Topic: [ANN][DASH] Dash (dash.org) | First Self-Funding Self-Governing Crypto Currency - page 6918. (Read 9723475 times)
February 14, 2014, 06:53:53 PM
Did anyone manage to compile it on a Digital Ocean droplet? Will there be an improvement?
February 14, 2014, 06:39:05 PM
There is no windows delivery yet?
For the new version of minerd
For the new version of minerd
February 14, 2014, 06:36:46 PM
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
From 166 kh/s to 184 kh/s.
So, it is about 11 ~ 12% improvement.
I guess you have generic kernel, right?
With the i5-2400 that I tried, the boost was about 21%
What cpu? And do u compiled with -march=native? Just to know so we can make a statistic. Because, for sure, improvement depends by AVX and AES performances of the processor.
For now this version boost just the echo algo, if I'll have time I want to try to boost some other from the 11 algos
.CPU: Atom C2758
AES-NI: Yes
AVX: No
Generic kernel?
Not sure what that means. I'm using a standard Ubuntu 13.10:
$ uname -a
Linux d510 3.11.0-15-generic #25-Ubuntu SMP Thu Jan 30 17:22:01 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
Ok, this is right. I tried before to compile with AES-NI without AVX, and I had a 14% improvement. Adding AVX I achieved 21%.
February 14, 2014, 06:14:22 PM
Thanks CHAOSiTEC. I'll be waiting to come back there.
February 14, 2014, 06:12:48 PM
play it safe....create 6 completely unique imaginary people.
Warning: VPS crypto mining may be dangerous for your mental health. (This is a joke
)I am mining on the lotterymining pool right now and I suppose there is something off with it. The block finding rate is very unstable even though the hashrate doesn't change that much. It differs from several blocks per hour to a couple blocks a day. I think there might be a problem with pool software or network connection.
As far as I understand very long time between blocks doesn't lead to larger rewards but only to wasted hashpower.
Yep, that's why I moved out of there temporarily.
i will restart the server tomorrow when im in harbour, i dont dare to do so now, since i dont know when i have internet, and when i dont, so i might end up restarting it, but unable to start up stratum when its done rebooting..
February 14, 2014, 06:01:43 PM
I haven't forgot about tip as well. Once I run all my workers it'll go. I'll report on how it goes on Xeons v2 vs Xeons v3.
February 14, 2014, 06:00:01 PM
New CPU Miner!!
I have created a new cpu-miner 64bit and I achieved about 22% more khs/sec .
This version is based on cpu-miner-1.2c by ig0tik3d, and plus uses AES-NI/AVX instructions from an other library, so it's usable with this list of Intel Processors:
http://ark.intel.com/search/advanced/?AESTech=true
and with these series by AMD:
AMD Bulldozer-based processors
AMD Piledriver-based processors.
AMD Jaguar-based processors.
AMD Kaveri-based processors.
I'll release it today, but it's in alpha stage and needs testing because I tried it just with a Sandy Bridge i5-2400 from a friend.
I have not cpus to mine, so bounties for my time will are very appreciated.
Let me know if you are interested.
XCO: XxSJMTPKqRzADNe7Y3pazXF8jPV95EvDnS
BTC: 1Kwewq26emgePgvJFwcVCgzQTnxwdF1Jcb
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
Awesome,
Gosh, I am getting old, was reading the install file
Well, looks like it's running a little bit better
from 19kh/s to 23 kh/s Still, accepteds very few and far between, wish I knew why 
Tip on it's way (sorry, I'm not a big holder, but hope it helps )From 19 to 23 it's a 21% improvement
. What cpu do you use?Thanks for the tip, it's appreciated
I'm sending a little more, I have one instance that went down to 15 kh/s Must be the way they put the cpu's together?? I'm using Amazon, per the instructions on page one. I think I'm going to open a few more, it's inexpensive and I think worth it?? Just for the weekend
I am assuming the XCO address is your darkcoin address? Better be, I sent twice, LOL!!! Makes me wonder, if you send to a non-existent account, will it reject and come back?
February 14, 2014, 05:49:41 PM
26% Improved. Thank !!!
February 14, 2014, 05:44:56 PM
New CPU Miner!!
I have created a new cpu-miner 64bit and I achieved about 22% more khs/sec .
This version is based on cpu-miner-1.2c by ig0tik3d, and plus uses AES-NI/AVX instructions from an other library, so it's usable with this list of Intel Processors:
http://ark.intel.com/search/advanced/?AESTech=true
and with these series by AMD:
AMD Bulldozer-based processors
AMD Piledriver-based processors.
AMD Jaguar-based processors.
AMD Kaveri-based processors.
I'll release it today, but it's in alpha stage and needs testing because I tried it just with a Sandy Bridge i5-2400 from a friend.
I have not cpus to mine, so bounties for my time will are very appreciated.
Let me know if you are interested.
XCO: XxSJMTPKqRzADNe7Y3pazXF8jPV95EvDnS
BTC: 1Kwewq26emgePgvJFwcVCgzQTnxwdF1Jcb
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
Awesome,
Gosh, I am getting old, was reading the install file
Well, looks like it's running a little bit better
from 19kh/s to 23 kh/s Still, accepteds very few and far between, wish I knew why Tip on it's way (sorry, I'm not a big holder, but hope it helps )From 19 to 23 it's a 21% improvement
. What cpu do you use?Thanks for the tip, it's appreciated
February 14, 2014, 05:02:41 PM
New CPU Miner!!
I have created a new cpu-miner 64bit and I achieved about 22% more khs/sec .
This version is based on cpu-miner-1.2c by ig0tik3d, and plus uses AES-NI/AVX instructions from an other library, so it's usable with this list of Intel Processors:
http://ark.intel.com/search/advanced/?AESTech=true
and with these series by AMD:
AMD Bulldozer-based processors
AMD Piledriver-based processors.
AMD Jaguar-based processors.
AMD Kaveri-based processors.
I'll release it today, but it's in alpha stage and needs testing because I tried it just with a Sandy Bridge i5-2400 from a friend.
I have not cpus to mine, so bounties for my time will are very appreciated.
Let me know if you are interested.
XCO: XxSJMTPKqRzADNe7Y3pazXF8jPV95EvDnS
BTC: 1Kwewq26emgePgvJFwcVCgzQTnxwdF1Jcb
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
Awesome,
Gosh, I am getting old, was reading the install file
Well, looks like it's running a little bit better
from 19kh/s to 23 kh/s Still, accepteds very few and far between, wish I knew why Tip on it's way (sorry, I'm not a big holder, but hope it helps ) February 14, 2014, 04:24:48 PM
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
From 166 kh/s to 184 kh/s.
So, it is about 11 ~ 12% improvement.
I guess you have generic kernel, right?
With the i5-2400 that I tried, the boost was about 21%
What cpu? And do u compiled with -march=native? Just to know so we can make a statistic. Because, for sure, improvement depends by AVX and AES performances of the processor.
For now this version boost just the echo algo, if I'll have time I want to try to boost some other from the 11 algos
. February 14, 2014, 04:17:49 PM
Source code released
[/size]Source:
https://github.com/elmad/darkcoin-cpuminer-1.3-avx-aes
#Read the README, compiles with march=native, only on CPUs supporting AES-NI instructions (it includes AVX)
Test it, and let me know
From 166 kh/s to 184 kh/s.
So, it is about 11 ~ 12% improvement.
I guess you have generic kernel, right?
February 14, 2014, 04:13:19 PM
ig0tik3d's version runs without problem. I don't know what the underlying hypervisor is.
Ok, because you have sse2. I was curious just about the message that said "not supporting x86_64".
But sure there is the problem with missing AES-NI and AVX.
February 14, 2014, 04:11:06 PM
ig0tik3d's version runs without problem. I don't know what the underlying hypervisor is.
February 14, 2014, 04:09:36 PM
Code:
flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 ht syscall nx lm nopl pni cx16 hypervisor
As u can see, here misses both avx, aes-ni, and ssse3. 1.2c version by ig0tik3d compiles with this virtual machine?
Is it XEN the host of this virtual machine? I tried just with vmare, and there, with a processor that support VT instructions, the CPU is fully exposed, while there I see that cpuinfo reports a Qemu cpu.
February 14, 2014, 04:06:48 PM
I guess my virtual machine does not use hardware virtualization. Unfortunately the hypervisor there is out of my control.
February 14, 2014, 04:03:15 PM
Code:
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 13
model name : QEMU Virtual CPU version (cpu64-rhel6)
stepping : 3
microcode : 0x1
cpu MHz : 3500.022
cache size : 4096 KB
physical id : 0
siblings : 8
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 4
wp : yes
flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 ht syscall nx lm nopl pni cx16 hypervisor lahf_lm
bogomips : 7000.04
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
vendor_id : GenuineIntel
cpu family : 6
model : 13
model name : QEMU Virtual CPU version (cpu64-rhel6)
stepping : 3
microcode : 0x1
cpu MHz : 3500.022
cache size : 4096 KB
physical id : 0
siblings : 8
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 4
wp : yes
flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 ht syscall nx lm nopl pni cx16 hypervisor lahf_lm
bogomips : 7000.04
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
February 14, 2014, 04:01:46 PM
Doesn't compile under KVM. Sure, there's limited CPU exposure.
Code:
error: CPU you selected does not support x86-64 instruction set
It is only 64 bit for now. Your OS in this case is 32bit? I tested with KVM on a virtual Machine, and it's ok. But guest OS needs to be 64 bit for now.
No, it's 64-bit.
Doesn't compile on Core 2 Quad either.
Code:
gcc -std=gnu99 -DHAVE_CONFIG_H -I. -pthread -fno-strict-aliasing -O3 -march=native -MT minerd-hash.o -MD -MP -MF .deps/minerd-hash.Tpo -c -o minerd-hash.o `test -f 'x5/echo512/ccalik/aesni/hash.c' || echo './'`x5/echo512/ccalik/aesni/hash.c
In file included from x5/echo512/ccalik/aesni/hash.c:24:0:
/usr/lib/gcc/x86_64-linux-gnu/4.6/include/wmmintrin.h:34:3: error: #error "AES/PCLMUL instructions not enabled"
x5/echo512/ccalik/aesni/hash.c: In function ‘Compress’:
x5/echo512/ccalik/aesni/hash.c:294:4: warning: implicit declaration of function ‘_mm_aesenc_si128’ [-Wimplicit-function-declaration]
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
...
In file included from x5/echo512/ccalik/aesni/hash.c:24:0:
/usr/lib/gcc/x86_64-linux-gnu/4.6/include/wmmintrin.h:34:3: error: #error "AES/PCLMUL instructions not enabled"
x5/echo512/ccalik/aesni/hash.c: In function ‘Compress’:
x5/echo512/ccalik/aesni/hash.c:294:4: warning: implicit declaration of function ‘_mm_aesenc_si128’ [-Wimplicit-function-declaration]
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
...
Yes, Core 2 Quad doesnt support AES-NI and AVX instructions. I posted early the list of the processors supported. For Intel, from Sandy Bridge and later processors.
What is the cpu of your virtual machine? Could you post your "cat /proc/cpuinfo" ? And, the version 1.2c by ig0tik3d compiles on the same machine? Say me so I can figure out where is the problem.
February 14, 2014, 03:57:42 PM
Doesn't compile under KVM. Sure, there's limited CPU exposure.
Code:
error: CPU you selected does not support x86-64 instruction set
It is only 64 bit for now. Your OS in this case is 32bit? I tested with KVM on a virtual Machine, and it's ok. But guest OS needs to be 64 bit for now.
No, it's 64-bit.
Doesn't compile on Core 2 Quad either.
Code:
gcc -std=gnu99 -DHAVE_CONFIG_H -I. -pthread -fno-strict-aliasing -O3 -march=native -MT minerd-hash.o -MD -MP -MF .deps/minerd-hash.Tpo -c -o minerd-hash.o `test -f 'x5/echo512/ccalik/aesni/hash.c' || echo './'`x5/echo512/ccalik/aesni/hash.c
In file included from x5/echo512/ccalik/aesni/hash.c:24:0:
/usr/lib/gcc/x86_64-linux-gnu/4.6/include/wmmintrin.h:34:3: error: #error "AES/PCLMUL instructions not enabled"
x5/echo512/ccalik/aesni/hash.c: In function ‘Compress’:
x5/echo512/ccalik/aesni/hash.c:294:4: warning: implicit declaration of function ‘_mm_aesenc_si128’ [-Wimplicit-function-declaration]
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
...
In file included from x5/echo512/ccalik/aesni/hash.c:24:0:
/usr/lib/gcc/x86_64-linux-gnu/4.6/include/wmmintrin.h:34:3: error: #error "AES/PCLMUL instructions not enabled"
x5/echo512/ccalik/aesni/hash.c: In function ‘Compress’:
x5/echo512/ccalik/aesni/hash.c:294:4: warning: implicit declaration of function ‘_mm_aesenc_si128’ [-Wimplicit-function-declaration]
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
x5/echo512/ccalik/aesni/hash.c:294:4: error: incompatible types when assigning to type ‘__m128i’ from type ‘int’
...
February 14, 2014, 03:55:56 PM
Doesn't compile under KVM. Sure, there's limited CPU exposure.
Code:
error: CPU you selected does not support x86-64 instruction set
It is only 64 bit for now. Your OS in this case is 32bit? I tested with KVM on a virtual Machine, and it's ok. But guest OS needs to be 64 bit for now.
Jump to: