Interesting Crypto Behavior

Korishan

Administrator
Joined
Jan 7, 2017
Messages
7,539
Ok, so I took Mike's advice and decided to run nanopool (that was a long and twisted road of learning!). So I ended up downloading Claymore for the CPU and the AMD GPU. Got everything configured and such.
However, I was noticing the hash rates were a lot lower than they were when I was running NiceHash. And yeah, they are both CryptoNight algorithms. And it was running about 1/2 as fast.
So I was talking with someone else about what was happening, and it dawned on me, couldn't I use the NH's programs and configure them for my needs? Change them from NH to nanopool.
Well, turns out I could. And way easy to configure, too. The two programs are xmrig and xmr-stak (CPU and GPU respectively). So I pulled them out of the folders, copied them to my Mining folder and started looking at the config files and adjusting them my pools wallet, etc.
So I run xmrig and I'm getting about 115H/s, about 20 more than I was getting running through NH. In xmr-stak, it was running about 120 H/s, again, about 20 more than from within NH.
At the time, I didn't realize, that by running xmr-stak in standalone, it was hashing on the CPU "and" the GPU at the same time. So the 120 was combined.
But here's the oddity: Running xmrig by itself yielded 115H/s, and running xmr-stak along side it, it still got 115H/s AND I was getting about 61-65H/s from xmr-stak on the CPU. So nearly 175H/s with the combined power of the two.
Now, my question is, why is this? I'm confused as to why I can run two algos on the CPU and get better performance out of it. I figured the xmrig would of decreased by at least 20H/s, if not more. But it ran stable at that speed for over 20minutes. Same with the xmr-stak results.

I haven't tried it with running the xmrig-AMD mining with the xmr-stak and see what results I get. I think I'll try that out now.
 
Ok, cleared up a little bit about the numbers. It's not as large as I initially thought. It's closer to about 5H/s faster, on average. The numbers that were being displayed in the program were average numbers over 30 or 60 seconds, depending on which one it was. I changed the refresh rate to 5 seconds, and I saw a more accurate picture of what was going on.

With the numbers being average, there were times when one miner was waiting for the next piece and wasn't hashing. At that time, the other program was hashing and jumped in performance. So the numbers got a bit skewed that way.

Drat, I was hopeing to get free hashes :p Ohwell! Live, learn, tweak, and learn some more!
 
Back
Top