Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 544. (Read 6590718 times)

just some info to sum things up:
I've got various 570s (hashing at 28.5-30MH) and vegas 64(40-41MH) running on winX anniversary with drivers: 17.10.2, 17.10.3, 17.11.1
Using overdriven tool for proper (not the same values for various stages, as claymore does) frequencies and volts - works great!
(actually I use it for vegas FE with 2017Q4 driver as well)

NEVER had an issue with setting "compute mode" - just fucking ignore the AMD GUI crap

Also for all those lazy people using AB - just DON'T. It should be the tool of last resort - only if nothing else works and immediate results are required... and can be achieved without fucking everything up.
Seriously. I'm tired of repeating myself on this issue. Don't use AB for AMD ever.

Sounds like it. There is a documented exploit for the Claymore Remote Manager API.

https://medium.com/@s3yfullah/hacking-cryptocurrency-miners-with-osint-techniques-677bbb3e0157

I would close any open ports on your router and change the default port of the Ethman manangement tool.

I am receiving those errors from time to time:

Remote management: read-only mode, command miner_getfile ignored
Remote management: read-only only, command miner_reboot ignored

Does someone hacked me?

I would not be surprised to see a hybrid POW/POS implementation sooner than people think. At the last Ethereum developer meeting it was discussed preparation for the POW/POS implementation is at a point where it could be released in the next hard fork, Constantinople, which should be within 6-9 months.

https://www.youtube.com/watch?v=ZN-AtGgtmtA&feature=youtu.be&t=2297

Guys, anybody have the exact information about ethereum switching from Pow to Pos? I just read an article that Casper is coming earlier than planned and maybe with the next ethereum upgrade (maybe even before the end of year or early in 2018). What exactly that means for us miners? Some say that it will be hybrid Pow/Pos... But, if it is just Pos, and if it happens in a couple of months, are we doomed?

any one who got a solution for afterburner not enabled controls with lastest driver 17.11.1

Ok, thank you for your suggestion. I will try.

I expressed only a guess that your problem is with the VRAM. If the GPU fails during miner initialization (OpenCL loading, DAG file generation), that means either the software does not allow proper initialization (check drivers, OS, recheck BIOS if flashed) or the GPU has massive hardware faults. If the memory has been fried earlier, it's too late. BTW, the VRAM chips in XFX GTR look covered by a bracket (most probably - of plastic, conducts heat badly) and that differs from XFX RS, where VRAM chips are uncovered.
Try connecting only the XFX RX480 to the MB, directly in the first PCI-E x16 slot and run a miner instance.

I had the samw issue (check here)

In my case, clean installation of Windows was the only solution. Make sure you disable windows updates because it will happen again if you don't (happened to me)

My card is XFX rx480 gtr xxx, is there the cooling situation equal to rx480 RS?  Do you think,  if i find a way to cool the vrams, my problem will be solved? Or are the vrams gone now?

Suddenly Claymore is weak. for you too?

Guys,

anybody have problem like this?

One of my Sapphire Rx 580Nitro+ 8Gb cards result this screen after I start Claymore. Of course I check all my cards, and found this one card which freeze windows. The card was bios mooded but I try also with his original bios and the result was the same.  

https://imgur.com/a/I1lr1

10:51:38:330   624   ETH: 1 pool is specified
10:51:38:334   624   Main Ethereum pool is eu1.ethermine.org:14444
10:51:40:888   624   OpenCL platform: AMD Accelerated Parallel Processing
10:51:40:893   624   OpenCL initializing...

10:51:40:896   624   AMD Cards available: 6
10:51:40:901   624   GPU #0: Ellesmere, 8192 MB available, 36 compute units
10:51:40:904   624   GPU #0 recognized as Radeon RX 480/580
10:51:40:908   624   GPU #1: Ellesmere, 8192 MB available, 36 compute units
10:51:40:912   624   GPU #1 recognized as Radeon RX 480/580
10:51:40:915   624   GPU #2: Ellesmere, 8192 MB available, 36 compute units
10:51:40:920   624   GPU #2 recognized as Radeon RX 480/580
10:51:40:923   624   GPU #3: Ellesmere, 8192 MB available, 36 compute units
10:51:40:927   624   GPU #3 recognized as Radeon RX 480/580
10:51:40:931   624   GPU #4: Ellesmere, 8192 MB available, 36 compute units
10:51:40:935   624   GPU #4 recognized as Radeon RX 480/580
10:51:40:939   624   GPU #5: Ellesmere, 8192 MB available, 36 compute units
10:51:40:943   624   GPU #5 recognized as Radeon RX 480/580
10:51:40:947   624   POOL/SOLO version
10:51:40:951   624   b421
10:51:40:955   624   Platform: Windows
10:51:40:993   624   start building OpenCL program for GPU 0...
10:51:41:289   624   done
10:51:41:496   624   start building OpenCL program for GPU 1... THAN freeze....


Any idea?

What is the difficulty adjustment mechanism in Siacoin and Decred?

update:
Decred community considers to project and manufacture an ASIC as well. Provided both Obelisks and Decred community ASICs will use 28nm chips, their energy efficiency will be close.
Actual Decred hashrate is ~350 TH/s, equivalent of 750 Obelisk DCR1.
Actual Siacoin hashrate is ~150 TH/s, equivalent of 500 Obelisk SC1.
I expect Decred hashrate to increase by x2-x5 till July 2018, before the release of Obelisk ASICs (unless Obelisk turns them on earlier, just like Chinese ASIC makers do). When Obelisks come online, network hashrate will double-triple within 2-3 weeks. GPU miners will be left nothing to catch 
Personally I would prefer a Decred ASIC over a Siacoin one - the underlying project offers more use cases.

XFX cards have poor cooling systems, the VRAM overheats seriously because it does not attach to heatsinks (in XFX RX480 RS in particular). I have measured temps up to 110-115C on VRAM, while the reported core temp was 65C.
I think your VRAM might be overheated or has been overheated in the past and VRAM chips started to fail, hence the random failures. Especially if your XFX was working under high temperature.

Hi, I'm using claymore 10.1. Sometimes my pc restart and after restart my xfx rx480 8gb isn't recognized in the system anymore (my 2 sapphire rx480 are shown) . The card is on main pci express on motherboard. When I turn off the pc with the button on the case and turn it on after, the card is shown in system and works like normal. What to do?

Kinda offtopic: what do you guys think of the Obelisk DCR ASIC? Sia too for that matter...gonna hurt those mining options.

Was wondering if someone could help me out with 2 strange problems.  (Using Claymore 10.0 but the same issue has carried over through 9.5 and 9.7)

1.  No matter what my config.txt file is, the miner reports it has trouble parsing line #14.

I have rearranged the lines in the file, added lines, even comment out all the flags.  Same parsing error.

2.  The clock I set for core and memory do not "take" until it is run a second time after reboot.  If I reboot my system, and then run Claymore's miner, it will mine at its original clock rates, not those I have specified in the config.txt file.  Upon the 2nd run after reboot (and all subsequent runs) it does run with the correct clocks.

Thanks in advance.Was wondering if someone could help me out with 2 strange problems.  (Using Claymore 10.0 but the same issue has carried over through 9.5 and 9.7)

1.  No matter what my config.txt file is, the miner reports it has trouble parsing line #14.

I have rearranged the lines in the file, added lines, even comment out all the flags.  Same parsing error.

2.  The clock I set for core and memory do not "take" until it is run a second time after reboot.  If I reboot my system, and then run Claymore's miner, it will mine at its original clock rates, not those I have specified in the config.txt file.  Upon the 2nd run after reboot (and all subsequent runs) it does run with the correct clocks.

Thanks in advance.

Blockchain driver did not work with windows 10 FCU for me. I would loose 5 Mh/s per card for RX580 rig with FCT and blockchain driver combination and the system is not stable either. I have one rig upgraded to FCU and I had to use 17.11.1 to keep the same hash rate. I don't really like the upgrade because I cannot use AB and the fan setting is not working properly. we are talking about days and weeks long run without restarting the rig not reinstall windows or driver every one or two days because you would lose 20-30% of your income if keep doing that.

Yes, I absolutely agree.
Also, many people should have the patience to read carefully at least some previous posts where they might find answers to their questions or useful information to solve their issues.

The problem is uninstalling drivers and installing FCU is not idiot proof so people are going to have issues.  You'd be surprised how many people just over write drivers and expect it to work.