It's easy to compare scrypt performances between whatever, your problem, which is the difference you are seeing between rigs, is due to the amount of errors you're getting, which is a lot. In fact, everything > 1 is bad. You should have 0 HW errors.
Go over your config and readjust it to fit your cards.
Topic: Comparing Applies and Oranges in cgminer for scrypt (Read 363 times)
I thought I understood all of cgminer's output when mining in BTC. Now, I find I am lost to really truly understand it when applied to scrypt especially when I want to compare performance on two different scrypt coins (or even more complex, two different coin switching pools).
For example, I have 2 identical machines (each 2.84 Mh/s) mining on two different pools. They started mining at the same time and have been going for over a week, so the statistics should not be affected much by luck.
MinerA on Pool#1 is pulling A:38495 R:612 errors:229 Utility:9.85 Diff1:39612 and pool (vardiff set) difficulty 256
MinerB on Pool#2 is pulling A:23717 R:261 errors:1027 Utility:5.99 Diff1:23982 and pool (vardiff set) difficulty 256
I don't think I can really compare accepted shares or utility, since they are mining different coins at different network difficulty all the time and it really is apples to oranges isn't it??? Or is that wrong? Is the accepted share count and/or utility a good basis for comparison (in which case, i should dump Pool#2)?
If the shares cannot be accurately compared due to different coins/network difficulty, then can I use the diff1 to compare diff1 equivalent shares?
If none of the above are valid for comparison from pool#1 to pool#2, is there a statistic I can use or implement that would compare performance (not in terms of return, since returns are really a function of price fluctuations and for now I really want to get to the issue of which pool works most efficient with my miners).
For example, I have 2 identical machines (each 2.84 Mh/s) mining on two different pools. They started mining at the same time and have been going for over a week, so the statistics should not be affected much by luck.
MinerA on Pool#1 is pulling A:38495 R:612 errors:229 Utility:9.85 Diff1:39612 and pool (vardiff set) difficulty 256
MinerB on Pool#2 is pulling A:23717 R:261 errors:1027 Utility:5.99 Diff1:23982 and pool (vardiff set) difficulty 256
I don't think I can really compare accepted shares or utility, since they are mining different coins at different network difficulty all the time and it really is apples to oranges isn't it??? Or is that wrong? Is the accepted share count and/or utility a good basis for comparison (in which case, i should dump Pool#2)?
If the shares cannot be accurately compared due to different coins/network difficulty, then can I use the diff1 to compare diff1 equivalent shares?
If none of the above are valid for comparison from pool#1 to pool#2, is there a statistic I can use or implement that would compare performance (not in terms of return, since returns are really a function of price fluctuations and for now I really want to get to the issue of which pool works most efficient with my miners).
Jump to: