Namsbreh
|
|
March 06, 2016, 04:04:25 PM |
|
hi, what is the speed of a single 750ti card on Decred?
Around 500mh depending on clocks.
|
|
|
|
|
|
|
|
|
"In a nutshell, the network works like a distributed
timestamp server, stamping the first transaction to spend a coin. It
takes advantage of the nature of information being easy to spread but
hard to stifle." -- Satoshi
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
Vaccomondus
|
|
March 06, 2016, 09:47:06 PM |
|
can this version be improved?
|
|
|
|
bathrobehero
Legendary
Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
|
March 07, 2016, 11:08:15 PM Last edit: March 07, 2016, 11:21:25 PM by bathrobehero |
|
I only get 445 Mh/s per 750 Ti. What am I missing? Do I need a certain driver or cuda toolkit or something?
Edit: apparently I get close to 500 if I only use one card but with 6 cards and a dualcore Pentium G3240 I only get abou 445 per card. I presume because the first round of blake is done on the CPU as per sp_ said.
|
Not your keys, not your coins!
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 07, 2016, 11:34:18 PM |
|
I only get 445 Mh/s per 750 Ti. What am I missing? Do I need a certain driver or cuda toolkit or something?
Edit: apparently I get close to 500 if I only use one card but with 6 cards and a dualcore Pentium G3240 I only get abou 445 per card. I presume because the first round of blake is done on the CPU as per sp_ said.
It could be the CPU that is he bottleneck (you could check usage) but try increasing your pagefile size, assuming you're on Windows.
|
|
|
|
bathrobehero
Legendary
Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
|
March 08, 2016, 12:06:08 AM |
|
I only get 445 Mh/s per 750 Ti. What am I missing? Do I need a certain driver or cuda toolkit or something?
Edit: apparently I get close to 500 if I only use one card but with 6 cards and a dualcore Pentium G3240 I only get abou 445 per card. I presume because the first round of blake is done on the CPU as per sp_ said.
It could be the CPU that is he bottleneck (you could check usage) but try increasing your pagefile size, assuming you're on Windows. 16GB pagefile with ~27% CPU usage but it's still slower than 1 card.
|
Not your keys, not your coins!
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 08, 2016, 03:34:55 AM |
|
I only get 445 Mh/s per 750 Ti. What am I missing? Do I need a certain driver or cuda toolkit or something?
Edit: apparently I get close to 500 if I only use one card but with 6 cards and a dualcore Pentium G3240 I only get abou 445 per card. I presume because the first round of blake is done on the CPU as per sp_ said.
It could be the CPU that is he bottleneck (you could check usage) but try increasing your pagefile size, assuming you're on Windows. 16GB pagefile with ~27% CPU usage but it's still slower than 1 card. Small cache maybe?
|
|
|
|
myagui
Legendary
Offline
Activity: 1154
Merit: 1001
|
|
March 08, 2016, 08:04:53 AM |
|
^ Also maybe worthwhile checking if multiple ccminer instances will help. So you could launch instance 1 with -d 0,1,2, and instance 2 with -d 3,4,5. In addition, ensure that cpu affinity is reserving specific cpu cores for instance one, and other cpu cores for instance 2.
|
|
|
|
Masked_Immortal
Member
Offline
Activity: 67
Merit: 10
|
|
March 08, 2016, 01:29:50 PM |
|
hmm.. my miner crash after some hours,this is the error report Cuda error in func 'decred_cpu_setBlock_52' at line 321 : unknown error. any idea about this?
|
|
|
|
s7icky
Member
Offline
Activity: 106
Merit: 10
|
|
March 08, 2016, 02:40:37 PM |
|
hmm.. my miner crash after some hours,this is the error report Cuda error in func 'decred_cpu_setBlock_52' at line 321 : unknown error. any idea about this? turn down intensity it was doing that with mine.
|
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 08, 2016, 03:37:15 PM |
|
^ Also maybe worthwhile checking if multiple ccminer instances will help. So you could launch instance 1 with -d 0,1,2, and instance 2 with -d 3,4,5. In addition, ensure that cpu affinity is reserving specific cpu cores for instance one, and other cpu cores for instance 2.
I had to think about this a bit. ccminer should already create multiple CPU threads spread over all cores. This can be confirmed by using the -D option to enable debug output. Even so, with three threads per core it could introduce scheduling latency. That combined with the small cache could easilly cause a 10% degradation in performance. The algo probbaly factors into it as well. Does the degradation occur with other algos? I have seen some odd performnance differences while testing cryptonight on cpuminer that I still don't understand. At first i thought it was due to some affinity tricks but I haven't found anything in the code to explain it. In short cryptonight performs radically differently on different CPUs/OSs. On a 6700K running Linux I get best performance CPU mining with 4 threads. More threads causes the total hashrate to drop to as low as half the 4 thread rate. Most other algos perform much better with more threads. The CPUs also run pretty cool on cryptonight suggesting they are often stalled waiting for data (ie memory bound). There shouldn't be any scheduling delays because the number of running threads is less than the available virtual cores. Any thread contention would occur during execution and be mitigated by hyperthreading. That leaves cache performance as the most likley cause for both issues. If the total memory requirements of all threads exceeds the available cache it will significantly affect cache performance. It's a step function as each cache level overflows. Seems like going too cheap with a CPU for a mining rig isn't a good idea.
|
|
|
|
pallas
Legendary
Offline
Activity: 2716
Merit: 1094
Black Belt Developer
|
|
March 08, 2016, 03:43:16 PM |
|
That leaves cache performance as the most likley cause for both issues. If the total memory requirements of all threads exceeds the available cache it will significantly affect cache performance. It's a step function as each cache level overflows.
^^^^^ This Seems like going too cheap with a CPU for a mining rig isn't a good idea.
True. But it's very difficult to roi with cpu anyways, 'cause of botnets.
|
|
|
|
myagui
Legendary
Offline
Activity: 1154
Merit: 1001
|
|
March 08, 2016, 04:19:39 PM |
|
@joblo, Cryptonight on CPU is a particular case. There's a 2MB scratchpad per thread (or something else which proper name I don't recall). For whatever CPU you have, the ideal number of threads will always be cache-size/2. Most i7's have 8MB cache, so optimal threads = 4. As far as the rest of the details that you posted, way over my head. /searching <nearest exit>
|
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 08, 2016, 04:20:44 PM |
|
That leaves cache performance as the most likley cause for both issues. If the total memory requirements of all threads exceeds the available cache it will significantly affect cache performance. It's a step function as each cache level overflows.
^^^^^ This Seems like going too cheap with a CPU for a mining rig isn't a good idea.
True. But it's very difficult to roi with cpu anyways, 'cause of botnets. My comments were aimed at the CPU performance of GPU mining. My CPU mining comparison was made only to illustrate a possible similar problem. Howerer, I agree that it's very difficult, if not impossible, to ROI with CPU mining.
|
|
|
|
Epsylon3 (OP)
Legendary
Offline
Activity: 1484
Merit: 1082
ccminer/cpuminer developer
|
|
March 08, 2016, 04:25:43 PM |
|
its still possible to get some cents from datacenter "sleeping" servers i have some cheap dedicated ones
|
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 08, 2016, 04:30:49 PM |
|
@joblo, Cryptonight on CPU is a particular case. There's a 2MB scratchpad per thread (or something else which proper name I don't recall). For whatever CPU you have, the ideal number of threads will always be cache-size/2. Most i7's have 8MB cache, so optimal threads = 4. As far as the rest of the details that you posted, way over my head. /searching <nearest exit> I don't agree with it being over your head, your comment was spot on and perfectly illustrates what I was saying. As you add more threads the memory requirements increase and when it becomes larger that the cache size performance drops noticibly in spite of any benefits provided by hyperthreading. Perhaps a similar thing is occurring on a small CPU when GPU mining ETH with many threads.
|
|
|
|
MeatPopsicle
Newbie
Offline
Activity: 49
Merit: 0
|
|
March 09, 2016, 07:24:47 PM |
|
[2016-03-09 19:23:25] GPU #1: aes_cpu_init invalid device symbol [2016-03-09 19:23:25] GPU #3: aes_cpu_init invalid device symbol [2016-03-09 19:23:25] GPU #2: aes_cpu_init invalid device symbol [2016-03-09 19:23:25] GPU #0: aes_cpu_init invalid device symbol [2016-03-09 19:23:26] GPU #1: aes_cpu_init invalid device symbol [2016-03-09 19:23:26] GPU #3: aes_cpu_init invalid device symbol [2016-03-09 19:23:26] GPU #2: aes_cpu_init invalid device symbol [2016-03-09 19:23:26] GPU #0: aes_cpu_init invalid device symbol [2016-03-09 19:23:27] GPU #1: quark_blake512_cpu_setBlock_80:318 invalid device symbol Cuda error in func 'cuda_check_cpu_setTarget' at line 41 : invalid device symbol.
Any ideas?
|
|
|
|
Epsylon3 (OP)
Legendary
Offline
Activity: 1484
Merit: 1082
ccminer/cpuminer developer
|
|
March 09, 2016, 07:36:58 PM |
|
|
|
|
|
MeatPopsicle
Newbie
Offline
Activity: 49
Merit: 0
|
|
March 09, 2016, 07:43:32 PM |
|
Ah, thanks, missed that somehow.
|
|
|
|
chrysophylax
Legendary
Offline
Activity: 2814
Merit: 1091
--- ChainWorks Industries ---
|
|
March 10, 2016, 02:42:44 AM |
|
hi all ... epsylon3 - ive tried to contact you on irc - but as usual you are a very busy man ... i have a small issue that i know has been looked at before a long time ago - so some of you may know how to 'fix' it ... i have just rebuild the main server in the network ( from a crash last week due to power issues ) and im at wits end as to how to rectify the compilation error im getting ... the rebuild was done with centos 7 x64 ( to match the multitude of vps systems granite currently has ) and the compilation of the miners ( both sp and tpruvot ) bomb with different issues ... ccminer-spmod bombs with a declaration issue in whirlpool - so i just changed code to declare the variable and it compile nicely ... so ccminer-spmod78 works and currently hashes without issue ( i have two 750ti oc cards in the server to test ) ... ccminer-tpruvot174 bombs with a json error - and i am unsure how to fix this ... i believe this is an icompatibility issue with the jansson version in centos 7 x64 - but cant figure out what to do to get ccminer-tpruvot174 compiled ... below is the output of the issue ... g++ -DHAVE_CONFIG_H -I. -fopenmp -pthread -fno-strict-aliasing -I/usr/local/cuda/include -DUSE_WRAPNVML -O3 -march=native -D_REENTRANT -falign-functions=16 -falign-jumps=16 -falign-labels=16 -MT ccminer-util.o -MD -MP -MF .deps/ccminer-util.Tpo -c -o ccminer-util.o `test -f 'util.cpp' || echo './'`util.cpp pools.cpp: In function ‘bool parse_pool_array(json_t*)’: pools.cpp:326:32: error: ‘json_array_foreach’ was not declared in this scope json_array_foreach(obj, idx, p) ^ pools.cpp:327:2: error: expected ‘;’ before ‘{’ token { ^ pools.cpp:393:1: error: expected ‘}’ at end of input } ^ make[2]: *** [ccminer-pools.o] Error 1 make[2]: *** Waiting for unfinished jobs.... ccminer.cpp: In function ‘bool work_decode(const json_t*, work*)’: ccminer.cpp:635:34: error: ‘json_array_foreach’ was not declared in this scope json_array_foreach(txs, idx, p) { ^ ccminer.cpp:635:36: error: expected ‘;’ before ‘{’ token json_array_foreach(txs, idx, p) { ^ ccminer.cpp:3514:1: error: expected ‘}’ at end of input } ^ ccminer.cpp:3514:1: error: expected ‘}’ at end of input ccminer.cpp:3514:1: error: expected ‘}’ at end of input make[2]: *** [ccminer-ccminer.o] Error 1
any help with this would be very much appreciated ... the other 'issue' - though not an issue with the miner itself - is that i have absolutely no idea how to create a config file for ccminer ... there seems to be NO documentation about it at all ... can anyone help? ... #crysx
|
|
|
|
joblo
Legendary
Offline
Activity: 1470
Merit: 1114
|
|
March 10, 2016, 04:44:17 AM |
|
hi all ... epsylon3 - ive tried to contact you on irc - but as usual you are a very busy man ... i have a small issue that i know has been looked at before a long time ago - so some of you may know how to 'fix' it ... i have just rebuild the main server in the network ( from a crash last week due to power issues ) and im at wits end as to how to rectify the compilation error im getting ... the rebuild was done with centos 7 x64 ( to match the multitude of vps systems granite currently has ) and the compilation of the miners ( both sp and tpruvot ) bomb with different issues ... ccminer-spmod bombs with a declaration issue in whirlpool - so i just changed code to declare the variable and it compile nicely ... so ccminer-spmod78 works and currently hashes without issue ( i have two 750ti oc cards in the server to test ) ... ccminer-tpruvot174 bombs with a json error - and i am unsure how to fix this ... i believe this is an icompatibility issue with the jansson version in centos 7 x64 - but cant figure out what to do to get ccminer-tpruvot174 compiled ... below is the output of the issue ... g++ -DHAVE_CONFIG_H -I. -fopenmp -pthread -fno-strict-aliasing -I/usr/local/cuda/include -DUSE_WRAPNVML -O3 -march=native -D_REENTRANT -falign-functions=16 -falign-jumps=16 -falign-labels=16 -MT ccminer-util.o -MD -MP -MF .deps/ccminer-util.Tpo -c -o ccminer-util.o `test -f 'util.cpp' || echo './'`util.cpp pools.cpp: In function ‘bool parse_pool_array(json_t*)’: pools.cpp:326:32: error: ‘json_array_foreach’ was not declared in this scope json_array_foreach(obj, idx, p) ^ pools.cpp:327:2: error: expected ‘;’ before ‘{’ token { ^ pools.cpp:393:1: error: expected ‘}’ at end of input } ^ make[2]: *** [ccminer-pools.o] Error 1 make[2]: *** Waiting for unfinished jobs.... ccminer.cpp: In function ‘bool work_decode(const json_t*, work*)’: ccminer.cpp:635:34: error: ‘json_array_foreach’ was not declared in this scope json_array_foreach(txs, idx, p) { ^ ccminer.cpp:635:36: error: expected ‘;’ before ‘{’ token json_array_foreach(txs, idx, p) { ^ ccminer.cpp:3514:1: error: expected ‘}’ at end of input } ^ ccminer.cpp:3514:1: error: expected ‘}’ at end of input ccminer.cpp:3514:1: error: expected ‘}’ at end of input make[2]: *** [ccminer-ccminer.o] Error 1
any help with this would be very much appreciated ... the other 'issue' - though not an issue with the miner itself - is that i have absolutely no idea how to create a config file for ccminer ... there seems to be NO documentation about it at all ... can anyone help? ... #crysx Have you tried downloading a fresh copy? Did you dl the release or the latest git? The problem is the compiler can't find json_array_foreach which is #defined in compat/jansson/jansson.h. It works for me with the release version. It seems like either the dl is corrupt or a bug was introduced after release. Can't help with the config file, never used it.
|
|
|
|
|