Topic: SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching - page 117. (Read 237247 times)

No. XMR-Stak and Cast_XMR work with a little bit lower hash rate.

seriously?

Hey did you get a resolution to this? I also am having this problem. 1 V56 and 2 V64

Were you disabling P0-P6 in ODT? You should. Just click each one to grey it out.
Also ODT version 0.2.6 is very buggy for me, it has a problem setting core properly so i downgraded to 0.2.5 which works perfectly.
What I do is enable "extend official overclock limits" in Afterburner, then from there I only use ODT 0.2.5, you have to do that or you get mem clock setting error in ODT.

OK, so ~40W at wall, that's more realistic.

I found my 550/560s need more voltage than 570/580s for the same freq to be stable, usually 1000mV+ for 1250MHz. But OK, maybe MSI 560 is a good series.

I have msi 560 that came with hynix mem and later core steppings, 1050 mV at stock core clock 1196.
They are great undervolters, doing 895 mV at 1180 on v8, they could do 885 mV with v7. gpuz reads a solid 36W, kill-a-watt doesn't go over 40W, bounces between 36-40W.
They were doing 29-30W on v7, so power consumption went up around 20%.

i think the http://cryptonightv8.com is the best and easiest

@playfast
I don't wanna be spiteful, but 450h@36W(at the wall) is barely possible.
As all polaris GPUs are 100% consistent in relative consumption it would mean an RX 570 can do 900 h/s at 72W but it can't. (or very rarely at least, 1 out of 100 cards maybe)
Did you measure it with killawatt or it's some 'calculated' or GPU-Z value?

Thanks for the reply - you got my last sMerit for it!

After making that post I poked around the log file created by SRB and found the intensity it was setting for the cards: all are at 26.0. I was using OverDrivenTool to set all the clocks and voltages for each power state but I found it cumbersome and once MSI AB was able to deal with 3+ AMD cards again (ie - v4.5.0) I switched back to it. Unfortunately, I haven't been able to unlock core voltage adjustment yet, and as I recall that was another reason why I switched from AB to ODT.

At any rate, my 4 cards are consistently doing 460, 420, 420, 460 and using an estimated 54W each (based on Kill-A-Watt monitor, minus the amount used by the system when not mining) and 1200 Core / 1900 Mem and -20% on the power limit, so you are totally kicking my ass in power consumption vs. hashrate. Clearly I have some more tweaking to do but right now I am just happy I got it working stably on v8.

I have RX 560 2gb rig mining XMR. I use manual intensity, all cards needed lower intensity than the auto intensity. I recommend start at 26 and keep decreasing by 0.5 increment if not stable. Some of my cards had to go down to 24 and the lowest is 23.5
Core 1180@895mV/Mem 2000-2100, power limit +10% each card running 450-465 H/s and using ~36W with 18.6.1, the best driver for me. My cards needed +5 mV on the memory compared to v7. I tried 18.5.x it hashes a tiny bit more but uses more power ~40W
v8 is more hungry for virtual memory, I have it set at 3x total vram.
I also recommend using a reasonably low static difficulty with XMR, variable difficulty can get too high and cause stability/pool problems with weaker cards like the 560.

SRB html monitor https://www.memrey.com/note/srb-dash-mon

https://imgur.com/a/eFYJnqr

[2018-10-26 01:00:28] GPU0 :    984 H/s [T: 53c, RPM: 2038, CC: 1452 MHz, MC: 875 MHz][BUS:14]
[2018-10-26 01:00:28] GPU1 :    1046 H/s [T: 50c, RPM: 1873, CC: 1452 MHz, MC: 875 MHz][BUS:7]
[2018-10-26 01:00:28] GPU2 :    1150 H/s [T: 54c, RPM: 1899, CC: 1452 MHz, MC: 875 MHz][BUS:3]
[2018-10-26 01:00:28] GPU3 :    1275 H/s [T: 50c, RPM: 0, CC: 1452 MHz, MC: 875 MHz][BUS:11] - vega56, why low, driver 18.6.1 "intensity" : 110, "worksize" : 16, "threads" : 2

I am using the "guide to reach 2k+ on v8 on vega 64"

Got 2 64s and 1 56.
The 56 is in slot 003, however upon launching the miner I get the following error:

2018-10-24 00:56:27] Loading kernel for DEVICE BUS ID 8 ...

[2018-10-24 00:56:27] Error CL_BUILD_PROGRAM_FAILURE when calling clBuildProgram [ctx->Program] for DeviceID 0 (Thread 0)

I got pretty familiar with XMR Stak but I have been having lots of stability issues with it so wanted to try the SRB miner. I'm sure it is a simple fix, just don't know where to start.



Any pointers are appreciated.

Any news on support for posible integration CryptoNight-Webchain?

Please help. I have one Vega56 and one RX580-8G. Got the following error when running SRBMinber 1.6.8. Tried different drivers (18.5.1, 18.5.2, 18.6.1), DDU clean install every time. Also tried the default kernel and compiled kernel.  One wired thing I noticed is that the miner recognizes my RX580 as "Radeon RX Vega [ellesmere]". The miner is working properly only when I use one card at a time to hash with fresh compiled kernel.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
[2018-10-25 14:17:10] Miner version: 1.6.8
[2018-10-25 14:17:10] Windows version: 10.0 build 17763
[2018-10-25 14:17:10] Video driver version: 24.20.11021.1000
[2018-10-25 14:17:11] AMD Platform ID: 0
[2018-10-25 14:17:11] AMD platform FOUND
[2018-10-25 14:17:11] Found 2 AMD devices
[2018-10-25 14:17:11] CPU AES-NI: FALSE
[2018-10-25 14:17:11] GPU0: Radeon RX Vega [ellesmere] [8192 MB][Intensity 61.0][W: 16][T: 2][K: 1][BUS: 15]
[2018-10-25 14:17:11] GPU1: Radeon RX Vega [gfx900] [8176 MB][Intensity 120.0][W: 16][T: 2][K: 1][BUS: 5]
[2018-10-25 14:17:11] ADL is enabled
[2018-10-25 14:17:11] CryptonightV8 mode enabled
[2018-10-25 14:17:11] Starting init of mining threads
[2018-10-25 14:17:11] Heating up system, please wait...
[2018-10-25 14:17:11] Created OCL context
[2018-10-25 14:17:11] Created OCL command queue for DeviceID 0 (Thread 0)
[2018-10-25 14:17:11] Created OCL input buffer for DeviceID 0 (Thread 0)
[2018-10-25 14:17:11] Created OCL output buffer for DeviceID 0 (Thread 0)
[2018-10-25 14:17:12] Loading kernel for DEVICE BUS ID 15 ...
[2018-10-25 14:17:12] Error CL_BUILD_PROGRAM_FAILURE when calling clBuildProgram [ctx->Program] for DeviceID 0 (Thread 0)
[2018-10-25 14:17:12] Error initing GPU's. Stopping miner process

Try 59.5 and other values with decimals. 5 minutes is too small period of time to stabilize speed especcially on heavy algo. You must every try wait at least 10-15 minutes. Sometimes speed is became stable after 10-12-15 minutes. And often it became faster!

59 is the Max, but tried less intensity, the total hashrate is random when wait 5 minuts mining

What intensity you're use?

I deleted the kernels, used ddu uninstaller, instaled driver 18.6.1 and same problema, im crazy with this problem, Mining monero v7 berore same problem with the hash

that didnt solved the problem... adding 2 thread reduces hs to 450

looking a bios flash to see if there is something wrong...