SHA256 ASICs were for BTC.

Don't believe the hype about some alt-coins, like darkcoin or quark about being ASIC proof. Yes, they use a lot of algorithms, which would make ASIC design time consuming. But all the algorithms used are specifically designed to be ASIC friendly - the "ASIC proof" bit comes from simply having a lot of different algorithms that require circuit design. The scrypt family of algorithms instead rely on using a large amount of memory. Memory, no matter how you build it, takes a lot of die space and adds a lot of latency, and with current technology this will always limit performance and drive up cost.

 

[geokilla]

scryptguild is pretty bad for rejects/stales unless you run a low intensity with --gpu-threads 2. So 7970/280x users are ok, but otherwise the cards you run at 19/20 intesity are going to do 5-8% rejects.

Only 0.5% rejects here with ScryptGuild. Two GPU threads, xintensity of 4, 650KH/s on my 7950.

 

[thilanliyan]

Only 0.5% rejects here with ScryptGuild. Two GPU threads, xintensity of 4, 650KH/s on my 7950.

intensity of only 4 and you're getting 650kh/s??!! :awe:

 

[JDG1980]

intensity of only 4 and you're getting 650kh/s??!! :awe:

Not regular intensity, xintensity. That's a new parameter available on the Kalroth fork of cgminer that allows somewhat more fine-grained control over the number of shaders used.

 

[Attic]

I'm going to try and upgrade my bamt sticks to use xintensity.

Basically regular intensity is:
13 - 8192
14 -16384
15 -32768
16 -65536
17 -131072
18 -262144
19 -524288
20 -1048576
21 -2097152
22 -4194304

Hard to dial in past a certain point because of the exponential growth of the power of 2 model it uses.

xintensity (-xI for bat?, or -xintensity for bat, "xintensity" : "value" for .conf)
is:
4 -(4n)
5 -(5n)
6 -(6n)
....

where n=your cards shader count.

xintensity of 4 is going to be relative to the card due to varying shader counts.

You need sgminer 4.1.0 or later or other supported mining software to use it.

I'm trying to get it going on bamt 1.3, but i'm a newb in linux and still confirming procedure.

Anyone have any good --gpu-threads 2 settings for 270/270x/290/290x? All the cards typically find best hash with gputhreads 1, but mining on scryptguild I need to get gputhreads to 2.

 

[IEC]

Power is cheap here so I previously didn't bother voltage optimizing my 290s, but with hot weather coming in a few months decided I'd better do something about the heat, as it will soon be a nuisance to heat the house.

So I've backed them all down to under 1GHz core clock, and am in the process of figuring out optimal undervolt offsets in MSI afterburner and applying it to each card individually. Currently have the first card I've tried it on mining stable at 855 kH/s @ 977/1250 at 187.5W VDDC Power In as measured by GPU-Z, VDDC 1.070V. I think I can probably reduce it further.

 

[Attic]

Power is cheap here so I previously didn't bother voltage optimizing my 290s, but with hot weather coming in a few months decided I'd better do something about the heat, as it will soon be a nuisance to heat the house.

So I've backed them all down to under 1GHz core clock, and am in the process of figuring out optimal undervolt offsets in MSI afterburner and applying it to each card individually. Currently have the first card I've tried it on mining stable at 855 kH/s @ 977/1250 at 187.5W VDDC Power In as measured by GPU-Z, VDDC 1.070V. I think I can probably reduce it further.

850c/1000m will do 800kh/s and should allow for substantial under volts, likely under 1000mv.

Let me know what you find. I'm in the same process and figuring if it's worth it to install windows so i can undervolt via afterburner. I'm unaware of anyway to undervolt in linux and there's no hawaii bios editor that works for modifying volts on the 290/290x.

 

[slashbinslashbash]

Don't believe the hype about some alt-coins, like darkcoin or quark about being ASIC proof. Yes, they use a lot of algorithms, which would make ASIC design time consuming. But all the algorithms used are specifically designed to be ASIC friendly - the "ASIC proof" bit comes from simply having a lot of different algorithms that require circuit design. The scrypt family of algorithms instead rely on using a large amount of memory. Memory, no matter how you build it, takes a lot of die space and adds a lot of latency, and with current technology this will always limit performance and drive up cost.

IMO the "ASIC proof" algorithms are just as flawed as SHA256, when it comes to concentration of hashpower. They're usually pretty GPU resistant as well, so they end up running efficiently on CPU's. Which means those with botnets or Amazon server farms can put everybody else out of business. Botnets in particular; they have 0 electricity costs.

This is why I'm buying into BitShares and other PoS-based crypto ideas. PoW is pretty much fundamentally flawed, in the long run. Just a giant waste of electricity to provide meaningless security. Bitcoin at difficulty 100 is just as secure as Bitcoin at difficulty 100,000,000,000; it's the blockchain that provides the security, not the difficulty. (Although a high difficulty can tell you that there are a lot of miners on the network, which all else being equal, does mean greater security; but again, if that mining power is all concentrated in the hands of a few, the security is an illusion.)

 

[IEC]

850c/1000m will do 800kh/s and should allow for substantial under volts, likely under 1000mv.

Let me know what you find. I'm in the same process and figuring if it's worth it to install windows so i can undervolt via afterburner. I'm unaware of anyway to undervolt in linux and there's no hawaii bios editor that works for modifying volts on the 290/290x.

I'll have to try 850/1000 clocks later, but at 947/1250 (avg 820 kH) or 977/1250 (avg 860 kH) my 290s seem to do okay. Here's what I've had a chance to test for a couple hours now:



One of my "test" mining rigs:


Edit: I should note that GPU-Z reports Watt values between 165W for the lowest card and 198W for the highest card now. If 850/1000 clocks can cut that further and still maintain an average of 800 kH per card I'd be very happy.

 

[Erenhardt]

snip

Nice. I would love to see Kill-A-Watt readings and how much efficiency improved!

 

[wand3r3r]

Is anyone else on 14.2?

I had been playing with my cards and switched them up with a fresh install of windows etc. but was having stability issues. In the end I removed AB, 13.12, and installed 14.2 the latest beta with a fresh copy of cgminer and started mining. So far I haven't played with the settings much and just basically started with stock settings and it was stable for a day or two. Now I just installed AB so I can start playing with the cards some more.

So far 14.2b looks ok. I haven't got my cards up to the hashrate they were at quite yet though. 14.1 had a dramatic decrease for whatever reason.

Morale of the story = don't switch your cards around when you have them working well. I'm fine tinkering with them again since it's looking promising and I wanted 14.2 for mantle.