Author

Topic: [ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner - page 117. (Read 877859 times)

full member
Activity: 169
Merit: 100
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m

Now that is impressive.

thanks, your work on neoscrypt and Xxx is even more impressive.

Eh, Neo maybe, but Xxx algos are just tedious.

I thought he was referring to your wallpaper.  Smiley

Haha, really?

Just trying to be funny, your work on all the algos and your wallpaper are equally impressive.  Well algos are probably more impressive.
full member
Activity: 169
Merit: 100
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m

Now that is impressive.

thanks, your work on neoscrypt and Xxx is even more impressive.

Eh, Neo maybe, but Xxx algos are just tedious.

I thought he was referring to your wallpaper.  Smiley
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m

Now that is impressive.

That's hot. I was hoping Lyra2RE would run cooler, but I guess that's what happens when you optimize the kernel to run without bottlenecks?

Nah, it shouldn't run that hot - he must have it in a case.

yes it's in a case and a reference design, so runs up to 95C and after that the centrifugal fan starts to sound like a vacuum cleaner :-)
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m

Now that is impressive.

thanks, your work on neoscrypt and Xxx is even more impressive.
member
Activity: 81
Merit: 1002
It was only the wind.


The kernelfile option? It's in Nicehash's build, probably.

is that a new Nicehash build to deal with the R9 290x bottle neck ?

No, I don't think they've made one.
full member
Activity: 169
Merit: 100
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m

Now that is impressive.

That's hot. I was hoping Lyra2RE would run cooler, but I guess that's what happens when you optimize the kernel to run without bottlenecks?
full member
Activity: 169
Merit: 100

Probably GPU programming or parallel programming. Most coders don't think like they need to to work on GPU - it's very different.

Thanks.  This gave me some ideas.
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Lyra2RE r9 290@1000/1500

GPU 0:  92.0C 2516RPM | 1.781M/1.782Mh/s | R:  0.0% HW:0 WU:3.123/m
member
Activity: 81
Merit: 1002
It was only the wind.
full member
Activity: 169
Merit: 100


I haven't really messed around with it that much but it was about 310 kh/s with a reference 290x. Intensity 14, w64, tc8192, 2 threads. I think those might be the example settings at your github miner.

What are yours up to now? Please don't tell me you're already hitting 1mh/s.  Haha

Halfway. 550kh/s on 290X.

https://ottrbutt.com/miner/neoscryptwolf-12152014.png (NSFW) - 280X cards, 7950s, and a 290X.

Oh, and my 2 270X cards: https://ottrbutt.com/miner/neoscryptwolf-12152014-2.png

Wow.  That's nice.  Looks like your 270x may catch my 290x.  That'll be a sad day.  I need to find a friend that knows how to program.

Not just programming, it's kind of a specialized area. Just broke 560kh/s on my 290X, but I'm gonna keep messing with it.

What would you call that specialized area if I need to describe it to someone?
full member
Activity: 169
Merit: 100

I think that may be the case.  CGWatcher uses temporary config files and something may go wrong with that step.  You can disable that setting.  I'll try that.

On a side note what is the character called that is between w64 and 4ku0 (w64l4ku0)?



l4 for 32-bit, l8 for 64-bit

Thanks.  That's why yours are always l8, I used to use your 64bit builds.

What hashrate do you get for Neoscrypt?

I haven't really messed around with it that much but it was about 310 kh/s with a reference 290x. Intensity 14, w64, tc8192, 2 threads. I think those might be the example settings at your github miner.

What are yours up to now? Please don't tell me you're already hitting 1mh/s.  Haha

Halfway. 550kh/s on 290X.

https://ottrbutt.com/miner/neoscryptwolf-12152014.png (NSFW) - 280X cards, 7950s, and a 290X.

Oh, and my 2 270X cards: https://ottrbutt.com/miner/neoscryptwolf-12152014-2.png

Wow.  That's nice.  Looks like your 270x may catch my 290x.  That'll be a sad day.  I need to find a friend that knows how to program.
legendary
Activity: 1400
Merit: 1050
Hey djm34, there's something is wrong with the compiled binary that you post https://bitcointalksearch.org/topic/m.9817141

When hashing lyra2RE (vtc), the hashrate shown on the miner is indeed faster 5-10% than the metalicjames one, but on pool it will only record half the hashrate.
All stats normal, low rejected shares, no hardware errors that are shown on the miners.
Tested it on two different pools, coinotron and hashlink.eu, still when using your windows binary, only half hashrate will be recorded on the web.
Somehow only half the shares are sent by the miners or accepted by the pool.
And it's not  just estimations only, the coin received is halved for the same time period.
When back on using metalicjames version, the hashrate going back to normal again.

Where's that half hashrate gone??
Could you please recheck the binaries. Thanks.
must be related to the difficulty adjustment..., I will re-upload it. In the mean time there is also a difficulty multiplier option in sgminer (it shows as deprecated in the help, however it still works).
Also, I think that several pool are still tuning their hash report things...  Grin

Nah man it isn't the pool. I gave this a try and thx for it...finally got to that sweet 640-660 kh/s x 5 (3.2) but on Givemecoins pool showing exactly half (~.1.65). Change back to sgminer RC1 back to 550 kh/s but showing correct on pool.

What you are telling shows that it depends on which pool you run... so it is up to the pool admi to set up the pool corectly so that it coressponds the hashrate the miner tell you.

So 3 pools haven't been set up right for your fork? Because b/w me and one other person posting about this that makes 3. But yet these 3 pools report correctly with another miner? Seems strange to me...

Yep same for me, already mentioned that briefly in the Vertcoin thread.

https://bitcointalksearch.org/topic/m.9826952

Sticking to the first sgminer release for now.
ah that you didn't say you were using the binaries, yes I haven't updated the binaries yet, only the source...
however you can still use it, but you need to append to the command line: "--difficulty-multiplier 0.0078125"  (which corresponds to 1/128)
member
Activity: 61
Merit: 10
Hey djm34, there's something is wrong with the compiled binary that you post https://bitcointalksearch.org/topic/m.9817141

When hashing lyra2RE (vtc), the hashrate shown on the miner is indeed faster 5-10% than the metalicjames one, but on pool it will only record half the hashrate.
All stats normal, low rejected shares, no hardware errors that are shown on the miners.
Tested it on two different pools, coinotron and hashlink.eu, still when using your windows binary, only half hashrate will be recorded on the web.
Somehow only half the shares are sent by the miners or accepted by the pool.
And it's not  just estimations only, the coin received is halved for the same time period.
When back on using metalicjames version, the hashrate going back to normal again.

Where's that half hashrate gone??
Could you please recheck the binaries. Thanks.
must be related to the difficulty adjustment..., I will re-upload it. In the mean time there is also a difficulty multiplier option in sgminer (it shows as deprecated in the help, however it still works).
Also, I think that several pool are still tuning their hash report things...  Grin

Nah man it isn't the pool. I gave this a try and thx for it...finally got to that sweet 640-660 kh/s x 5 (3.2) but on Givemecoins pool showing exactly half (~.1.65). Change back to sgminer RC1 back to 550 kh/s but showing correct on pool.

What you are telling shows that it depends on which pool you run... so it is up to the pool admi to set up the pool corectly so that it coressponds the hashrate the miner tell you.

So 3 pools haven't been set up right for your fork? Because b/w me and one other person posting about this that makes 3. But yet these 3 pools report correctly with another miner? Seems strange to me...

Yep same for me, already mentioned that briefly in the Vertcoin thread.

https://bitcointalksearch.org/topic/m.9826952

Sticking to the first sgminer release for now.
member
Activity: 81
Merit: 1002
It was only the wind.
hero member
Activity: 896
Merit: 1000
hero member
Activity: 896
Merit: 1000

Halfway. 550kh/s on 290X.

https://ottrbutt.com/miner/neoscryptwolf-12152014.png (NSFW) - 280X cards, 7950s, and a 290X.

Oh, and my 2 270X cards: https://ottrbutt.com/miner/neoscryptwolf-12152014-2.png

Do you use the public kernel? What is your setting?
full member
Activity: 169
Merit: 100

I think that may be the case.  CGWatcher uses temporary config files and something may go wrong with that step.  You can disable that setting.  I'll try that.

On a side note what is the character called that is between w64 and 4ku0 (w64l4ku0)?



l4 for 32-bit, l8 for 64-bit

Thanks.  That's why yours are always l8, I used to use your 64bit builds.

What hashrate do you get for Neoscrypt?

I haven't really messed around with it that much but it was about 310 kh/s with a reference 290x. Intensity 14, w64, tc8192, 2 threads. I think those might be the example settings at your github miner.

What are yours up to now? Please don't tell me you're already hitting 1mh/s.  Haha
full member
Activity: 136
Merit: 100
Hey djm34, there's something is wrong with the compiled binary that you post https://bitcointalksearch.org/topic/m.9817141

When hashing lyra2RE (vtc), the hashrate shown on the miner is indeed faster 5-10% than the metalicjames one, but on pool it will only record half the hashrate.
All stats normal, low rejected shares, no hardware errors that are shown on the miners.
Tested it on two different pools, coinotron and hashlink.eu, still when using your windows binary, only half hashrate will be recorded on the web.
Somehow only half the shares are sent by the miners or accepted by the pool.
And it's not  just estimations only, the coin received is halved for the same time period.
When back on using metalicjames version, the hashrate going back to normal again.

Where's that half hashrate gone??
Could you please recheck the binaries. Thanks.
must be related to the difficulty adjustment..., I will re-upload it. In the mean time there is also a difficulty multiplier option in sgminer (it shows as deprecated in the help, however it still works).
Also, I think that several pool are still tuning their hash report things...  Grin

Nah man it isn't the pool. I gave this a try and thx for it...finally got to that sweet 640-660 kh/s x 5 (3.2) but on Givemecoins pool showing exactly half (~.1.65). Change back to sgminer RC1 back to 550 kh/s but showing correct on pool.

What you are telling shows that it depends on which pool you run... so it is up to the pool admi to set up the pool corectly so that it coressponds the hashrate the miner tell you.

So 3 pools haven't been set up right for your fork? Because b/w me and one other person posting about this that makes 3. But yet these 3 pools report correctly with another miner? Seems strange to me...
full member
Activity: 169
Merit: 100

I think that may be the case.  CGWatcher uses temporary config files and something may go wrong with that step.  You can disable that setting.  I'll try that.

On a side note what is the character called that is between w64 and 4ku0 (w64l4ku0)?



l4 for 32-bit, l8 for 64-bit

Thanks.  That's why yours are always l8, I used to use your 64bit builds.
full member
Activity: 169
Merit: 100
 Here are all the bins getting produced:

  I get this from my output.log:

Code:
[14:07:01] Building binary neoscryptHawaiigw64l4ku0big7hs.bin
[14:07:01] Error -11: Building Program (clBuildProgram)
[14:07:01] "C:\Users\ANIMAL~1\AppData\Local\Temp\OCL4772T27.cl", line 368: warning:
          variable "t" was declared but never referenced
   uint4 t, st[4];
        ^

"C:\Users\ANIMAL~1\AppData\Local\Temp\OCL4772T27.cl", line 495: error:
          identifier "MAX_GLOBAL_THREADS" is undefined
   __global ulong16 *V = (__global ulong16 *)(padcache + (0x8000 * (get_global_id(0) % MAX_GLOBAL_THREADS)));
                                                                                      ^

"C:\Users\ANIMAL~1\AppData\Local\Temp\OCL4772T27.cl", line 513: warning:
          argument of type "__global ulong16 *" is incompatible with parameter
          of type "__global uint16 *"
   SMix(X, V, flag);
          ^

1 error detected in the compilation of "C:\Users\ANIMAL~1\AppData\Local\Temp\OCL4772T27.cl".

Frontend phase failed compilation.

Any ideas?  Could CGWatcher be interfering somehow when bins are made?

You have wrong marucoin-mod.cl.
Try to find right, and replace in ./kernels
fix https://bitcointalk.org/index.php?topic=854257.240

Thanks, I have actually been using that marucoin-mod.cl from the thread.  To be sure I double checked and it does have the correct line 96.  I'm all ears if you've got any other ideas.

You're building a Neoscrypt binary, not an X13 one.

Yeah, it's trying to build a weird one right, Building binary neoscryptHawaiigw64l4ku0big7hs.bin ?  Here are some other strange ones that actually got built.

darkcoin-modHawaiigw64l4ku0.bin (Correct)
darkcoin-modHawaiigw64l4ku0big7hs.bin
darkcoin-modHawaiigw64l4tc8192.bin
marucoin-modHawaiigw64l4ku0.bin
marucoin-modHawaiigw64l4ku0big7hs.bin (Correct)
marucoin-modHawaiigw64l4tc8192.bin
neoscryptHawaiigw64l4tc8192.bin (Correct)




CGWatcher may be screwing up.

I think that may be the case.  CGWatcher uses temporary config files and something may go wrong with that step.  You can disable that setting.  I'll try that.

On a side note what is the character called that is between w64 and 4ku0 (w64l4ku0)?

Jump to: