Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 307. (Read 3426976 times)
June 25, 2014, 04:43:24 AM
Crytonote coin are too user unfriendly. I think those buying crytonote coins are mostly miner. Even if investor buy the coin, they will just leave it at exchange. Setting up a working wallet is too much work.
June 25, 2014, 04:42:39 AM
Well on AMD side 290x's are smoking 280x's out of the water with nearly double the hashrate. Probably thanks to the 512bit vs 384bit memory bus 
Hmm yeh,
650h/s = 512bit - 290X
450h/s = 384bit - 280X
220h/s = 128bit - 750Ti
nvidia still seems to have a better hashrate to bus speed ratio
June 25, 2014, 04:40:23 AM
I think right now higher end GPUs like the 780 and 780Ti should be mining something else. You can earn a lot more on a lot of other coins with them as the rate of a 780Ti over that of a 750Ti is not a lot.
On a side note, as the 780Ti has 3GB memory compared to the 750Ti's 2GB
How performance driven is this algo? Yes it is highly Memory bound BUT does having more memory on a slower card out way less on a higher card?
i.e. will a cheap 4GB gtx 730 perform the same as a 2GB 750Ti?
Yes one has more memory and the bus speed is the same, BUT
I would love to know this, as honestly, this algo may be doomed if a cheap £60 730 can get close to the 750Ti
On a side note, as the 780Ti has 3GB memory compared to the 750Ti's 2GB
How performance driven is this algo? Yes it is highly Memory bound BUT does having more memory on a slower card out way less on a higher card?
i.e. will a cheap 4GB gtx 730 perform the same as a 2GB 750Ti?
Yes one has more memory and the bus speed is the same, BUT
- What about memory versions? GDDR3 vs GDDR5?
- How is the actual performance of the card effecting it?
I would love to know this, as honestly, this algo may be doomed if a cheap £60 730 can get close to the 750Ti
Well on AMD side 290x's are smoking 280x's out of the water with nearly double the hashrate. Probably thanks to the 512bit vs 384bit memory bus
June 25, 2014, 04:25:35 AM
I think right now higher end GPUs like the 780 and 780Ti should be mining something else. You can earn a lot more on a lot of other coins with them as the rate of a 780Ti over that of a 750Ti is not a lot.
On a side note, as the 780Ti has 3GB memory compared to the 750Ti's 2GB
How performance driven is this algo? Yes it is highly Memory bound BUT does having more memory on a slower card out way less on a higher card?
i.e. will a cheap 4GB gtx 730 perform the same as a 2GB 750Ti?
Yes one has more memory and the bus speed is the same, BUT
I would love to know this, as honestly, this algo may be doomed if a cheap £60 730 can get close to the 750Ti
On a side note, as the 780Ti has 3GB memory compared to the 750Ti's 2GB
How performance driven is this algo? Yes it is highly Memory bound BUT does having more memory on a slower card out way less on a higher card?
i.e. will a cheap 4GB gtx 730 perform the same as a 2GB 750Ti?
Yes one has more memory and the bus speed is the same, BUT
- What about memory versions? GDDR3 vs GDDR5?
- How is the actual performance of the card effecting it?
I would love to know this, as honestly, this algo may be doomed if a cheap £60 730 can get close to the 750Ti
June 25, 2014, 04:18:53 AM
Minor update + Win binary: https://github.com/tsiv/ccminer-cryptonight/releases
No longer wasting CPU cycles on busy waiting, some checks on the initial memory allocations before getting to work and per device launch settings.
Seems like we have a lot of Win 8 users with problems, but apparently it works for some.
No longer wasting CPU cycles on busy waiting, some checks on the initial memory allocations before getting to work and per device launch settings.
Seems like we have a lot of Win 8 users with problems, but apparently it works for some.
I have no problems with windows 8.1, just the performance on 780ti is pretty bad compared to 750ti cards. i.e. I'm getting similar hashrates to what others have reported here for the 780ti ~270hash/s (using 4x120).
June 25, 2014, 03:18:01 AM
Yep, i dont dump atall now especially with the huge diff jump so i cant solomine now 
June 25, 2014, 03:17:04 AM
I could switch to Linux and compile the cryptonight miner, but to be honest it doesn't look like profitability merits the hassle right now. I mean this from the point of view that we have enough Beta testers.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
Honestly i dont think it will be too unstable. I am willing to try it but the income off it isnt enough to warrant the switch. Especially with how the price is fluctuating. Even JPC isnt fluctuating as much right now but is still a gamble. Here is a sample of my current page i was tracking some things with
JPC is what i am mining hence the 0 price.
That is accurate for 7 days. So that will give you an example of some price drops. And current value is calculated off the highest current buying offer at poloniex.
That is for 3x750Ti, and a 780
I'm sticking with JPC right now, too. I've stopped my daily dump as I feel there's another pump on the cards, as I see a MintPal listing within the next week or 2.
June 25, 2014, 03:01:16 AM
I could switch to Linux and compile the cryptonight miner, but to be honest it doesn't look like profitability merits the hassle right now. I mean this from the point of view that we have enough Beta testers.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
Honestly i dont think it will be too unstable. I am willing to try it but the income off it isnt enough to warrant the switch. Especially with how the price is fluctuating. Even JPC isnt fluctuating as much right now but is still a gamble. Here is a sample of my current page i was tracking some things with
JPC is what i am mining hence the 0 price.
That is accurate for 7 days. So that will give you an example of some price drops. And current value is calculated off the highest current buying offer at poloniex.
That is for 3x750Ti, and a 780
June 25, 2014, 02:44:46 AM
I could switch to Linux and compile the cryptonight miner, but to be honest it doesn't look like profitability merits the hassle right now. I mean this from the point of view that we have enough Beta testers.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
If you need Win 8.1 testers on any subsequent release I will be happy to assist. As it stands right now, it is too unstable on 8.1 to be any kind of mining solution.
June 25, 2014, 02:34:53 AM
Minor update + Win binary: https://github.com/tsiv/ccminer-cryptonight/releases
No longer wasting CPU cycles on busy waiting, some checks on the initial memory allocations before getting to work and per device launch settings.
Seems like we have a lot of Win 8 users with problems, but apparently it works for some. It is a mystery, my money is still on hitting TDR. If that's the case http://support.microsoft.com/kb/2665946 will provide a workaround. It's not pretty but increasing the timeout from the default 2 seconds to something like 5 seconds shouldn't hurt anyone. The real fix would be breaking up the kernel into smaller pieces that are run in sequence as opposed to one huge kernel that does the whole hash in one go and takes near forever (well, over 2 seconds) to do it. Looking into that, should be relatively easy. A compute 2.0/2.1 version shouldn't be too hard either but I'm not sure if they're going to perform too well and worse than bad is... very bad?
No longer wasting CPU cycles on busy waiting, some checks on the initial memory allocations before getting to work and per device launch settings.
Seems like we have a lot of Win 8 users with problems, but apparently it works for some. It is a mystery, my money is still on hitting TDR. If that's the case http://support.microsoft.com/kb/2665946 will provide a workaround. It's not pretty but increasing the timeout from the default 2 seconds to something like 5 seconds shouldn't hurt anyone. The real fix would be breaking up the kernel into smaller pieces that are run in sequence as opposed to one huge kernel that does the whole hash in one go and takes near forever (well, over 2 seconds) to do it. Looking into that, should be relatively easy. A compute 2.0/2.1 version shouldn't be too hard either but I'm not sure if they're going to perform too well and worse than bad is... very bad?
June 25, 2014, 02:32:32 AM
That is relying on 100% uptime pools. So at that its 0.005btc a day for 3 750tis which isnt bad but i think is too low compared to jackpot as a whole with my 780. Good if your willing to hold and hope xmr goes back to 0.01btc again
i don't remember if the block reward is based on the difficulty or is fixed
June 25, 2014, 02:24:43 AM
Anyone got this running on amazon ec2 grid? I ran it yesterday and barely managed 100 H/s with 1 GPU.
could someone try to post a tutorial on how to use amazon ec2 (or else... I think it is possible to get a few free hours... and I'd like to try that)Can anyone help me to go through all the procedure to compile this on linux? I can't comple it on ubuntu.
Thk
June 25, 2014, 02:18:54 AM
That is relying on 100% uptime pools. So at that its 0.005btc a day for 3 750tis which isnt bad but i think is too low compared to jackpot as a whole with my 780. Good if your willing to hold and hope xmr goes back to 0.01btc again
June 25, 2014, 02:14:19 AM
If i knew it was profitable i would maybe move some cards over as a test for windows 8. How much is it estimates to earn in a day at 220h/s?
Bare in mind im still sitting at jackpot coin making 0.0016btc a day per 750ti
Bare in mind im still sitting at jackpot coin making 0.0016btc a day per 750ti
the net hash is 11MH, block time is 60sec, block reward is about 16, this mean that everyday about 23k monero are mined
one rig 6 x 750 does 1.2k, which is about 1/10k of the net , 23k/10k = 2.3-2.5 monero a day
June 25, 2014, 02:04:31 AM
If i knew it was profitable i would maybe move some cards over as a test for windows 8. How much is it estimates to earn in a day at 220h/s?
Bare in mind im still sitting at jackpot coin making 0.0016btc a day per 750ti
Bare in mind im still sitting at jackpot coin making 0.0016btc a day per 750ti
June 25, 2014, 02:03:08 AM
Getting driver crashes with the TSIV compiled binary on Windows 7 with GTX 770 and GTX 780, as well as with Windows 8.1 with GT 750M. Hash rates show crazy numbers but no actual work. From the GT 750M laptop:
Is this because I don't have the latest drivers, or maybe it's because these are Kepler (not Maxwell) GPUs? Or maybe I just need to update drivers...trying that now.
Is this because I don't have the latest drivers, or maybe it's because these are Kepler (not Maxwell) GPUs? Or maybe I just need to update drivers...trying that now.
So, I think we can officially say windows 8 isn't supported ^^" It isn't really a mining OS though, best results are even under Linux.
But yeah, it should be fixed if possible
But I had a bad time too to get my 760 working, I tried restarting the miner, resetting in device manager, taking off any OC, rebooting... After some random tries, it suddenly started working. But I don't think Kepler is even worth it for Cryptonight, being almost half of the 750TI's rate, brrr.
I pointed it back to X11, leaving my 2 750TI's to do the monero dance
?
my win8 6x750ti rig works fine.
June 25, 2014, 02:01:39 AM
for XMR , can i mine directly payout to exchange address ?
No, don't even think about trying it. Will not work.
Exchanges need a both a base address and a transaction address (ID address).
Download the wallet and create and ID for mining.
Use this while the deamon catches up which will take a long time as it's over 1 GB block train.
Send ALL transactions from pools to your local wallet and then transfer to exchange.
If anyone is super stuck I can generate a wallet address for you to get started mining on XMR or QCN.
I'll transfer this info to you so you can take possession of coins but you should then generate a new wallet address as soon as you have caught up on the block chain.
Let me know if I can assist any of you. It's not hard (but different) then any other alt coins.
Carlo
PS I'm as trustworthy as I claim.
But kicking aside let me know if I can help any of you out getting started with CryptoNight. It's different for sure but not hard.
You just want his coins
Yea, because I'm the stupidest scammer on the plannet and will use my real credential to "scam" with.
But seriously, I would agree this in general isn't a good idea. But it can take a long time to sync up with the block chain.
I also figure if I offer to help, I've got enough "credentials" around here for people to know I'm not a scammer and really only offering to help and i'm not going to steal a few XMR worth of coins.
But yea, in truth there are only a few people on this forum I would trust with any large amount of coins if truth be told.
Chris, Bombadil, djm34, cryptomines, zelante, dma etc. (I'm sure there are a few more)
Carlo
Im not even in your list
June 25, 2014, 01:46:54 AM
so cryptonight run overnight quite well on three linux rigs and without any problems.
without problems is not exactly true... the pools are not that stable. one pool made a comfy coffee break for 5 hours.
sadly the kopiemTu monitor process do not work with tvis ccminer release at the moment, so I have not the comfort of failover anymore.
and all these cryptonight pools would badly need it.
I don't think you can pass launch configs to ccminer...cudaminer takes those, not cc, afaik.
I read it on tsiv's post & on github... this combination would not come to my mind otherwise.
this switch gives me 20-25h/s more per card.
without problems is not exactly true... the pools are not that stable. one pool made a comfy coffee break for 5 hours.
sadly the kopiemTu monitor process do not work with tvis ccminer release at the moment, so I have not the comfort of failover anymore.
and all these cryptonight pools would badly need it.
btw. your -l 8x60 is also with my rigs the fastest option i tried so far
Quote from: kopiemTu console output
[2014-06-24 17:57:13] accepted: 46/46 (100.00%), 1372.57 H/s (yay!!!)
I don't think you can pass launch configs to ccminer...cudaminer takes those, not cc, afaik.
I read it on tsiv's post & on github... this combination would not come to my mind otherwise.
this switch gives me 20-25h/s more per card.
June 25, 2014, 01:10:05 AM
Getting driver crashes with the TSIV compiled binary on Windows 7 with GTX 770 and GTX 780, as well as with Windows 8.1 with GT 750M. Hash rates show crazy numbers but no actual work. From the GT 750M laptop:
Yeah, I wish a GT 750M could get those hash rates! Same things basically on Win7 with the two GPUs:
Is this because I don't have the latest drivers, or maybe it's because these are Kepler (not Maxwell) GPUs? Or maybe I just need to update drivers...trying that now.
Code:
C:\Downloads\Monero\ccMiner-cryptonight>ccminer.exe -a cryptonight -o stratum+tcp://mine.cryptoescrow.eu:3333 -u 48JM22E3ZfPSoFCukcizpSR2hCsBnAExT4ACvrpYx5czFgEyR12LWwK9JpgYRZKjsRHp8ynDcQegbhCspvjHd7gaL8qbzYy -p x
*** ccMiner for nVidia GPUs by Christian Buchner and Christian H. ***
This is version 1.1 (beta)
based on pooler-cpuminer 2.3.2 (c) 2010 Jeff Garzik, 2012 pooler
based on pooler-cpuminer extension for HVC from
https://github.com/heavycoin/cpuminer-heavycoin
and
http://hvc.1gh.com/
Cuda additions Copyright 2014 Christian Buchner, Christian H.
LTC donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm
BTC donation address: 16hJF5mceSojnTD3ZTUDqdRhDyPJzoRakM
YAC donation address: Y87sptDEcpLkLeAuex6qZioDbvy1qXZEj4
[2014-06-25 00:50:05] Using JSON-RPC 2.0
[2014-06-25 00:50:05] Using 40 blocks of 8 threads for cryptonight kernel
[2014-06-25 00:50:05] 1 miner threads started, using 'cryptonight' algorithm.
[2014-06-25 00:50:05] Starting Stratum on stratum+tcp://mine.moneropool.com:3334
[2014-06-25 00:50:05] Pool set diff to 25000.1
[2014-06-25 00:50:05] Stratum detected new block
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 136.85 H/s
[2014-06-25 00:50:07] GPU #0: GeForce GT 750M, 8312687.31 H/s
[2014-06-25 00:50:09] GPU #0: GeForce GT 750M, 207583755.19 H/s
[2014-06-25 00:51:14] Stratum detected new block
[2014-06-25 00:51:14] GPU #0: GeForce GT 750M, 2379764.39 H/s
[2014-06-25 00:51:15] GPU #0: GeForce GT 750M, 205302327.42 H/s
Code:
[2014-06-24 22:54:14] Using JSON-RPC 2.0
[2014-06-24 22:54:14] Using 40 blocks of 8 threads for cryptonight kernel
[2014-06-24 22:54:14] Starting Stratum on stratum+tcp://mine.moneropool.com:3334
[2014-06-24 22:54:14] Binding thread 1 to cpu 1
[2014-06-24 22:54:14] Binding thread 0 to cpu 0
[2014-06-24 22:54:14] 2 miner threads started, using 'cryptonight' algorithm.
[2014-06-24 22:54:14] Pool set diff to 25000.1
[2014-06-24 22:54:14] Stratum detected new block
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:16] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 110.95 H/s
[2014-06-24 22:54:17] GPU #1: GeForce GTX 770, 6721000.00 H/s
[2014-06-24 22:54:16] GPU #0: GeForce GTX 780, 111.72 H/s
[2014-06-24 22:54:18] GPU #1: GeForce GTX 770, 230949064.20 H/s
[2014-06-24 22:54:27] GPU #0: GeForce GTX 780, 673.48 H/s
[2014-06-24 22:54:27] GPU #0: GeForce GTX 780, 673.48 H/s
[2014-06-24 22:54:27] GPU #0: GeForce GTX 780, 40641000.00 H/s
Try fiddling with the -l flag to try and find a launch configuration that doesn't make the driver immediately crash. My 780 Ti didn't like more than 4 threads (either the performance was lower for the same Blocks x Threads, or it would outright crash), and even using -l 4x135, I only got 275 H/s or so.
It's likely more profitable to mine x11, x13, et al., rather than XMR, on higher-end GPUs, at least with the current Nvidia cryptonight miner from tsiv.
June 25, 2014, 01:09:18 AM
Oh man, if anybody has a Fermi/Compute 2.0 build of tsiv's CryptoNight fork that would be amazing... (GTX 570)
(GTX 570)Would be hugely appreciated if someone does have the time for the above,have some old 560 TI's knocking about i could use
Jump to: