all that stuff about advanced code is fishermen's talk,
what beats me why so drastic fall of performance with 14.9
on badman's mod with public kernels
Probably AMD changed something small in the compiler - not sure if the one doing AMD IL -> ISA is on the card or not, but if it's the driver (and I think it is) even a small change could make that mess slow to a crawl. yep, but what... even scrypt is damn slow, 450 vs 650 with same gpu settings, pallas groestl is 7.5 after 17.5 with 14.7, great I've tried tweaking it for 14.9 and I was able to improve it a bit but far from 14.6/14.7 speed. |
|
|
|
|
It looks like the compiler included in 14.9 drivers produces binaries which run considerably slower than older releases.
The same happens for other algorythms as well, for example on X11.
I've tweaked the code a bit but I still can't reach full speed, so I will keep on trying or, eventually, wait for a new driver release.
Meanwhile, if you are on 14.9, use the provided bin file instead of the kernel source in .cl format.
|
|
|
|
|
block explorer looks to be down:
HTTP request sent, awaiting response... Read error (Connection reset by peer) in headers.
|
|
|
|
|
Just wanted to say I'm following this coin to see the groestl and pos rollout.
Really looking forward to merge mining with DMD and possibly other coins.
I believe my groestl kernel will work with noble and I'm here for support (best if you post on the kernel thread of my signature, though).
|
|
|
|
|
I believe that the coin aim is to replace the useless mining of the other coins with useful computation.
In this regard, the dogecoin effort doesn't deliver, nor does the current curecoin implementation.
But version 2 will, so I am really looking forward to it!
Let's hope people will understand the importance of this aspect of cryptocurrencies.
|
|
|
|
I can not replace the bin file. After restarting the miner, is presented again the old bin file. How to replace it? Thank you
if a bin file with the same name already exists, it shouldn't replace it. so best replace the bin file the miner creates with mine, using the same filename. |
|
|
|
I am getting better hash rates with this than with grs-sgminer (solo mining).
R290 - 20.8 mh/s vs 18.3 mh/s with grs
R280x - 17.3 mh/s vs 14.6 mh/s with grs
I am not seeing much advantage to increase intensity above 20.
I am seeing that both cards are running hotter - using more power. I have the 14.6 beta driver package installed.
However, I have two questions about what I am seeing.
1. My R290s are set to 1100 for GPU speed, yet they are actually running under 1000. I am seeing speeds of only 937. auto gpu is disabled. I increased powertune to 20 and the gpu speed increased slightly but not to the speed setting of 1100. Can anyone explain this?
2. I set the memory speeds on the R290s to 150. With grs-sgminer I don't have to, they run at that speed simply because the memory is not used. With this software, the memory on the R290s runs are 1250 and I cannot change that. On the R280X, the memory is running at 1500 and I cannot lower it no matter what I set it to. Can anyone explain this?
BTW, I was using the 13.12 driver package before and I could not get the hash rates that I am getting with the newer drivers. The best I could get out of the R290s was about 13.8 mh/s. With grs-sgminer and the 13.12 drivers I was getting 18.3 mh/s.
1. High ram clock = lower Max core clock 2. If on Linux try: http://epixoip.github.io/od6config/ |
|
|
|
Hi, 4gpu 280x = 69Mhs
This is for DMD, right? Could you share your full set of parameters passed to the sgminer? Me, on 280x I can't get past 7.45 Mhs, with the sph-sgminer version 4.1.0-105-gf02f. The hash rate increase due this optimized kernel has been about 7% only. Thanks. That is right in the MHs range I was getting with catalyst 14.9 driver ... use 14.7 RC3 or the bin provided by pallas and see what you get Could not locate 14.7RC3 (I am on Linux), only 14.4, will try. The bin provided by pallas is for Hawaii (290), not for Tahiti (280), correct? this is 14.9 (non-beta) for linux: http://support.amd.com/en-us/download/desktop?os=Linux+x86the hawaii binary should work on tahiti as well. Any benefit for upgrading to 14.9? using the old binary, made with 14.6, I noticed less power usage with 14.9. |
|
|
|
Hi, 4gpu 280x = 69Mhs
This is for DMD, right? Could you share your full set of parameters passed to the sgminer? Me, on 280x I can't get past 7.45 Mhs, with the sph-sgminer version 4.1.0-105-gf02f. The hash rate increase due this optimized kernel has been about 7% only. Thanks. That is right in the MHs range I was getting with catalyst 14.9 driver ... use 14.7 RC3 or the bin provided by pallas and see what you get Could not locate 14.7RC3 (I am on Linux), only 14.4, will try. The bin provided by pallas is for Hawaii (290), not for Tahiti (280), correct? this is 14.9 (non-beta) for linux: http://support.amd.com/en-us/download/desktop?os=Linux+x86the hawaii binary should work on tahiti as well. |
|
|
|
I suggest everybody keep "good binaries": make a backup of the fastest .bin files you have, so you can recover them in case of driver problems.
Also this will enable you to get 1 o 2 percent more hashrate because of compiler variance (try removing the bin and running 3/4 times to see the variance in action).
Or use the provided bin file (see the OP) which should be a good one.
I've added a "final advices" section to the second post with the quoted text and this one: "I've experienced lower power usage with catalyst 14.9 compared to 14.6 beta (a bit less compared to 13 but still better). Speaking of optimization, this should be kept in mind: buy a power meter for your miner(s)!" |
|
|
|
|
I suggest everybody keep "good binaries": make a backup of the fastest .bin files you have, so you can recover them in case of driver problems.
Also this will enable you to get 1 o 2 percent more hashrate because of compiler variance (try removing the bin and running 3/4 times to see the variance in action).
Or use the provided bin file (see the OP) which should be a good one.
|
|
|
|
@st.neman Tried installing catalyst 14.9 and it totally borked hashrate using super kernel. Stick with catalyst 14.7 RC3  I've experienced the hashrate drop on the new drivers too. I'll look into it: there might need to be two versions of the kernel. You can still use the new drivers but the old bin files, it should work. |
|
|
|
1.12$/day / (0,87kW * 24h) = 0.0536$ / kWh average electricity cost is about 0.054$/kWh if I have <1600kWh for month Could I ask you in which country you live? That's crazy cheap electricity, I'm envious! :-) |
|
|
|
price did stabilize now between 0.00035 and 0.00040
i hope we rest here a week and then start slow natural growth
instead that fast increase we had now
+1 let's hope btc does the same ;-) |
|
|
|
Can also confirm lost 2 MH/s total mining X11 combined with my 3x R9 270s and my 1x R9 280X on 14.9 drivers
All right, then I'm sure I fucked up my test somehow. Did you delete the old .bin's from whatever drivers you had before, and had new ones generated for 14.9 drivers? Of course I did. I'm gonna re-do the test now, though. 14.6 vs. 14.9 on both my X11 kernels. Ooookay... this is odd. The hashrate changed - so it seems the driver did as well. Further, I fetched the date on fglrx each time before starting (14.6 being May 29 2014 and 14.9 being Sept 2 2014) so it DID change. As for the bin files, yes, I removed them; I double-checked. Anyway, Savage, this is what I've got, columns are GPU/clocks/mine/stock, rows are obviously GPUs. Running driver version 14.6:
270X - 1155/1500 - 3.64 - 2.80
290X - 1045/1600 - 7.96 - 4.86
7950 - 1155/1500 - 5.03 - 4.01
280X - 1155/1500 - 5.77 - 4.54
Running driver version 14.9:
270X - 1155/1500 - 3.64 - 2.80
290X - 1045/1600 - 7.98 - 4.95
7950 - 1155/1500 - 5.03 - 4.05
280X - 1155/1500 - 5.75 - 4.60
changes are too small: probably just compiler variance. compiling the same .cl file with the same settings may result in different hashrates (I believe a 2% range is common). the file size changes too. it doesn't happen with all the kernels but I'm sure it does on x11 as well as on my groestl, for example. EDIT: so long for those fine optimizations which result in 1% improvement and you think it's good... and it's nothing instead ;-) I don't understand your last comment. If you're saying my optimizations get 1% improvement, then you're reading the table wrong, because the third COLUMN is my kernel's hashrate, the fourth is the stock kernel's hashrate with the same GPU, driver, and clocks. Example: On driver version 14.9, my 290X achieved a hashrate of 7.98MH/s with my kernel, and 4.95MH/s with the kernel that is on SGMiner's master branch right now. The clocks were the same for both runs, 1045/1600. sorry, I didn't mean that. I was just saying that to me and other developers I know, it often happened that an optimization which led to a 1% improvement, or even more, in the end turned out to be just that variance, so not an optimization at all. It probably happened to you as well. |
|
|
|
Can also confirm lost 2 MH/s total mining X11 combined with my 3x R9 270s and my 1x R9 280X on 14.9 drivers
All right, then I'm sure I fucked up my test somehow. Did you delete the old .bin's from whatever drivers you had before, and had new ones generated for 14.9 drivers? Of course I did. I'm gonna re-do the test now, though. 14.6 vs. 14.9 on both my X11 kernels. Ooookay... this is odd. The hashrate changed - so it seems the driver did as well. Further, I fetched the date on fglrx each time before starting (14.6 being May 29 2014 and 14.9 being Sept 2 2014) so it DID change. As for the bin files, yes, I removed them; I double-checked. Anyway, Savage, this is what I've got, columns are GPU/clocks/mine/stock, rows are obviously GPUs. Running driver version 14.6:
270X - 1155/1500 - 3.64 - 2.80
290X - 1045/1600 - 7.96 - 4.86
7950 - 1155/1500 - 5.03 - 4.01
280X - 1155/1500 - 5.77 - 4.54
Running driver version 14.9:
270X - 1155/1500 - 3.64 - 2.80
290X - 1045/1600 - 7.98 - 4.95
7950 - 1155/1500 - 5.03 - 4.05
280X - 1155/1500 - 5.75 - 4.60
changes are too small: probably just compiler variance. compiling the same .cl file with the same settings may result in different hashrates (I believe a 2% range is common). the file size changes too. it doesn't happen with all the kernels but I'm sure it does on x11 as well as on my groestl, for example. EDIT: so long for those fine optimizations which result in 1% improvement and you think it's good... and it's nothing instead ;-) |
|
|
|
is the block explorer down? it says "coucou".
bump... |
|
|
|
I've found it is needed to keep GPU temp below 72C with the powercolor R9 280x, GPU clock gets throttled down >= 72C ...
72C looks pretty low to me... my 290s throttles at 95 and 85. |
|
|
|
mmmm ... wonderful odor, I am steam curing a pound of virginia tobacco in oven for my own version of dark molasses cavendish ... nice rich smooth smoke and it's getting really dark after 8 hr sample.
I'm wondering what it has to do with this thread. Still, thanks for the bump ;-) |
|
|
|
|
If someone is interested, I've compiled the headless wallet on arm (for my QNAP nas).
If there is interest I can post a compile guide.
I can as well make a static version which works on many arm devices including android.
|
|
|
|
|
Show Posts
