Search content
Sort by
Showing 20 of 137 results by Grout
The rig boots fine, whatever the number of GPUs.
It's just that one of the GPUs is not recognized by the driver. I can still see it in lspci, though.
CPU is Intel(R) Celeron(R) CPU G1840 @ 2.80GHz
Update: I replaced all GPUs with 750Tis. 6 of them. And they all worked fine.
Another pointer towards the PSU: a Corsair CS750
It's just that one of the GPUs is not recognized by the driver. I can still see it in lspci, though.
CPU is Intel(R) Celeron(R) CPU G1840 @ 2.80GHz
Update: I replaced all GPUs with 750Tis. 6 of them. And they all worked fine.
Another pointer towards the PSU: a Corsair CS750
Wait, I was wrong, Bitchy is not always the same card.
After more shuffling around, it turns out it's always a card in the first OR second PCI-E slot. It just happenend to follow that card the first time and lead me to label Bitchy as the culprit. I'm sorry I doubted you Bitchy.
I swapped 6pin cables and it didn't change, although it doesn't mean much on a single-rail PSU.
Past experience tells me it's the PSU. It's always the PSU. Because it's the only part I don't have a spare for and Murphy's Law is a bitch...
Still interested in your experience with similar problems though.
After more shuffling around, it turns out it's always a card in the first OR second PCI-E slot. It just happenend to follow that card the first time and lead me to label Bitchy as the culprit. I'm sorry I doubted you Bitchy.
I swapped 6pin cables and it didn't change, although it doesn't mean much on a single-rail PSU.
Past experience tells me it's the PSU. It's always the PSU. Because it's the only part I don't have a spare for and Murphy's Law is a bitch...
Still interested in your experience with similar problems though.
Hi all,
I'm having a weird issue with a GTX 1060 :
With 4 GPUs, all cards are working fine.
When I add a 5th card, this particular card stops working (seen in lspci, not in nvidia-smi), let's call her 'Bitchy'.
I tried moving the cards around on different PCIe slots, it's always Bitchy that's unhappy.
It doesn't matter what the 5th card is, I tried with a 750ti and a 1060.
Bitchy just doesn't like when the rig gets too crowded...
Asrock H81 pro btc with molex plugged in and powered risers. Ubuntu 16.04. Driver Version: 375.74
750W Corsair PSU. 4 GTX 1060 (including bitchy), plus another card puts the total power draw to 600W tops...
Has anyone encountered that issue before ? Any idea ?
I'm having a weird issue with a GTX 1060 :
With 4 GPUs, all cards are working fine.
When I add a 5th card, this particular card stops working (seen in lspci, not in nvidia-smi), let's call her 'Bitchy'.
I tried moving the cards around on different PCIe slots, it's always Bitchy that's unhappy.
It doesn't matter what the 5th card is, I tried with a 750ti and a 1060.
Bitchy just doesn't like when the rig gets too crowded...
Asrock H81 pro btc with molex plugged in and powered risers. Ubuntu 16.04. Driver Version: 375.74
750W Corsair PSU. 4 GTX 1060 (including bitchy), plus another card puts the total power draw to 600W tops...
Has anyone encountered that issue before ? Any idea ?
As said above, so much of the rig's power use ends up in heat that you can consider a 100% transformation. I had one of my rigs in my bedroom for a while and it heated it fine.
The main issues to consider are the noise and light. Poorly built GPUs can make a lot of noise when running at full speed (I won't even talk about ASICs!). A blinking ethernet light can also be really annoying.
The main issues to consider are the noise and light. Poorly built GPUs can make a lot of noise when running at full speed (I won't even talk about ASICs!). A blinking ethernet light can also be really annoying.
Re: █▓▒░-< [ZPOOL.CA][BTC Multipool] The miners multipool >-░▒▓█
Oh Zpool. I like you very much but you make it really hard to write software that auto-switch to the most profitable pool (like MinerControl or my own software). Here is an extract from the values returned by your API for the neoscrypt pool :
This one abhorrent value (5375 BTC/Gh/day) screws up everyone that switches based on an average. I think it is default behavior for MinerControl (it definitely is for my software) to average the last 2 hours values to make a decision. This single 'error', makes everyone switch to you for the next two hours. Even if you are not the most profitable.
Guess I'll have to work around that and hope it is unintentional.
Code:
2.95757
2.94374
2.94605
2.95738
2.96167
2.96167
2.96167
5375.56269
13.16649
1.68837
2.10255
11.16034
5.42788
5.42788
5.19891
3.63841
3.70073
3.94264
3.94264
2.76667
2.76667
2.94374
2.94605
2.95738
2.96167
2.96167
2.96167
5375.56269
13.16649
1.68837
2.10255
11.16034
5.42788
5.42788
5.19891
3.63841
3.70073
3.94264
3.94264
2.76667
2.76667
This one abhorrent value (5375 BTC/Gh/day) screws up everyone that switches based on an average. I think it is default behavior for MinerControl (it definitely is for my software) to average the last 2 hours values to make a decision. This single 'error', makes everyone switch to you for the next two hours. Even if you are not the most profitable.
Guess I'll have to work around that and hope it is unintentional.
That must have been from UFO. It was diff blocked for a day or two and so when a block was found the chain, well, that would make the diff drop super low to allow it to catchup with blocks and thus making it jump the profitability. I would expect that it would report the same on all pools with UFO in their mix...
Well, the only other pool mining UFO I have data on is Hashpower, and they don't show the same fluctuation. In the same timeperiod, their highest value is 5.8. I only have one 'dot' per minute though, so it might have slipped between my API calls.
Anyway, I added a bit of code to ignore these 'anomalies' and it should be OK in the future.
Re: █▓▒░-< [ZPOOL.CA][BTC Multipool] The miners multipool >-░▒▓█
Oh Zpool. I like you very much but you make it really hard to write software that auto-switch to the most profitable pool (like MinerControl or my own software). Here is an extract from the values returned by your API for the neoscrypt pool :
This one abhorrent value (5375 BTC/Gh/day) screws up everyone that switches based on an average. I think it is default behavior for MinerControl (it definitely is for my software) to average the last 2 hours values to make a decision. This single 'error', makes everyone switch to you for the next two hours. Even if you are not the most profitable.
Guess I'll have to work around that and hope it is unintentional.
Code:
2.95757
2.94374
2.94605
2.95738
2.96167
2.96167
2.96167
5375.56269
13.16649
1.68837
2.10255
11.16034
5.42788
5.42788
5.19891
3.63841
3.70073
3.94264
3.94264
2.76667
2.76667
2.94374
2.94605
2.95738
2.96167
2.96167
2.96167
5375.56269
13.16649
1.68837
2.10255
11.16034
5.42788
5.42788
5.19891
3.63841
3.70073
3.94264
3.94264
2.76667
2.76667
This one abhorrent value (5375 BTC/Gh/day) screws up everyone that switches based on an average. I think it is default behavior for MinerControl (it definitely is for my software) to average the last 2 hours values to make a decision. This single 'error', makes everyone switch to you for the next two hours. Even if you are not the most profitable.
Guess I'll have to work around that and hope it is unintentional.
Re: [ANN]pledge for the release of the a 50% faster neoscrypt miner for nvidia cards
87a38017a306d45152261735da5a2c21473bd7891886783cd960b05bb2f45a67
Even if you don't reach the goal, consider this a thank you for your previous work and an incentive to opensource a little more, and troll a little less ;-)
Thanks for your support, can't promise much on the trolling part though Even if you don't reach the goal, consider this a thank you for your previous work and an incentive to opensource a little more, and troll a little less ;-)
(which is usually a reaction on something I consider as trolling)As far as windows 10........ I will never..never....never..install it on any of my computers...NEVER!
I'm with you there. I'm win 7 all the way and if that gets utterly obsolete, I'll join team linux instead.
We have cookies ;-)
Re: [ANN]pledge for the release of the a 50% faster neoscrypt miner for nvidia cards
87a38017a306d45152261735da5a2c21473bd7891886783cd960b05bb2f45a67
Even if you don't reach the goal, consider this a thank you for your previous work and an incentive to opensource a little more, and troll a little less ;-)
Even if you don't reach the goal, consider this a thank you for your previous work and an incentive to opensource a little more, and troll a little less ;-)
0.01 BTC guys. This is all I am asking 
There you go
697ff02bb2d2e1589eb2b8c40cbfd5584f477cbfe4ff6a942076ee54b8888075
Re: [POOL] YAAMP.COM multipool multialgo profit switch with exchange
Hi Guys,
It seems my original post has caused so much confusion I would like to explain a bit further. I was/am having trouble connecting to the pool. Here is what I was/am trying to do. I wanted to mine the 'Quark' algo...not the coin (payout to a different coin). While mining the 'Quark' algo, I entered a completely different coin/algo to be paid in (lets say Dash (x11), as an example). I set my GPU to the correct 'Quark' port of 4033. I entered my 'Dash' info (address), etc, but am getting the 'Authenication failure' error message.
Is that 'NOT' the way this pool operates? In order for me to mine a 'quark' algo, do I have to enter/mine to a quark address. If so, then your instructions are not very good and how do I get a pay-out to a different coin than what I mined?
I hope this clears up the problem and someone can now tell me what to do...other than tell me to read the friggin' directions.(not a noob, ya boob).
Thank You
It seems my original post has caused so much confusion I would like to explain a bit further. I was/am having trouble connecting to the pool. Here is what I was/am trying to do. I wanted to mine the 'Quark' algo...not the coin (payout to a different coin). While mining the 'Quark' algo, I entered a completely different coin/algo to be paid in (lets say Dash (x11), as an example). I set my GPU to the correct 'Quark' port of 4033. I entered my 'Dash' info (address), etc, but am getting the 'Authenication failure' error message.
Is that 'NOT' the way this pool operates? In order for me to mine a 'quark' algo, do I have to enter/mine to a quark address. If so, then your instructions are not very good and how do I get a pay-out to a different coin than what I mined?
I hope this clears up the problem and someone can now tell me what to do...other than tell me to read the friggin' directions.
Thank You
yiimp doesn't do coin conversion so you mine a coin, not an algo, and get paid in that coin. It doesn't work like the old yaamp.
Thank You...that was such a simple answer which got right to the jest of the question. So much extra was made of the original....Much appreciated!!
So, I assume if you want to mine GraniteCoin there, you have to point your miner to the x11 port, specify a GRN wallet address as username, and 'c=GRN' as your password, correct ?
You will not mine the highest profit (BowsCoin at the moment of writing), but only GraniteCoin. Correct?
I am fairly new to this and I have to ask, what are asic-resistant coins?
Basically, any coin that uses an algorithm for which there is no ASICs yet, is asic-resistant.
Some algorithms are made specifically in a way that makes building ASICs for them very hard/costly.
There's ASICs for SHA256 and Scrypt. And rumors circulating about X11...
Don't even try mining an ASIC'ed algo with a CPU/GPU, you'd just be burning money.
Any other algorithm, and you have a fighting chance.
(wouldn't go for evga either because they are probably more expensive and you buy mostly the brand... probably the case with corsair too tbh and they probably just rebrand seasonic psu )
yeah actually go for seasonic, they are building their psu...
)yeah actually go for seasonic, they are building their psu...
Generally I'd agree with that statement but the PSU I mentioned is off the charts regarding performance per dollar:
https://www.techpowerup.com/reviews/EVGA/SuperNOVA_G2_1300/9.html
I have terrible experience with lower capacity (<850W) EVGA PSU's though.
I found this PSU (the 1300W one) on the cheap and decided to give it a go.
20 minutes. I was still installing the OS when I heard a loud POP and everything went dark. No burnt smell though.
And I can't have been overloading it since the GPUs weren't plugged in at the time. Only CPU, mobo and HDD.
I spoke about it with a friend who often works with China and he told me it is common practice for some companies to replace capacitors with cheap ones and sell them (the capacitors) back to the manufacturer.
That's right, your brand new Asus motherboard (or EVGA PSU in that case) might not have the capacitors the manufacturer put on it...
They manage to make a profit because human labor is cheaper than the price difference between crappy capacitors and high-end ones. Unbelievable.
So take this as a lesson guys. A piece of hardware from a reputable brand shouldn't be bought from a non-reputable vendor, especially if the price is abnormally low.
Now I have to try to RMA this pretty expensive brick, and not insult their mothers in the process...
At least my rig didn't seem to suffer in the process. I slapped an old Corsair 550W in there and it seems fine.
you can find french retailer in france from evga their website I think.
Personnally I wouldn't buy a cooler master (I had problem with one) go for corsair, seasonic and other reputable brand (wouldn't go for evga either because they are probably more expensive and you buy mostly the brand... probably the case with corsair too tbh and they probably just rebrand seasonic psu)
yeah actually go for seasonic, they are building their psu...
Personnally I wouldn't buy a cooler master (I had problem with one) go for corsair, seasonic and other reputable brand (wouldn't go for evga either because they are probably more expensive and you buy mostly the brand... probably the case with corsair too tbh and they probably just rebrand seasonic psu
)yeah actually go for seasonic, they are building their psu...
I have two Corsair 550W gold that are perfect for 750Ti rigs. But given the 970 prices in France, my next cards are going to need more oomf...
I found a Seasonic 1250W gold for 255 on the same website. It should be fine if I mix the cards. Something like 4 970 + 2 750Ti per rig should do the trick...
And with more and more optimizations for different algos the power consumption will also increase so I personally like to use actual power consumption figures rather than what the advertised specs say.
Then, the EVGA wouldn't be enough either at 1300W. We'd have to go 5 GPUs, or 1500W PSUs, which are significantly less cost efficient.
I live in Central EU and there are 6 retailers selling the Evga PSU in my country and only one of them selling the Cooler Master V1200 so I have to assume there must be some webshops in France too selling the G2 for a reasonable price (260).
The main online retailers (LDLC, materiel.net) simply don't list EVGA PSUs, at all.
After a bit of digging, I found a smaller one that's not too shady and sells it for 230. Looks great!
It even lists a Super Flower Leadex 1600W - 80+ Gold at 290...
Now, the specs say 2 cables are 2*8pin, and 4 cables are 1*8pin. So you would still need adapters...
Most 970's have a 250W TDP so 6 (1500W) is way too much.
Also 6 pin power connectors can output 75W and 6+2 connectors can (should) output 150W so having 12 of them is somewhat pointless since you can't utilize all of them (1800W).
I use EVGA SuperNova G2 1300W PSU's and I think they are better than what I have have read about the Cooler Master V1200.
The G2 1300W is cheaper, have 3 years more warranty (10 instead of 7) and have lower ripple levels on the 12V rail.
It's "just" gold, not platinum but the difference in efficiency under standard mining load (70-80%) is negligible:
https://www.techpowerup.com/reviews/EVGA/SuperNOVA_G2_1300/6.html
https://www.techpowerup.com/reviews/CoolerMaster/V1200/6.html
Apparently they are louder than the V1200 but I never heard any of them even under heavy load, or at least much quiter than other components like GPUs.
Look into it and decide for yourself.
Also 6 pin power connectors can output 75W and 6+2 connectors can (should) output 150W so having 12 of them is somewhat pointless since you can't utilize all of them (1800W).
I use EVGA SuperNova G2 1300W PSU's and I think they are better than what I have have read about the Cooler Master V1200.
The G2 1300W is cheaper, have 3 years more warranty (10 instead of 7) and have lower ripple levels on the 12V rail.
It's "just" gold, not platinum but the difference in efficiency under standard mining load (70-80%) is negligible:
https://www.techpowerup.com/reviews/EVGA/SuperNOVA_G2_1300/6.html
https://www.techpowerup.com/reviews/CoolerMaster/V1200/6.html
Apparently they are louder than the V1200 but I never heard any of them even under heavy load, or at least much quiter than other components like GPUs.
Look into it and decide for yourself.
TDP is not real world consumption: my Gigabyte GTX 970 G1 Gaming (insane overclock out of the box: 1430 Mhz) pulls 180W from the wall mining Quark. 145W mining Lyra2v2.
If you really wanted to stay on the safe side, you could always replace 1-2 cards with 950s...
The EVGA looks nice too. But EVGA is a very US-centric company: it seems impossible to find in France. The only one I could find was on amazon.fr for 438. So that's a no-go for me...
There is also the Enermax Platimax EPM1350EWT 80PLUS Platinum for 295, but it only has 8 8pin connectors...
Any alternatives?
What do you guys think of this PSU for Nvidia GPU mining?
http://www.coolermaster.com/powersupply/enthusiast-v-series/v1200/
- Single rail, no connector headache
- 1200W platinum. Should power 6 GTX 970s while staying above 92% efficiency.
- 6 separate 2*8pin PCIe cables. No need for those pesky fire-prone adapters
- 3*3 molex cables. Should be able to handle 2 powered risers each
- 2*4 SATA cables, in case you want mine burst on the same rig ;-)
At 280, it seems like a pretty good option to me. What do you think? Do you know a better one?
http://www.coolermaster.com/powersupply/enthusiast-v-series/v1200/
- Single rail, no connector headache
- 1200W platinum. Should power 6 GTX 970s while staying above 92% efficiency.
- 6 separate 2*8pin PCIe cables. No need for those pesky fire-prone adapters
- 3*3 molex cables. Should be able to handle 2 powered risers each
- 2*4 SATA cables, in case you want mine burst on the same rig ;-)
At 280, it seems like a pretty good option to me. What do you think? Do you know a better one?
I just switched my desktop (the one with the GTX970 G1 Gaming) from windows to linux and got a very nice surprise: on windows, when mining, the mouse is lagging and the machine is pretty much unusable. On linux, no such thing! The machine is happily mining Quark at 17MH while I watch a 1080p video on Youtube.
What is this witchery? Not that I'm complaining...
BTW, it's about time for my donation, so here's a beer SP: 3bb1e5e50d3967cfc8f5e14aa700c316fc69c45380fcad938aed2d67fc62615c
which distro did you use? ... redhat based or debian based? ...What is this witchery? Not that I'm complaining...
BTW, it's about time for my donation, so here's a beer SP: 3bb1e5e50d3967cfc8f5e14aa700c316fc69c45380fcad938aed2d67fc62615c
My miner is on CentOS6, because nvidia's debian repository was acting up at the time.
My desktop is on Ubuntu. I don't like it that much, but once you clean it up, it's okay. And the list of supported OSes from nvidia is so short it doesn't leave much of a choice.
I just switched my desktop (the one with the GTX970 G1 Gaming) from windows to linux and got a very nice surprise: on windows, when mining, the mouse is lagging and the machine is pretty much unusable. On linux, no such thing! The machine is happily mining Quark at 17MH while I watch a 1080p video on Youtube.
What is this witchery? Not that I'm complaining...
BTW, it's about time for my donation, so here's a beer SP: 3bb1e5e50d3967cfc8f5e14aa700c316fc69c45380fcad938aed2d67fc62615c
What is this witchery? Not that I'm complaining...
BTW, it's about time for my donation, so here's a beer SP: 3bb1e5e50d3967cfc8f5e14aa700c316fc69c45380fcad938aed2d67fc62615c