Anyway I applied your settings to 2 of my GTX 1060 3GB and result is brilliant, 598 sol/s for both of them and they are only 3GB version.
That's great!
Let's use this thread if we find something better!
Topic: Feedback on overclock my 1070 (Read 289 times)
That's great!
Let's use this thread if we find something better!
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
I have actual +200 core and +750 memory with 70% power limit on G1 card, and is doing 455 sol/s best.
Whenever I tried to go down in core clock it lost 60 sol/s for -120 core clock and it lost about 85 sol/s when I switched from +750 memory to -500 one.
I think staying in the plus both in core and memory with 70% power limit should be the best settings.
-Power: 92
-Temp: 81
-CC: +61
-MC: +829
Now I'll try what philipma1957 suggested me.
Impressed at 505 sol/s from a Gtx 1070 , first time I see such performance from a card like that. Palit Super JetStream must be really good for mining. I think better stay here with a bit of more energy consuming rather to go almost 100 sol/s below with 70% power limit.
Anyway I applied your settings to 2 of my GTX 1060 3GB and result is brilliant, 598 sol/s for both of them and they are only 3GB version.
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
Ok, tried this for 10 mins now and it gives me 417-426 Sol/s @133W. The 1070 is rated @150W, so I think the efficiency is really good. But I still want to increase Sol/s.+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
Thank you for your input! This is surely very useful!
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
I have actual +200 core and +750 memory with 70% power limit on G1 card, and is doing 455 sol/s best.
Whenever I tried to go down in core clock it lost 60 sol/s for -120 core clock and it lost about 85 sol/s when I switched from +750 memory to -500 one.
I think staying in the plus both in core and memory with 70% power limit should be the best settings.
-Power: 92
-Temp: 81
-CC: +61
-MC: +829
Now I'll try what philipma1957 suggested me.
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
I have actual +200 core and +750 memory with 70% power limit on G1 card, and is doing 455 sol/s best.
Whenever I tried to go down in core clock it lost 60 sol/s for -120 core clock and it lost about 85 sol/s when I switched from +750 memory to -500 one.
I think staying in the plus both in core and memory with 70% power limit should be the best settings.
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
That's what I want to achieve! Is this a way of overclocking and undervolting the same card?
try
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
+125 core
-500 memory
70% tdp
I think you will get around 465 sols and save power
You are getting almost all what you can get from that card. Running 471 sol/s per second I think it is one of the best setups for a 1070 card.
You can try to increase the memory clock to max of +1000 if your cards hold to it but 800 it is a pretty good number. If it crashes when you increase to 1000 go back to 800 and stick to the settings you have found. Also increasing power limit to 90% or 100% can give a little boost in performance but I think you have already are in some of the best settings for your card.
My GIGABYTE G1 1070 does only 432 sol/s on Nicehash and 455 sol/s on EBWF cuda miner with +750 memory and +200 clock. I am going to follow your example and try to make -120 core clock to see if I get any improvement.
You can try to increase the memory clock to max of +1000 if your cards hold to it but 800 it is a pretty good number. If it crashes when you increase to 1000 go back to 800 and stick to the settings you have found. Also increasing power limit to 90% or 100% can give a little boost in performance but I think you have already are in some of the best settings for your card.
My GIGABYTE G1 1070 does only 432 sol/s on Nicehash and 455 sol/s on EBWF cuda miner with +750 memory and +200 clock. I am going to follow your example and try to make -120 core clock to see if I get any improvement.
Thanks for your reply.
Yes, I've read on some threads that mining zcash requires more memory than core, and they suggested to reduce Core clock and rise Memory clock. I don't know why, but it works
In any case I've saved this setup on Afterburner, so I can switch between coins with the best settings.
--a little edit to my previous post--
the precise settings, for nicehash, are:
Power: 100%
Temp: 83%
CC:+23
MC:+382
This allows me to reach 42.07MH/s
on EBWF cuda miner I go with:
Power: 91%
Temp: 83%
CC: -83
MC: +813
and I get 471sol/s
Fan speed, in both cases is between 20% and 40%.
What I want to do now is to play with power limit, since I believe it can give interesting results.
Ive got a Zotac 1070 Mini,
With core +180 & mem +540 (tdp 100%) i get 475sols on ebwf
With core +180 & mem +540 (tdp 100%) i get 475sols on ebwf
You are getting almost all what you can get from that card. Running 471 sol/s per second I think it is one of the best setups for a 1070 card.
You can try to increase the memory clock to max of +1000 if your cards hold to it but 800 it is a pretty good number. If it crashes when you increase to 1000 go back to 800 and stick to the settings you have found. Also increasing power limit to 90% or 100% can give a little boost in performance but I think you have already are in some of the best settings for your card.
My GIGABYTE G1 1070 does only 432 sol/s on Nicehash and 455 sol/s on EBWF cuda miner with +750 memory and +200 clock. I am going to follow your example and try to make -120 core clock to see if I get any improvement.
You can try to increase the memory clock to max of +1000 if your cards hold to it but 800 it is a pretty good number. If it crashes when you increase to 1000 go back to 800 and stick to the settings you have found. Also increasing power limit to 90% or 100% can give a little boost in performance but I think you have already are in some of the best settings for your card.
My GIGABYTE G1 1070 does only 432 sol/s on Nicehash and 455 sol/s on EBWF cuda miner with +750 memory and +200 clock. I am going to follow your example and try to make -120 core clock to see if I get any improvement.
Greetings!
I'm starting OC my cards, one by one. I've a 1070 Palit SJS and I'm mining Zcash on nanopool and sometimes use nicehash for the most profitable coin.
Now, at stock settings, it reach 440sol/s and 40.1 MH/s on Nicehash.
I would love to increase performance but I don't know what I should modify to rise earnings.
My first profile setting is:
Power Limit: 80%
Temp Limit: 80%
Core Clock: -120MHz
Memory Clock: +800MHz
giving me 471Sol/s
From this point, every slide I move, result in a decrease of the hash per second.
The problem is that I have not understood how this whole process works.
I'm also hoping there are some general settings in order to make my gpu more profitable.
N.B. so far, I'm using just one gpu, the others have to wait the cables to arrives. But I would love to learn this process, so when my hardware arrive, I will be ready to OC them.
Thanks
I'm starting OC my cards, one by one. I've a 1070 Palit SJS and I'm mining Zcash on nanopool and sometimes use nicehash for the most profitable coin.
Now, at stock settings, it reach 440sol/s and 40.1 MH/s on Nicehash.
I would love to increase performance but I don't know what I should modify to rise earnings.
My first profile setting is:
Power Limit: 80%
Temp Limit: 80%
Core Clock: -120MHz
Memory Clock: +800MHz
giving me 471Sol/s
From this point, every slide I move, result in a decrease of the hash per second.
The problem is that I have not understood how this whole process works.
I'm also hoping there are some general settings in order to make my gpu more profitable.
N.B. so far, I'm using just one gpu, the others have to wait the cables to arrives. But I would love to learn this process, so when my hardware arrive, I will be ready to OC them.
Thanks
Jump to: