Author

Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 797. (Read 2347664 times)

full member
Activity: 229
Merit: 100
X11 is another option, but it has to improve a lot before it will be comparable to either of them. ROI currently for this kernel is negative.

My latest private does 3.1MHASH@50watt. (750ti)
Nicehash x11 miner does 11MHASH@(250watt!!) (r9 280x)

15,5/11= 40% bether.

My NVIDIA kernal is mining x11 40% more efficient than the fastest x11 kernal on AMD cards..


My 3card r9 280x rig pulled over 800watt in the wall when I tested the latest Nicehash miner. Quark/qubit was pulling around 700 with only 3 cards. (Watt messured including cpu/motherboard and harddrive)


Nicehash X11 does 20MHASH @ 100 watt with the right card.

so 7 x 750ti = also 20Mhash but 350 watt.

 Cool

But than again, price of a 750ti is what? 150 Bucks? no idea.

The "right card" = 450bucks.

Ah, it's still more profitable :p
legendary
Activity: 2940
Merit: 1091
--- ChainWorks Industries ---
I tried an experiment last night going back to solo mining the quark algo instead of pool mining and, while it wasn't a complete failure, performance was miserable. Same hash rates, only about 70% acceptance. Running ccminer in full auto mode. Looking for suggestions to get the acceptance ratio back to the 97.5% territory?

750ti, sp-mod 1.5.77 compiled with CUDA 7.5 on Xubuntu 14.04, nVidia driver 352.68.

Which coin? Did you try another one?

i get high share acceptance with ccminer-spmod78 c7.5 linux ( fedora 23 x64 ) solo mining granite - which is x11 ...

so pallas might be on the right track - maybe its the coin or the settings you are using ...

#crysx
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
I tried an experiment last night going back to solo mining the quark algo instead of pool mining and, while it wasn't a complete failure, performance was miserable. Same hash rates, only about 70% acceptance. Running ccminer in full auto mode. Looking for suggestions to get the acceptance ratio back to the 97.5% territory?

750ti, sp-mod 1.5.77 compiled with CUDA 7.5 on Xubuntu 14.04, nVidia driver 352.68.

Which coin? Did you try another one?
sr. member
Activity: 346
Merit: 250
I tried an experiment last night going back to solo mining the quark algo instead of pool mining and, while it wasn't a complete failure, performance was miserable. Same hash rates, only about 70% acceptance. Running ccminer in full auto mode. Looking for suggestions to get the acceptance ratio back to the 97.5% territory?

750ti, sp-mod 1.5.77 compiled with CUDA 7.5 on Xubuntu 14.04, nVidia driver 352.68.
sp_
legendary
Activity: 2954
Merit: 1087
Team Black developer
X11 is another option, but it has to improve a lot before it will be comparable to either of them. ROI currently for this kernel is negative.

My latest private does 3.1MHASH@50watt. (750ti)
Nicehash x11 miner does 11MHASH@(250watt!!) (r9 280x)

15,5/11= 40% bether.

My NVIDIA kernal is mining x11 40% more efficient than the fastest x11 kernal on AMD cards..


My 3card r9 280x rig pulled over 800watt in the wall when I tested the latest Nicehash miner. Quark/qubit was pulling around 700 with only 3 cards. (Watt messured including cpu/motherboard and harddrive)
legendary
Activity: 3164
Merit: 1003
Isn't there a better thread to talk about eth mining on nvidia?
I don't mine eth and I had to skip a couple pages at least.
actually there is a whole forum to talk about ethereum  Grin (but it is like miners, they want all the algo in the same miner and
all the discussion in the same thread... actually the old Buchner cudaminer thread has always been the thread where everything was discussed, so I am not really chocked by this)
True djm.
legendary
Activity: 1400
Merit: 1050
Isn't there a better thread to talk about eth mining on nvidia?
I don't mine eth and I had to skip a couple pages at least.
actually there is a whole forum to talk about ethereum  Grin (but it is like miners, they want all the algo in the same miner and
all the discussion in the same thread... actually the old Buchner cudaminer thread has always been the thread where everything was discussed, so I am not really chocked by this)
legendary
Activity: 1764
Merit: 1024

z.gif

oh wait what gpu-z version is this?

It seems your vram actually IS overclocked!! at 1900 mhz (stock is 1750)

which translates to +300 mhz in eg for precision x

or 7600 mhz effective

Yup 3800 with offset (not sure where you're getting 7600 from, it's doubled for the readout in Nvidiainspector)... That doesn't make that big of a difference though. He should be about 20Mhs or so much like most other people with OC'd memory. Doesn't explain the extra 2-3Mhs.

Stock is 1500 before it's doubled or quadrupled (not many people use this).

Isn't there a better thread to talk about eth mining on nvidia?
I don't mine eth and I had to skip a couple pages at least.

I think it's very applicable here as a lot of people are mining Eth right now, you should be too unless you have some sort of super secret coin you're mining right now that's really profitable. It's the most profitable thing to mine, even though Nvidia cards are crippled.

I don't complain when people are talking about some sort of random issue that doesn't apply to me. I don't know why I would anyway, this thread is all about miner information dissemination. Talking is part of that process. If that stops, this thread loses it's relevancy.
legendary
Activity: 3164
Merit: 1003
Really quick test, all cards p0 and memory +500 = 4010 MHz

asus strix samsung ~199 GB
asus strix elpida ~199 GB
palit reference hynix ~190 GB

Samsung and elpida are hashing 22MH but hynix only 19MH. Hynix needs +750 (!!!) on memory before it crosses 20MH line.

thanks for the stats

My cards are hashing in P 2 state but at stock memory clock (bios mod).
BUT in P 2 state the oc for the memory doesn't stick, it only works in P 0 state to oc successfully.

Thats why even at +500 my bandwidth didn't change  Cheesy

explanation on how to switch to p0 state were given in the pledge thread, it should apply to your card as well
Yes go here   Grim     https://bitcointalk.org/index.php?topic=1317772.120
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Isn't there a better thread to talk about eth mining on nvidia?
I don't mine eth and I had to skip a couple pages at least.
legendary
Activity: 1400
Merit: 1050
Really quick test, all cards p0 and memory +500 = 4010 MHz

asus strix samsung ~199 GB
asus strix elpida ~199 GB
palit reference hynix ~190 GB

Samsung and elpida are hashing 22MH but hynix only 19MH. Hynix needs +750 (!!!) on memory before it crosses 20MH line.

thanks for the stats

My cards are hashing in P 2 state but at stock memory clock (bios mod).
BUT in P 2 state the oc for the memory doesn't stick, it only works in P 0 state to oc successfully.

Thats why even at +500 my bandwidth didn't change  Cheesy

explanation on how to switch to p0 state were given in the pledge thread, it should apply to your card as well
legendary
Activity: 1400
Merit: 1050
full member
Activity: 155
Merit: 100
what gpu-z version is this?
It seems your vram actually IS overclocked!! at 1900 mhz (stock is 1750)
GPU-Z 0.8.6
you have another 970 to test Huh
one card 970..
full member
Activity: 155
Merit: 100
the last mhash (you said 23-24)
22-23 , not 23-24.
@ Vita1ico do you have any other 970 to test?
Unfortunately, others do not. Only 970 + 960 (two cards)

sr. member
Activity: 506
Merit: 252
ETH is memory constrained thats why you cant really optimize it. (which was a VERY smart move)

Yes you can. The current miner is not hashing as fast as the copyspeed limit.
Creating the growing dagger file was a stupid move. Soon it will grow to big for the GPU, and then Etherum will go POS or be food for the botnets.


While the TLB trashing sucks I still think it is one of the most brilliant algos around.
More modern gpu designs seem to push the TLB trashing limit upwards.

ETH goes POS anyway... (now that is a mistake, imo)
dunno about Factom tho...
sp_
legendary
Activity: 2954
Merit: 1087
Team Black developer
ETH is memory constrained thats why you cant really optimize it. (which was a VERY smart move)

Yes you can. The current miner is not hashing as fast as the copyspeed limit.
Creating the growing dagger file was a stupid move. Soon it will grow to big for the GPU, and then Etherum will go POS or be food for the botnets.
sr. member
Activity: 506
Merit: 252
Also this link provides info on how to hex edit a amd bios for tighter vram timings.

http://www.overclock.net/t/1554360/tahiti-memory-timings-patch-for-hynix-vram
sr. member
Activity: 506
Merit: 252
run this and compare to your other cards

oclmembench.exe
https://drive.google.com/file/d/0B9OoHSmkeSeNNEE3ZkpsSWlnZ1k/view

full member
Activity: 155
Merit: 100
However any changes of numbers BIOS to do increase +5 Mhs. Because the factory BIOS is not more than 17 .
sr. member
Activity: 506
Merit: 252
The bios tools available are lucky to tell you the brand and let you adjust the frequency only.
You can not adjust vram voltage neither can you adjust vram timings. (sadly)

see post #9155
https://bitcointalksearch.org/topic/m.13591064
Jump to: