XPM - Issue with mining....reward 5 XPM

[Kavre]

Hello,

here is my problem. I have a server with 8 cpu (8 x Intel Xeon L5420 @ 2.50GHz).
I'm mining around 2000 PPS, I hardly believe that is doing just 2k pps..I have setgenerate true -1
I'm sure that is possible to get much more pps (i think).
Running on Ubuntu 10.04. Version hp4 from mikaelh.
So i'm giving a reward 5 XPM to solve this problem.
I tried almost everything. I was playing with setgenerate true 1........30.....100...200 but still no more than 2000 PPS.
I'm completely lost. Can't get it work properly.
Any idea what is the problem?

Thanks

[1713835497]

1713835497

[1713835497]

1713835497

[1713835497]

1713835497

[BitRock]

Hello,

here is my problem. I have a server with 8 cpu (8 x Intel Xeon L5420 @ 2.50GHz).
I'm mining around 2000 PPS, I hardly believe that is doing just 2k pps..I have setgenerate true -1
I'm sure that is possible to get much more pps (i think).
Running on Ubuntu 10.04. Version hp4 from mikaelh.
So i'm giving a reward 5 XPM to solve this problem.
I tried almost everything. I was playing with setgenerate true 1........30.....100...200 but still no more than 2000 PPS.
I'm completely lost. Can't get it work properly.
Any idea what is the problem?

Thanks

according to mikaelh's post

Q: Why is my performance worse on Linux?
A: This version relies on libgmp to do most of the complex calculations. Make sure your libgmp is up-to-date. Old versions do not support newer CPUs. I suggest compiling your own libgmp from latest source code for maximum performance.

[logicsol]

set generate set how many cores you are using. -1 will use all of them.

Make sure you are using the 64bit version of the miner as well.

[FiiNALiZE]

Compile the source here: https://bitcointalk.org/index.php?topic=255782.0

You should get a massive boost in PPS.

[solstoce]

Hello,

here is my problem. I have a server with 8 cpu (8 x Intel Xeon L5420 @ 2.50GHz).
I'm mining around 2000 PPS, I hardly believe that is doing just 2k pps..I have setgenerate true -1
I'm sure that is possible to get much more pps (i think).
Running on Ubuntu 10.04. Version hp4 from mikaelh.
So i'm giving a reward 5 XPM to solve this problem.
I tried almost everything. I was playing with setgenerate true 1........30.....100...200 but still no more than 2000 PPS.
I'm completely lost. Can't get it work properly.
Any idea what is the problem?

Thanks

I think this is what your looking for let know if this helps http://www.ppcointalk.org/index.php?topic=341.msg2361#msg2361

[Eli0t]

try chemists sse build, since that cpu doesn't have avx some of the newer hp avx builds dont work to well

https://github.com/Chemisist/primecoin
https://bitcointalk.org/index.php?topic=253662.msg2716037#msg2716037

AeSZAm1BWParxJJZbNwPqnd5CQCncychqG
if you found this helpful

[Kavre]

Hello,

here is my problem. I have a server with 8 cpu (8 x Intel Xeon L5420 @ 2.50GHz).
I'm mining around 2000 PPS, I hardly believe that is doing just 2k pps..I have setgenerate true -1
I'm sure that is possible to get much more pps (i think).
Running on Ubuntu 10.04. Version hp4 from mikaelh.
So i'm giving a reward 5 XPM to solve this problem.
I tried almost everything. I was playing with setgenerate true 1........30.....100...200 but still no more than 2000 PPS.
I'm completely lost. Can't get it work properly.
Any idea what is the problem?

Thanks

according to mikaelh's post

Q: Why is my performance worse on Linux?
A: This version relies on libgmp to do most of the complex calculations. Make sure your libgmp is up-to-date. Old versions do not support newer CPUs. I suggest compiling your own libgmp from latest source code for maximum performance.

libgmp is not supported for ubuntu 10.04, i had to install libgmp3

set generate set how many cores you are using. -1 will use all of them.

Make sure you are using the 64bit version of the miner as well.

yes it's 64bit
tried all of them still nothing

Compile the source here: https://bitcointalk.org/index.php?topic=255782.0

You should get a massive boost in PPS.

i have already this 

Hello,

here is my problem. I have a server with 8 cpu (8 x Intel Xeon L5420 @ 2.50GHz).
I'm mining around 2000 PPS, I hardly believe that is doing just 2k pps..I have setgenerate true -1
I'm sure that is possible to get much more pps (i think).
Running on Ubuntu 10.04. Version hp4 from mikaelh.
So i'm giving a reward 5 XPM to solve this problem.
I tried almost everything. I was playing with setgenerate true 1........30.....100...200 but still no more than 2000 PPS.
I'm completely lost. Can't get it work properly.
Any idea what is the problem?

Thanks

I think this is what your looking for let know if this helps http://www.ppcointalk.org/index.php?topic=341.msg2361#msg2361

not sure what do you mean :S

try chemists sse build, since that cpu doesn't have avx some of the newer hp avx builds dont work to well

https://github.com/Chemisist/primecoin
https://bitcointalk.org/index.php?topic=253662.msg2716037#msg2716037

AeSZAm1BWParxJJZbNwPqnd5CQCncychqG
if you found this helpful

i had Chemisist version but was doing around 1000-1200PPS


___________
thanks any1 for help

[AgentME]

PPS is not a good measure for performance.

[logicsol]

can you see if all your cpus are getting pegged at 100%?

[Kavre]

PPS is not a good measure for performance.

True. But how to explain that my laptop is doing 1500PPS and 8 cpu together just a bit more......doesn't seem right to me (i think)

can you see if all your cpus are getting pegged at 100%?

not sure how to check it   
"top"  or some other command?

[nfuse]

my E5630 does 3096 pps and my E5645 4318 pps both windows (running modified version hp4

[Kavre]

my E5630 does 3096 pps and my E5645 4318 pps both windows (running modified version hp4

you are talking for 1 cpu-each, right?
seems fair that pps

[OnlyC]

Try ubuntu 12.04 if you have time

[logicsol]

PPS is not a good measure for performance.

True. But how to explain that my laptop is doing 1500PPS and 8 cpu together just a bit more......doesn't seem right to me (i think)

can you see if all your cpus are getting pegged at 100%?

not sure how to check it   
"top"  or some other command?

use system monitor, it should be able to tell you.

[Kavre]

PPS is not a good measure for performance.

True. But how to explain that my laptop is doing 1500PPS and 8 cpu together just a bit more......doesn't seem right to me (i think)

can you see if all your cpus are getting pegged at 100%?

not sure how to check it   
"top"  or some other command?

use system monitor, it should be able to tell you.

have just terminal/cmd

[logicsol]

PPS is not a good measure for performance.

True. But how to explain that my laptop is doing 1500PPS and 8 cpu together just a bit more......doesn't seem right to me (i think)

can you see if all your cpus are getting pegged at 100%?

not sure how to check it   
"top"  or some other command?

use system monitor, it should be able to tell you.

have just terminal/cmd

You're running GNOMEless then? hmm. top should work

[Kavre]

PPS is not a good measure for performance.

True. But how to explain that my laptop is doing 1500PPS and 8 cpu together just a bit more......doesn't seem right to me (i think)

can you see if all your cpus are getting pegged at 100%?

not sure how to check it   
"top"  or some other command?

use system monitor, it should be able to tell you.

have just terminal/cmd

You're running GNOMEless then? hmm. top should work

primecoind  %CPU  between 700-800
what mean this?

[mustyoshi]

Open up your makefile, and add -O3 -march=native -s and recompile it from source.

Also, do that with libgmp for even more performance.

[logicsol]

primecoind  %CPU  between 700-800
what mean this?

Sounds like the miner is properly using the CPUs at least, that number means 90-100% usage across all 8 cpus. I think recompiling with the latest libgmp is your best route.

I'd hazard a guess that it's not properly threading the CPUs, as you get similar pps as my i7-920 @ stock. AKA it's running your 8 4 core CPUs as 1 8 core cpu.

[Kavre]

primecoind  %CPU  between 700-800
what mean this?

Sounds like the miner is properly using the CPUs at least, that number means 90-100% usage across all 8 cpus. I think recompiling with the latest libgmp is your best route.

I'd hazard a guess that it's not properly threading the CPUs, as you get similar pps as my i7-920 @ stock. AKA it's running your 8 4 core CPUs as 1 8 core cpu.

huh
i have too 920 on desktop pc, i'm doing with it around 3300....

Open up your makefile, and add -O3 -march=native -s and recompile it from source.

Also, do that with libgmp for even more performance.

DON'T do that with GMP. GMP is already tuned for greatest performance, and using "-O3" can sometimes break code.

Maybe i will try it, just to know if are any differnece.
Like I said I'm using libgmp3. Should run like libgmp?