That's good enough for me. It'll be fixed in the next release. |
|
|
|
Trying to mine SHA256t on Zpool, every share is rejected with "reject reason: Low difficulty share", unless using "-m 256"
I do wonder if its pool side issue or miner?
cpuminer-avx2.exe -o stratum+tcp://sha256t.mine.zpool.ca:3339 -u MM8RmXUgxDwHJxrC54muF7KHciSCFS3gx3 -p c=LTC -a sha256t
ps. tested oldest versions, and I see same issue in 3.5.9.1 and 3.6.3
It worked with the original coin that introduced the algo. That error often means the hash is incomplete but I've also seen it due to a pool config error. This could be a slightly different algo using the same name but I'd have to compare with a working miner. It looks like the original coin was revived. https://bitcointalk.org/index.php?topic=2313497.msg23484925#msg23484925It's a simple to change the difficulty targetting in software but I want some confirmation that zpool is correct before I make the change. Has anyone tried other pools? |
|
|
|
[a long unrelated post]
And also would be great to add argon2d algos to the next build  )) It would be great if you didn't quote an entire, long, unrelated post and did some reading before posting. |
|
|
|
I've contacted JayDDee last year and wanted to help porting the project to the latest visual studio edition, and improve performance. Here I am.
I told you at the time the project files would have to be rebuilt from scratch. |
|
|
|
Trying to mine SHA256t on Zpool, every share is rejected with "reject reason: Low difficulty share", unless using "-m 256"
I do wonder if its pool side issue or miner?
cpuminer-avx2.exe -o stratum+tcp://sha256t.mine.zpool.ca:3339 -u MM8RmXUgxDwHJxrC54muF7KHciSCFS3gx3 -p c=LTC -a sha256t
ps. tested oldest versions, and I see same issue in 3.5.9.1 and 3.6.3
It worked with the original coin that introduced the algo. That error often means the hash is incomplete but I've also seen it due to a pool config error. This could be a slightly different algo using the same name but I'd have to compare with a working miner. |
|
|
|
I am proposing fuurther canges to the Windows binaries suite to be better aligned with the CPU features actually used by cpuminer-opt. Comments are welcome. The problem: 1. AVX is meaningless to cpuminer-opt. AVX adds no useful 128 bit SIMD instructions, full 128 bit support only requires SSE4.2. As a result some CPUs are prevented from using optimzations they support. This also occurs when compiling from source. 2. cpuminer-sse is actually built with SSSE3. Some AMD CPUs with SSE2 do not have SSSE3 and can't run cpuminer. The proposal: 1. Eliminate cpuminer-aes-avx binary and eliminate AVX as a reportable feature and replace it with SSE4.2 cpuminer-aes-sse42 will include all optimizations from the deleted AVX build. 2. Remove SSSE3 from cpuminer-sse2 build. Pro/con: The only theoretical con is that Intel core2 CPUs will not make use of SSSE3 when using cpuminer-sse2 windows bnary. There are no significant differences with SSSE3 so no performance impact is expected. Some previously unsupported AMD CPUs will become supported. Some optimizations currently tagged as requiring AVX will become available on Intel Westmere CPUs and possibly some AMD CPUs with similar features. Fewer binary builds, should be less confusing for users. Current suite: cpuminer-sse2.exe "-march=core2" Core2, Nehalem
cpuminer-aes-sse42.exe "-maes -msse4.2" Westmere
cpuminer-aes-avx.exe "-march=corei7-avx" Sandybridge, Ivybridge
cpuminer-avx2.exe "-march=core-avx2" Haswell...
cpuminer-avx2-sha.exe "-march=core-avx2 -msha" Ryzen
Proposed: cpuminer-sse2.exe "-msse2" Core2, Nehalem
cpuminer-aes-sse42.exe "-maes -msse4.2" Westmere, Sandybridge, Ivybridge
cpuminer-avx2.exe "-march=core-avx2" Haswell...
cpuminer-avx2-sha.exe "-march=core-avx2 -msha" Ryzen
|
|
|
|
...
can you link a post where someone explains how to run the miner. which command etc. that would be really nice. Or maybe link that specific post or instructions in the ANN post?
cpuminer-aes-sse42 -h replace cpuminer-aes-sse42 with the build your using, pools will also have quickstart commands/options listed ok. so i just have to type this when i'm in the specific path of the cpuminer. where do i know what build I'm using? README.txt |
|
|
|
Mining noob here look for some help trouble shooting. I'm trying to mine zoin. I believe i have everything set up correctly but cpuminer keeps closing out on me. I'm on an older laptop running windows 8.1, intel i5-4210u cpu which does support avx2. Here's my .bat "cpuminer-avx2.exe -a lyra2zoin -o stratum+tcp://zoin.netabuse.net:3000 -u mynamemyworker -p mypassword -t1" I double click the .bat, the cmd window opens and i start hashing. I check my worker on the pools page and it shows it running. Shares submitted, shares accepted, running fine for 4-5 minutes then the cmd window closes and no more hashing. I've given permission to the av, i'm only running 1 thread so cpu usage/temperature isn't an issue. It just quits. Like i said i'm a total noob when it comes to mining so maybe there's something obvious i'm overlooking, but as far as i can tell it's all set up right.  If the miner is reporting shares accepted and the pool confirms it you seem to be doing (almost) everything right and likely have a problem with your PC causing the miner to crash. Almost because you should put a pause in your bat so the window stays open and you can see any error messages that may have been displayed. |
|
|
|
Hi can someone help me?? why i get this error??  im using ./build.sh and doing the same thing like 1 month ago without any issue, but somehow now i got that error , idk what is wrong thank you for your help guys Your CPU has AVX512 I presume. Read back a few posts, you can get more details here... https://github.com/JayDDee/cpuminer-opt/issues/116 |
|
|
|
|
If you'r e the lucky owner of a CPU with AVX512, maybe you aren't so lucky because
v3.8.4 fails to compile. I included some experimental AVX512 code but was unable
to test compile it, and sure enough, it won't compile.
Until a new release is available you can remove the erroneous backslash in avxdefs.h line 1244
or remove the entire AVX512 block of code.
|
|
|
|
work well on cpuminer-opt-3.8.4
Yes, according to my calculations, the hashrate is about 10-12% higher. Thank you very much to the developer optminer for the prompt update of the product You're welcome. Thanks for reporting your results. |
|
|
|
Can you take a look at the Argon2d algo and add it? There seem to be 2 variants, for 2 coins (wonder whats the difference between the algos, but the hashrate is about 3 times faster on Credits coin than on Dynamic) https://github.com/BeastPool/cpuminer-argon2dIt looks like the same algo but with different parameters. It should be a straightforward implementation because it was forked from cpuminer-opt. But I'll probably rework it to reduce code duplication/triplication. |
|
|
|
cpuminer-opt-3.8.4Added yescryptr32 algo for WAVI coin. Added URL to API data. Improved detection of __int128 support (Linux only) Compile support for CPUs without SSSE3 (no binary support) https://github.com/JayDDee/cpuminer-opt/releases/tag/v3.8.4Additional notes: This release may fix some recently reported compile issues related to unsupported SIMD instructions and lack of support for __int128. This may help some older AMD CPUs and Linux distributions that previously failed to compile. There may be some other unknown underlying issues so YMMV. In particular Centos 6 requires a newer version of gmp that is avalable for that release. There is no Windows binary for CPUs with SSE2 but not SSSE3. cpuminer-sse2 is built for Intel Core2 which includes SSSE3. However, I can be bribed.
|
|
|
|
|
Must be a slow day. The way ROI is being used is to demark the point when the investment is paid and
the product starts earning a Return On the Investment. Is the distinction really big enough to deserve a rant?
How about people who "could care less" instead of "couldn't care less? It only takes a subtle grasp of
the english language to figure that one out but so many get it wrong.
Or how about people who think "myself" is synonymous with "me"?
Misusing ROI just doesn't rate.
|
|
|
|
Spam! Spam! Spam! Spam!
No Spam here please. |
|
|
|
|
Forget about turboboost, it's just a slick way to make something bad sound good.
With all cores running the CPU produces too much heat at full speed so the clock is throtled back.
Your CPU has 2 physical cores but can run 4 logical threads. HT allows 2 threads to run on
a single core but they share resources. Two threads on one core using HT is not as fast as
each thread on a dedicated core. That's why you saw the numbers you did.
|
|
|
|
I belive it's calling TurboBoost... Maybe the YescryptR16 algo it's very sensitive to the cpu clock frequency and when only one core worked TurboBoost run this core on maximum clock frequency and slow down clock frequency when more core utilized.
Please don't speculate. YescryptR16 is less sensitve to CPU clock because it's memory hard. Adding more CPU threads when memory bandwidth is saturated just causes CPU stalls. You also don't seem to understand turboboost either. Turboboost only varies the CPU clock by about 15% and all cores run at the same freq. Neither is the case with the data provided. |
|
|
|
|
It's called Hyperthreading, google it.
|
|
|
|
Hi joblo, thx for this. It seems there is a bug in the Code Freshly downloaded the sources on another machine and gone through configure.ac My version shows as follows, which explains the problem we got. (I manually edited and corrected it, and it works fine now) Another thing is that I have to install libgmpv4-dev to meet another requirement, but after that cpuminer binary got compiled fine  if test "x$OS" = "xWindows_NT" ; then # MinGW AC_CHECK_LIB([pthread], [pthread_create], PTHREAD_LIBS="-lpthreadGC2",[]) else AC_CHECK_LIB([pthread], [pthread_create], PTHREAD_LIBS="-lpthreadGC2",[]) fi Where did you get the source code? That's not what's on my github. |
|
|
|
Neoscrypt crashing on i7-3610QM too.
mine too, i7-7700HQ Same problem... any suggestions for neoscrypt??? cpuminer-opt has the same problem since moving to Ubuntu-16.04/MinGW64 build environment for Windows binaries but the legacy version still works. |
|
|
|
|
Show Posts
