VS
Windows:
Topic: Wolf's XMR/BCN/DSH CPUMiner - 2x speed compared to LucasJones' - NEW 06/20/2014 - page 33. (Read 547096 times)
June 10, 2014, 01:53:35 PM
Linux:
VS
Windows:
VS
Windows:
June 08, 2014, 11:20:23 AM
Oh, and by the way - this affects non-AES-NI, too. So now my implementation of non-AES-NI is a bit faster than Lucas'. 
June 10, 2014, 11:19:24 AM
I'd give all kinds of naughty favours for win64 binaries. With AES preferred for i7-4930k 
*nudge* *nudge*
*nudge* *nudge*I'll compile them soon.
I tried out the 9th of June on Win 7 64-bit and it gives a 3-4% reduction on an i5-2500k but maybe a 1% increase on an i7-4930k.
Can hugepages even be done on Windows?
What speed you get with an i7 4930k?
There is a difference with a i7 3930K?
I'm curious to find out the speed of the new processor i7 4970k!
About 400/500 h/s? 
It's only about 250 my dear friend, I'm using Windows 7, the real speeds are in Linux.
thanks!
And what's the difference?
June 08, 2014, 11:14:39 AM
Bump for update.
Are you making a compiled version for windows? I don't see it in the post.
No need. Windows doesn't give me such control over memory management - so this optimization can't be done for Windows.
June 10, 2014, 10:34:35 AM
What are the speed differences in Linux from Windows? Might try Ubuntu again if the speed increase is worth it.
June 10, 2014, 08:20:51 AM
I'd give all kinds of naughty favours for win64 binaries. With AES preferred for i7-4930k *nudge* *nudge*
*nudge* *nudge*I'll compile them soon.
I tried out the 9th of June on Win 7 64-bit and it gives a 3-4% reduction on an i5-2500k but maybe a 1% increase on an i7-4930k.
Can hugepages even be done on Windows?
What speed you get with an i7 4930k?
There is a difference with a i7 3930K?
I'm curious to find out the speed of the new processor i7 4970k!
About 400/500 h/s? It's only about 250 my dear friend, I'm using Windows 7, the real speeds are in Linux.
June 10, 2014, 07:01:34 AM
I'd give all kinds of naughty favours for win64 binaries. With AES preferred for i7-4930k *nudge* *nudge*
*nudge* *nudge*I'll compile them soon.
I tried out the 9th of June on Win 7 64-bit and it gives a 3-4% reduction on an i5-2500k but maybe a 1% increase on an i7-4930k.
Can hugepages even be done on Windows?
What speed you get with an i7 4930k?
There is a difference with a i7 3930K?
I'm curious to find out the speed of the new processor i7 4970k!
About 400/500 h/s? June 08, 2014, 09:50:05 AM
Bump for update.
June 09, 2014, 11:59:01 PM
I'm using GCC 4.8.2 and LD 2.23.52 compiled.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
What distro? It probably doesn't include hugepages.
Nice!
Avast antivirus just started registering this minerd.exe as dangerous
Avast antivirus just started registering this minerd.exe as dangerous
AV companies do that to boost the number of malware they "found" so they look good on paper. The miner is used by malware to mine on people's computers, but instead of just tagging the actual malware, they add the miner as well.
CentOS6.3 x64
Second report I've gotten of CentOS not working... I think the first was fixed by updating glibc.
many thanks!
I will try...
June 09, 2014, 11:49:28 PM
Not 100% sure. Since most Win users can't compile their own binaries, I'd have to revert every commit, one by one, and build a Win64 binary for each >.<
I always thought that the binary one for a particular type of Windows. Why did they each compile? June 08, 2014, 03:59:05 AM
Hi
i need a some help here, i manage to compile, but when i run ./minerd i got this
but i have openssl installed and this archives at /usr/lib and /usr/lib64
My enviroment is centos 6.4
Tnks
i need a some help here, i manage to compile, but when i run ./minerd i got this
Code:
./minerd: error while loading shared libraries: libcrypto.so.1.0.0: cannot open
shared object file: No such file or directory
Code:
libcrypto.so libcrypto.so.10 libcrypto.so.1.0.1e
My enviroment is centos 6.4
Tnks
Wait... so you compiled against libcrypto.so.1.0.0, but when you ran it, you had only libcrypto.so.1.0.1e?
You must have upgraded OpenSSL in the middle. Rebuild the miner.
June 09, 2014, 10:47:01 PM
I'm using GCC 4.8.2 and LD 2.23.52 compiled.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
What distro? It probably doesn't include hugepages.
Nice!
Avast antivirus just started registering this minerd.exe as dangerous
Avast antivirus just started registering this minerd.exe as dangerous
AV companies do that to boost the number of malware they "found" so they look good on paper. The miner is used by malware to mine on people's computers, but instead of just tagging the actual malware, they add the miner as well.
CentOS6.3 x64
June 09, 2014, 10:25:59 PM
Nice!
Avast antivirus just started registering this minerd.exe as dangerous
Avast antivirus just started registering this minerd.exe as dangerous
June 07, 2014, 04:26:45 PM
AMD1055T won't run, you can solve? please .
No. Doesn't have AES-NI.
June 09, 2014, 10:25:14 PM
I'm using GCC 4.8.2 and LD 2.23.52 compiled.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
when I make , I got a error:
cpu-miner.c: In function 'miner_thread':
cpu-miner.c:1065:88: error: 'MADV_HUGEPAGE' undeclared (first use in this function)
madvise(persistentctx, sizeof(struct cryptonight_ctx), MADV_RANDOM | MADV_WILLNEED | MADV_HUGEPAGE);
who can help me ?
thanks.
June 09, 2014, 09:05:00 PM
Yeah hugepages can be done - the YAM miners by yvg1900 (for MMC, PTS, etc.) use hugepages and are the fastest CPU miners for the coins it mines.
June 07, 2014, 10:00:08 AM
Do you have plans to add merged mining (for FCN) ?
How much would it be worth if I got it almost as fast as my cpuminer?
June 09, 2014, 08:59:47 PM
I'd give all kinds of naughty favours for win64 binaries. With AES preferred for i7-4930k *nudge* *nudge*
*nudge* *nudge*I'll compile them soon.
I tried out the 9th of June on Win 7 64-bit and it gives a 3-4% reduction on an i5-2500k but maybe a 1% increase on an i7-4930k.
Can hugepages even be done on Windows?
June 09, 2014, 07:27:46 PM
All of my tests were done on an AMD FX-8320 @ 4.2GHz, 8GB 1833, Win7x64, ran @ -t 7. The average is from when lightly using the computer as normal and the peak, where it spends most of its time, when left alone. Now if only I can figure out how to compile it on windows myself lol.
5/29
Peak – 322
Average – 270
5/30
Peak – 312
Average – 265
Experimental
Peak – 328
Average – 275
6/8
Peak – 331
Average – 315
5/29
Peak – 322
Average – 270
5/30
Peak – 312
Average – 265
Experimental
Peak – 328
Average – 275
6/8
Peak – 331
Average – 315
June 09, 2014, 05:11:37 PM
hi wolfie, ty so much for your miner and flurry! nor fury 
im testing with i74702MQ and yeah prolly drop -5% or zero different than 5-30-2014
and flurry! nor fury im testing with i74702MQ and yeah prolly drop -5% or zero different than 5-30-2014
Jump to: