t0mpr1c3
Newbie
Offline
Activity: 13
Merit: 0
|
|
June 02, 2014, 05:31:59 PM |
|
Thanks. Those settings give me 0 WU and many HW errors on catalyst 13.12 under Linux.
Edit: To clarify, I am running Ubuntu 12.04 32 bit.
You have wrong kernel from sources (x13mod.cl = 31 KB), working kernel for 5850 and binary for windows is in my link (x13mod.cl = 281 KB). Deleted old *.bin files, recompiled with the larger x13mod kernel, and it's giving me "Error -46: Creating Kernel from program. (clCreateKernel #name)" followed by a segmentation fault. @t0mpr1c3 - read that : The big X13mod, works on Radeons 5xxx-6xxx. Its obvious that this is using "hacked" x11mod from Girino's exe (with opencl file >270KB) example cmd line for 5850 sgminer -k x13mod -o stratum+tcp://boost.suprnova.cc:3311 --auto-fan -w 256 -g 1 -I 18 --thread-concurrency 7200 --lookup-gap 2 Works with Cat 13.12, works with 14.4 too. tested on 5850, 6970 and 5750 x13mod.cl - 281KB https://dl.dropboxusercontent.com/u/88276654/x13mod.7zSorry but can't remember how I found it, was posted somewhere in a forum - with complete sgminer build, made 28-05 early in the morning Cheesy PS: https://dl.dropboxusercontent.com/u/88276654/sgminerx11.zip thats the file I downloaded, the x13mod kernel inside is NOT compatible with later sgminer builds, so ... OK thanks for the info.
|
|
|
|
flex65
|
|
June 02, 2014, 06:40:03 PM |
|
i'm trying x13mod and have nothing )) first ver crashes (problems with kernel), second ver - one of my cars has many HW. does anyone have working x13mod?
|
|
|
|
qunity3d
Member
Offline
Activity: 100
Merit: 10
|
|
June 02, 2014, 06:55:32 PM |
|
I use -k x13mod, but have error: Error -11: Building Program (clBuildProgram) "/usr/local/bin/simd.cl", line 37: warning: incompatible redefinition of macro "ROL32" (declared at line 6 of "/usr/local/bin/skein.cl") #define ROL32 SPH_ROTL32 "/tmp/OCLNDA4lf.cl", line 325: error: function "UBI_BIG" declared Failed to init GPU thread 0, disabling device 0.
|
|
|
|
Sehae
Newbie
Offline
Activity: 29
Merit: 0
|
|
June 02, 2014, 07:06:59 PM |
|
Has anyone got Boostcoin at x13 running properly on 280x's? I seem to fail hard at getting thigs to work with my rigs? Could someone please post the config file he uses and the version ( I try with the x13 from here lol)? Cheers
|
|
|
|
Bojcha
|
|
June 02, 2014, 08:30:13 PM |
|
That working x13 kernel for 5xxx and 6xxx cards is not near close hashrate how it should be. Posted from Bitcointa.lk - #Af5Y9PGMa4GNVbSN
|
|
|
|
speedhash.eu
|
|
June 02, 2014, 08:54:48 PM |
|
Two quick questions:
- does the newest PIMP (June 1st v1.4.2) support X13 and X11_mod (better hashrates) without problems? - what is the power consumption for X13? Smiliar to X11 (~ 50% of Scrypt)?
Thanks for help in advance!
|
|
|
|
ekoice
|
|
June 02, 2014, 09:18:22 PM |
|
x13 : x11: cheers.
|
|
|
|
punisher1
Newbie
Offline
Activity: 25
Merit: 0
|
|
June 02, 2014, 09:21:38 PM |
|
i have a nvidia 650ti and windows 7 x64 . what i need to mine x13 and what settings should i use ? thank you !
|
|
|
|
bullus
|
|
June 02, 2014, 09:26:24 PM Last edit: June 02, 2014, 09:38:03 PM by bullus |
|
Thanks. Those settings give me 0 WU and many HW errors on catalyst 13.12 under Linux.
Edit: To clarify, I am running Ubuntu 12.04 32 bit.
You have wrong kernel from sources (x13mod.cl = 31 KB), working kernel for 5850 and binary for windows is in my link (x13mod.cl = 281 KB). Deleted old *.bin files, recompiled with the larger x13mod kernel, and it's giving me "Error -46: Creating Kernel from program. (clCreateKernel #name)" followed by a segmentation fault. @t0mpr1c3 - read that : The big X13mod, works on Radeons 5xxx-6xxx. Its obvious that this is using "hacked" x11mod from Girino's exe (with opencl file >270KB) example cmd line for 5850 sgminer -k x13mod -o stratum+tcp://boost.suprnova.cc:3311 --auto-fan -w 256 -g 1 -I 18 --thread-concurrency 7200 --lookup-gap 2 Works with Cat 13.12, works with 14.4 too. tested on 5850, 6970 and 5750 x13mod.cl - 281KB https://dl.dropboxusercontent.com/u/88276654/x13mod.7zSorry but can't remember how I found it, was posted somewhere in a forum - with complete sgminer build, made 28-05 early in the morning Cheesy PS: https://dl.dropboxusercontent.com/u/88276654/sgminerx11.zip thats the file I downloaded, the x13mod kernel inside is NOT compatible with later sgminer builds, so ... OK thanks for the info. Thanks, tested your 2nd link and the hd6970 is hashing and recieved accepted shares and no HW-errors btw is this the miner with the 2% fee?? sgminer 4.1.0 - Started: [2014-06-02 23:22:07] -------------------------------------------------------------------------------- (5s):1.371M (avg):1.381Mh/s | A:16 R:0 HW:0 WU:0.014/m ST: 1 SS: 0 NB: 6 LW: 262 GF: 0 RF: 0 Connected to am02.eu.trademybit.com:5550 (stratum) diff 0.003 as user xxxx Block: 37c56433... Diff:602 Started: [23:25:07] Best share: 0.057 -------------------------------------------------------------------------------- [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit GPU 0: 57.5C 2466RPM | 1.382M/1.387Mh/s | R: 0.0% HW:0 WU:0.014/m I:18 -------------------------------------------------------------------------------- [23:23:51] Accepted 95ec8d27 Diff 0.007/0.003 GPU 0 at am02 tmb x13 [23:24:02] Accepted dc3dc4da Diff 0.005/0.003 GPU 0 at am02 tmb x13 [23:24:12] Network diff set to 592 [23:24:12] New block detected on network before pool notification [23:24:24] Accepted 6f611824 Diff 0.009/0.003 GPU 0 at am02 tmb x13 [23:24:37] Network diff set to 584 [23:24:37] Stratum from am02 tmb x13 detected new block [23:24:42] Accepted cf8027fe Diff 0.005/0.003 GPU 0 at am02 tmb x13
Is the hashrate ok? 6970@960/1375 "kernel" : "x13mod", "api-allow" : "W:127.0.0.1", "api-listen" : true, "api-mcast-port" : "4028", "api-port" : "4028", "expiry" : "28", "failover-only" : true, "kernel-path" : "/usr/local/bin", "log" : "5", "queue" : "1", "scan-time" : "7", "scrypt" : true, "shares" : "0", "failover-switch-delay" : "10", "gpu-threads" : "2", "gpu-dyninterval" : "7", "gpu-engine" : "0", "gpu-fan" : "55", "gpu-platform" : "0", "gpu-memclock" : "0", "gpu-memdiff" : "0", "gpu-powertune" : "8", "gpu-vddc" : "0", "intensity" : "18", "temp-target" : "75", "temp-overheat" : "85", "temp-cutoff" : "95", "temp-hysteresis" : "3", "vectors" : "1", "worksize" : "64", "lookup-gap" : "0", "shaders" : "1536", "thread-concurrency" : "8000", "no-pool-disable" : true, "device" : "0"
|
btc 16LWhms487bzCxQWq5oeW8SMDPmU2rvQjf
|
|
|
hashbrown9000
|
|
June 02, 2014, 10:05:59 PM |
|
For the guys running R9 270 cards: I had to switch over to girino's 2% version under BAMT 1.6. To get any sort of stability past 8 hours of mining. Now my rigs are rock stable at near 1965 khs per 270x. A hair lower with 270 non-x.
Note all my cards are undervolted in the BIOS to 1.100v using VBE7. I could probably push them to 2.2 Mhs if they were still stock @ 1.250v, however electric rates are going up as we speak where I am.
My settings off the top of my head:
Xintensity: 512 Gpu-threads: 2 Lookup-gap: 2 Thread-concurrency: 5121 Gpu-engine: 1110 Gpu-memckock: 1250 Gpu-powertune: 20 Auto-fan: true Shaders: 1280 Gpu-fan: 35-85
If someone can verify Pimp can run the free x11mod and x13mod on 270 and 270x cards without needing to hard reboot after a few hours I can switch back. But for now I'm stable with girino
|
Pinkcoin: ETH: VTC: BTC:
|
|
|
treyson
Newbie
Offline
Activity: 12
Merit: 0
|
|
June 02, 2014, 10:06:25 PM |
|
On x13mod on a 290 rig i've found that my memory speed is bouncing between what I have it set at, 1250mhz, and 150mhz. It continually drops to 150mhz and then regains to 1250mhz. It happens roughly every 15seconds. Still hashes, still submits, but this appears to be dropping hashing speed over time.
My 280x rig doesn't do this with same software.
Anybody else seeing this and any ideas?
This was fixed earlier by downloading MSI AB and then checking Unofficial overclock in the Settings. Apparently this has something to do with the powertune. I tried this without success. I wonder if I need to set anything specific for powertune? I'm at 0 now, but issue persisted at 20 as well. The powertune lead makes sense because it appears memory is downclocking or throttling, though I'm not overclocking the cards (950c/1250m) and volts are untouched. The issue has continued through AMD Drivers 13.12's, 14.4's and i'm on MSI AB beta 19.
|
|
|
|
NRD
Newbie
Offline
Activity: 37
Merit: 0
|
|
June 02, 2014, 10:09:09 PM Last edit: June 02, 2014, 10:22:48 PM by NRD |
|
i have a nvidia 650ti and windows 7 x64 . what i need to mine x13 and what settings should i use ? thank you !
No x13 on NV yet as far as I know, check the CC/cudaminer thread. They'll probably be adding it soon enough. I never tried running more threads as I wanted to see how stable this setup is. Here's a 23 hour run: http://i152.photobucket.com/albums/s196/nicknrd/Crypto/x11-1438mh-23h.png3 X 290 tri-x 1080/1500 1 X 7870 OC 1300/1500 setx GPU_MAX_ALLOC_PERCENT 100 setx GPU_USE_SYNC_OBJECTS 1 export GPU_FORCE_64BIT_PTR=1 ##export GPU_MAX_HEAP_SIZE=50 del *.bin timeout /t 4 sgminer.exe -k x11mod --rawintensity 1306624,1306624,1306624,1048576 -g 1,1,1,2 --shaders 2560,2560,2560,1280 -w 256,256,256,64 --no-submit-stale -c sgminer.conf *Note, I removed TC from my post because it's irrelevant with X11, however I still have it in my .bat file.
|
|
|
|
RoyalSands
|
|
June 02, 2014, 10:25:11 PM Last edit: June 03, 2014, 12:13:03 AM by RoyalSands |
|
For the guys running R9 270 cards: I had to switch over to girino's 2% version under BAMT 1.6. To get any sort of stability past 8 hours of mining. Now my rigs are rock stable at near 1965 khs per 270x. A hair lower with 270 non-x.
Note all my cards are undervolted in the BIOS to 1.100v using VBE7. I could probably push them to 2.2 Mhs if they were still stock @ 1.250v, however electric rates are going up as we speak where I am.
My settings off the top of my head:
Xintensity: 512 Gpu-threads: 2 Lookup-gap: 2 Thread-concurrency: 5121 Gpu-engine: 1110 Gpu-memckock: 1250 Gpu-powertune: 20 Auto-fan: true Shaders: 1280 Gpu-fan: 35-85
If someone can verify Pimp can run the free x11mod and x13mod on 270 and 270x cards without needing to hard reboot after a few hours I can switch back. But for now I'm stable with girino
"1965 khs per 270x" Is that mining X11? Because if it is then that is horribly low for a 270x.
|
|
|
|
virtualista
Newbie
Offline
Activity: 22
Merit: 0
|
|
June 02, 2014, 10:33:20 PM Last edit: June 02, 2014, 11:10:40 PM by virtualista |
|
X13 MSI 280X 3G 2.515+ Mh/s "intensity" : "16", "worksize" : "256", "kernel" :"marucoinmod", "lookup-gap" : "2", "thread-concurrency" : "8192", "shaders" : "2048", "gpu-threads" : "2", "gpu-engine" : "1150", "gpu-memclock" : "1300", "gpu-memdiff" : "0", "gpu-powertune" : "20" https://i.imgur.com/Q5GiBON.png
|
|
|
|
mrbrdo
|
|
June 03, 2014, 12:14:40 AM |
|
Hey guys, Since we merged all (or most) functionality from sph-sgminer into official sgminer ( https://github.com/sgminer-dev/sgminer/tree/v5_0), and also added both X11-mod and X13-mod kernels, it would be cool if we can work together there, so we can keep all the improvements in one place. I am happy to merge any improvements from this version into there. Also, I would be really interested to hear if you guys still get the SICK problem on our v5_0 branch (we did not increase the sick time)? Since there are many new bugfixes there and it could be that the issue is not directly caused by X11-mod. Personally I haven't had any SICK yet (I don't over-do it on the GPU settings though).
|
|
|
|
hero18688
|
|
June 03, 2014, 02:47:50 AM |
|
290 for x13mod,please.I got only the same hash as 280x.And it is sick minutes after. Posted from Bitcointa.lk - #1YvS5J6uNUEVz5KG
|
|
|
|
goodluck0319
|
|
June 03, 2014, 03:03:57 AM |
|
me too. I cannot get pass the 2.2-2.4 MH/s mark.
|
|
|
|
bengx
|
|
June 03, 2014, 03:35:11 AM |
|
Thanks for the kernel. PM your bitcoin address to me?
|
|
|
|
jenveno
Member
Offline
Activity: 350
Merit: 10
|
|
June 03, 2014, 04:21:55 AM |
|
Coming back to share some settings, as promised, now that I've found something that seems to be working stable. Gigabyte R9-270 | win7 64, amd catalyst 14.4 | on x13mod hashing ~1.65mh/s per card "intensity" : "18", "kernel" : "x13mod", "worksize" : "128", "lookup-gap" : "2", "gpu-threads" : "2", "shaders" : "1280", "gpu-engine" : "1110", "gpu-memclock" : "1475", "gpu-powertune" : "20" Sapphire R9-280X | Xubuntu 12.04, amd catalyst 14.4 | on x13mod hashing ~2.65mh/s "intensity" : "18", "kernel" : "x13mod", "worksize" : "128", "lookup-gap" : "2", "gpu-threads" : "2" "shaders" : "2048", "gpu-engine" : "1160", "gpu-memclock" : "1500", "gpu-powertune" : "20" all you gotta do is add your pool/fan/temp details as normal and it should blast off. Note that I have not thoroughly tested these for days on end yet. I'll report back if shit goes haywire and I need to reconfigure.
|
|
|
|
flex65
|
|
June 03, 2014, 05:49:14 AM |
|
Thanks for the kernel. PM your bitcoin address to me? with this shminer one card has +100500 HW, dont know why...
|
|
|
|
|