CPUMINER-OPT-3.3 I tried the Windows binary. On my E5640. I'm pretty sure it has AES_NI, but the software doesn't like it. Says "rebuild with -march=native"
Then I tried the other two binaries, they won't even start at all, just crashes. In my E5-2650 doesn't work, Says "rebuild with -march=native". But in my other rigs (I5-3550 , i7 2600) see an speed increment of 10 to 20 percent  It does work. I know it's confusing but that message is just a tip, not an error so the miner will continue to start mining. The message should be ignored for first generation AES enabled CPUs. Edit: I'm tracking this issue in my thread and have posted an update: https://bitcointalk.org/index.php?topic=1326803.msg15045809#msg15045809Fixed in v3.3.2, see link in sig. |
|
|
|
Wow! Thats really cool, I hope it brings some more attention to HOdlcoin!  I hope the this hashrate increase is not nicehash work. The nicehash minimum limit is way to high for a slow hashing algo like HOdl. This will be disruptive but ultimately good for the long term prospects of the coin and the algo. Currently Nicehash is mining at around 700KH which more than doubles the current network hash rate. At the moment it's still a mess but when things settle down and if Nicehash realizes the screw up with the minimum limit things should improve. |
|
|
|
Thanks, OP is updated with the link. |
|
|
|
Thanks, you may have noticed that I implemented unordered_map as you suggested a while back. |
|
|
|
cpuminer-opt v3.3.2 is released. https://drive.google.com/open?id=0B0lVSGQYLJIZczlQLV93RnUzV1U- Fixed low difficulty shares mining hodl - Changed minimum CPU requirement for AES optimized mining, now requires a CPU with AES and AVX1. This excludes first generation AES CPUs like Nehalem and similar AMD. These CPUs can still mine using SSE2 optimizations. - Updated build.sh and README.md Windows users may want to wait until Cryptomining Blog build binaries for v3.2.2. A download link will be posted when it becomes available. |
|
|
|
cpuminer-opt v3.3.2 is released. https://drive.google.com/open?id=0B0lVSGQYLJIZczlQLV93RnUzV1U- Fixed low difficulty shares mining hodl - Changed minimum CPU requirement for AES optimized mining, now requires a CPU with AES and AVX1. This excludes first generation AES CPUs like Nehalem and similar AMD. These CPUs can still mine using SSE2 optimizations. - Updated build.sh and README.md Windows users who don't like to compile their own can wait for Cryptomining Blog to provide precompiled Windows executables. I will post the link in the OP when it becomes available. |
|
|
|
AMD FX7600 (SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4A, XOP, AVX, FMA, FMA4, AES)
cpuminer v3.3.1 - 15-17 khash
Wolf0 - 33-35 khash
I will definitely follow up for v3.3.3, need to get the other fixes out first. Thanks. This is precisely the issue I'm trying to solve, where to find the best miner for a specific algo. I want cpuminer-opt to the the one stop shop for all CPU mining. Thanks for that tip. |
|
|
|
joblo, today I decided to test the built-in Magi XMG M7M algo and for my surprise the miner says the algo in not AES-NI. Magi has AES-NI support in cpuminer from atleast an year ago, by Wolf0. Here is the github link: https://github.com/wolf9466/wolf-m7m-cpuminerHow does your performance compare? |
|
|
|
When using cpuminer-opt with first generation AES enabled CPUs (Nehalem) a message may be displayed suggesting rebuilding could produce better performance. This message is incorrect and is the result in a discrepency between the strict definition of AES and the miner's AES optimized code. The message should be ignored and the miner will hash at the best rate possible. Follow updates in my thread. https://bitcointalk.org/index.php?topic=1326803.msg15045809#msg15045809 |
|
|
|
I have an update on a couple of issues that have existed for a while but are now better understood. 1. The definition of AES The miner's AES optimized code also includes AVX instructions that are not included in the first generation of AES CPUs, however the miner's CPU capabilities check uses the pure drefinition of AES (without AVX). This results in the following message being displayed on startup on CPUs with AES but not AVX: Rebuild with "-march=native" for better performance. This is not an error, the miner will work, but not at AES performance levels. This message should be ignored for Nehalem series CPUs and similar AMD. These CPUs can't mine using the miner's AES & AVX code. I other words the miner will mine at the best rate possible for that CPU in spite of the suggestion otherwise. The minimum requirement for AES mining will be changed to Sandybridge to conform with the actual limitations of the AES mining code and the capabilities check will be changed to also check for AVX support in the CPU. Update: I have changed the terminology to be more precise. It will now be called AES-AVX1 to more clearly exclude CPUs that have AES but not AVX1. Also the rebuild warning will not be displayed for those CPUs. Documentation will also be updated. 2. Low difficulty shares mining hodl Since v3.2 hodl mining has started producing low difficulty shares at a rate of 1 or 2%. Prior to that there were virtually no rejects. I may have found the problem and am testing a fix. At this point the first 170 shares have been valid. Update: I am optimistic the fix works. The first test run produced no low diff shares in 268 accepts 3. Hodl support on Windows for non AES-AVX1 CPUs. This is a more challenging issue and will not be fixed in the next release. I will build a new release when these 2 issues are solved. In the meantime users should ignore the warning to rebuild for faster performance and hodl users with Linux can use v3.1.18 if the rejects are unacceptible. Hodl miners n Windows will have to live with the rejects until the fix is released. Note to Cryptimining Blog: You might want to drop the Westmere build for the next release as it seems to have no practical use. A core2 build would expand the range of supported CPUs and would also work on Westmere. Update:It appears both issues are fixed. The fix to hodl was more of a trial-and-error proceess so I want to test longer to raise the confidence level. I don't know why it broke and I don't know why the fix works. The AES issue, on the other hand, is now well understood so I am confident in the fix although I don't have the resources to test it properly. An interesting contrast in processes. V3.3.2 coming in a matter of hours. |
|
|
|
Not very good at compiling things or Linux, but will try following the instructions.
That won't be necessary, I misunderstood what you wrote and thought you already had a Linux machine and a Windows machine. I have done my own testing and am convinced the rejects problem was introduced in v3.2. Now I just have to fix it.. Edit: I am tracking this issue in my thread and have posted an update. https://bitcointalk.org/index.php?topic=1326803.msg15045809#msg15045809 |
|
|
|
CPUMINER-OPT-3.3 I tried the Windows binary. On my E5640. I'm pretty sure it has AES_NI, but the software doesn't like it. Says "rebuild with -march=native"
Then I tried the other two binaries, they won't even start at all, just crashes. In my E5-2650 doesn't work, Says "rebuild with -march=native". But in my other rigs (I5-3550 , i7 2600) see an speed increment of 10 to 20 percent It does work. I know it's confusing but that message is just a tip, not an error so the miner will continue to start mining. The message should be ignored for first generation AES enabled CPUs. Edit: I'm tracking this issue in my thread and have posted an update: https://bitcointalk.org/index.php?topic=1326803.msg15045809#msg15045809 |
|
|
|
It looks like you are trying to solo mine the wallet. I have not been able to test this and didn't know if it worked.
Thanks for being my tester. I'll look into it to see what I can do.
Edit: can you provide the output without -q?
********** cpuminer-opt 3.3 ***********
A CPU miner with multi algo support and optimized for CPUs
with AES_NI extension.
BTC donation address: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT
Forked from TPruvot's cpuminer-multi with credits
to Lucas Jones, elmad, palmd, djm34, pooler, ig0tik3d,
Wolf0 and Jeff Garzik.
Checking CPU capatibility...
Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz
CPU arch supports AES_NI...YES.
SW built for AES_NI........YES.
Algo supports AES_NI.......YES.
Start mining with AES_NI optimizations...
[2016-06-01 13:25:04] Binding thread 0 to cpu 0 (mask 1)
[2016-06-01 13:25:04] Binding thread 1 to cpu 1 (mask 2)
[2016-06-01 13:25:04] Binding thread 2 to cpu 2 (mask 4)
[2016-06-01 13:25:04] 4 miner threads started, using 'hodl' algorithm.
[2016-06-01 13:25:04] Binding thread 3 to cpu 3 (mask 8)
[2016-06-01 13:25:05] Current block is 64418
[2016-06-01 13:25:05] Unrecognized block version: 4
[2016-06-01 13:25:05] Long-polling on http://192.168.1.8:11989
[2016-06-01 13:25:06] json_rpc_call failed, retry after 10 seconds
[2016-06-01 13:25:07] JSON-RPC call failed: Method not found
[2016-06-01 13:25:17] Current block is 64418
[2016-06-01 13:25:17] Unrecognized block version: 4
[2016-06-01 13:25:18] json_rpc_call failed, retry after 10 seconds
[2016-06-01 13:25:29] Current block is 64418
[2016-06-01 13:25:29] Unrecognized block version: 4
[2016-06-01 13:25:30] json_rpc_call failed, retry after 10 seconds The output with Debug option without -q. Also regarding the low difficulty share issue, it also occurred to me before. i5-2500K on ubuntu 14.04 x64. Edit: Output with protocol dump (-P)[2016-06-01 13:44:32] JSON protocol request:
{"method": "getblocktemplate", "params": [{"capabilities": ["coinbasetxn", "coinbasevalue", "longpoll", "workid"]}], "id":0}
* Rebuilt URL to: http://192.168.1.8:11989/
* Found bundle for host 192.168.1.8: 0x7fead8013d00
* Re-using existing connection! (#0) with host 192.168.1.8
* Connected to 192.168.1.8 (192.168.1.8) port 11989 (#0)
* Server auth using Basic with user 'hodl'
> POST / HTTP/1.1
Authorization: Basic Zm9ydHJlc3Nob2RsOjdBMjV2TjlGbGljSnhwcVk=
Host: 192.168.1.8:11989
Accept: */*
Accept-Encoding: deflate, gzip
Content-Type: application/json
Content-Length: 126
User-Agent: cpuminer-opt/3.3
X-Mining-Extensions: longpoll reject-reason
Expect: 100-continue
* Done waiting for 100-continue
< HTTP/1.1 200 OK
< Date: Wed, 01 Jun 2016 05:44:36 +0000
< Connection: keep-alive
< Content-Length: 583
< Content-Type: application/json
* Server bitcoin-json-rpc/v1.0.0 is not blacklisted
< Server: bitcoin-json-rpc/v1.0.0
<
* Connection #0 to host 192.168.1.8 left intact
[2016-06-01 13:44:33] JSON protocol response:
{
"error": null,
"result": {
"transactions": [],
"sizelimit": 1000000,
"previousblockhash": "00000018958d65d48ce5f3ef329ab9b8a49eb9e6ad8c56c67cd4518b992841f5",
"mutable": [
"time",
"transactions",
"prevblock"
],
"longpollid": "00000018958d65d48ce5f3ef329ab9b8a49eb9e6ad8c56c67cd4518b992841f5331",
"version": 4,
"mintime": 1464759040,
"height": 64428,
"capabilities": [
"proposal"
],
"coinbaseaux": {
"flags": ""
},
"sigoplimit": 20000,
"coinbasevalue": 5000000000,
"target": "0000004556e10000000000000000000000000000000000000000000000000000",
"curtime": 1464759870,
"noncerange": "00000000ffffffff",
"bits": "1d4556e1"
},
"id": 0
}
[2016-06-01 13:44:33] Current block is 64428
[2016-06-01 13:44:33] Unrecognized block version: 4
[2016-06-01 13:44:33] JSON protocol request:
{"method": "getmininginfo", "params": [], "id":8}
* Rebuilt URL to: http://192.168.1.8:11989/
* Found bundle for host 192.168.1.8: 0x7fead8013d00
* Re-using existing connection! (#0) with host 192.168.1.8
* Connected to 192.168.1.8 (192.168.1.8) port 11989 (#0)
* Server auth using Basic with user 'hodl'
> POST / HTTP/1.1
Authorization: Basic Zm9ydHJlc3Nob2RsOjdBMjV2TjlGbGljSnhwcVk=
Host: 192.168.1.8:11989
Accept: */*
Accept-Encoding: deflate, gzip
Content-Type: application/json
Content-Length: 51
User-Agent: cpuminer-opt/3.3
X-Mining-Extensions: longpoll reject-reason
Expect: 100-continue
* Done waiting for 100-continue
< HTTP/1.1 200 OK
< Date: Wed, 01 Jun 2016 05:44:37 +0000
< Connection: keep-alive
< Content-Length: 246
< Content-Type: application/json
* Server bitcoin-json-rpc/v1.0.0 is not blacklisted
< Server: bitcoin-json-rpc/v1.0.0
<
* Connection #0 to host 192.168.1.8 left intact
[2016-06-01 13:44:34] JSON protocol response:
{
"error": null,
"result": {
"difficulty": 0.01356886,
"errors": "",
"currentblocksize": 1000,
"currentblocktx": 0,
"pooledtx": 0,
"blocks": 64427,
"genproclimit": -1,
"networkhashps": 402233,
"generate": false,
"testnet": false,
"chain": "main",
"hashespersec": 0
},
"id": 8
}
[2016-06-01 13:44:34] json_rpc_call failed, retry after 10 seconds
I think this is your problem. [2016-06-01 13:25:05] Unrecognized block version: 4 I think there is a new block version for hodl and believe there is a bounty for someone to write the code to support it. Until then you'll have to mine in a pool. |
|
|
|
NICE! I guess it's important to use the right version, or?? cpuminer-sandybridge-ivybridge.exe cpuminer-haswell-broadwell.exe cpuminer-amd.exe cpuminer.exe Yes it is. Or?? it does not work. ...or works slower. Haswell is for core i 4000 series, sandy for Core i 2000. If your not sure give them all a try and use the one that works best. If the algo is not AES_NI optimized you won't notice much difference. I tried the Windows binary. On my E5640. I'm pretty sure it has AES_NI, but the software doesn't like it. Says "rebuild with -march=native" Then I tried the other two binaries, they won't even start at all, just crashes. Oh well. the e5640 is technically a Westmere cpu. One generation prior to sandy bridge. I know that the intel site says that it has AES-NI support, but it has a slightly different instruction set which could be the problem. It is missing AVX. I don't know enough about exactly how the miner works to say whether or not that is why you are facing your issue though. If that is indeed the problem, then you would need a compiled version of the miner without AVX support. Otherwise your guess as to why it is not working would be as good as mine. I think you got it. The miner's AES code contains AVX instructions that aren't supported on first generation AES CPUs. So the miner's definition of AES does not agree with the compiler's version or Intel's version. To get the benefit of the AES optimizations in the miner the CPU needs AVX as well as AES. That means Ssandybridge is the minimum architecture required for top performance. I will tighten up the documentation to include AVX as a requirement for AES performance and also try to improve the CPU capabilities report by checking for AVX before saying YES to AES. In your case Westmere is what you have amd you don't have AVX so you don't get the AES miner optimizations. The error is the miner saying you can. |
|
|
|
Windows 7 on Intel i7 5820K does produce some low diff shares using the cpuminer.exe from the link, using the hashwell exe is much better only once in a while a low diff share.
What's your reject rate? I think it's a problem introduced in v3.2. Give v3.1.18 a try on Linux to see if you get any rejects. I expect zero after several hundred accepts. Here's the link in case you don't have it. https://drive.google.com/file/d/0B0lVSGQYLJIZc1R0ajdpQ2R3NzQ/view?usp=sharing |
|
|
|
|
I think the low difficulty rejects in hodl were left over from the mess in v3.2. v3.1.18 seems to be clean.
If anyone has evidence to the contrary please let me know.
|
|
|
|
Updated for 3.3.1 with x11evo, that was quick. where is 3.3.1？ They didn't update the link, only the download. |
|
|
|
CRYPTOMINING-BLOG ARTICLE--
Congratulations! Your software was featured on the blog! I tried it, and found that it does at least 100% better than CPU-Multi on Yescript! I do see that Yescript software algo is not yet optimized. I hope that you will find the time to do that bit of code soon.
Running on i7 2600 here, I get about 2kh/s for Yescript. With CPU-Multi, it was closer to 800H/s.
Thank you! --scryptr
Unfortunately I can't optimize scrypt algos, only the chained algos like x11, quark etc. You get 2x multi, not sure why, I just took whatever was there and it worked. Never compared performance. Pretty cool to be featured on CMB, they already updated to v3.3.1. |
|
|
|
Updated for 3.3.1 with x11evo, that was quick. |
|
|
|
cpuminer-opt now supports Revolver coin with x11evo algo. x11evo has been optimized 109% on CPUs with AES_NI and 42% for CPUs with only SSE2. Windows support was also recently added to cpuminer-opt using mingw_64 and msys. See README.md for instructions on how to install mingw_w64 and msys. No precompile Windows binaries are provided, must compile from source. There is a small undocumented change in the build procedure for this release. A new option has been added to the configure command. This procedure works for both Linux and Windows. ./autogen.sh
./configure CFLAGS="-O3 -march=native" CXXFLAGS="$CFLAGS -std=gnu++11" --with-curl
make
Download it here or visit the link in my sig for more details. https://drive.google.com/file/d/0B0lVSGQYLJIZWlJaOWtPZ2tESmM/view?usp=sharingcompiled and working at ubuntu 15.10 and ubuntu 14.04 lts But i show error... Windows not supporded? What error? Cryptomining blog just posted Windows binaries of cpuminer-opt v3.3, which unfortunately doesn't have x11evo. You'll have to wait for them to build v3.3.1. |
|
|
|
|
Show Posts
