GUI mining - updated Dec 3 with 7970 bugfix, also supports Stratum!

[fizgig]

no,
on pools you work on easy hashes,
means say the current requirement is for 14 leading zeros in the hash to solve a block, you are trying to find hashes with say at least 3 leading zeros.
you'll find and submit hashes with 3,4,5,6,7,8,9..... leading zeros (which all get you a share in the pool) and every once in a while someone finds a hash with the required 14 leading zeros and bang, pool gets the block, everyones happy.

Interesting.  Why would you submit easy hashes since, I assume, easy hashes will never be accepted into the block chain?  Is it to demonstrate that you actually are trying to solve the problem and not just tagging along doing nothing?

And what does the "1" in "Difficulty 1" mean? 

[1714225593]

1714225593

[1714225593]

1714225593

[1714225593]

1714225593

[1714225593]

1714225593

[1714225593]

1714225593

[BitLex]

Interesting.  Why would you submit easy hashes

as a proof-of-work that your miners did

And what does the "1" in "Difficulty 1" mean? 

it means that it's (currently)    55590times easier to find a winning hash on pools than on the main-bitcoin-network,
because the difficulty to find one on the main-bitcoin-network is 55590

[Kiv]

no,
on pools you work on easy hashes,
means say the current requirement is for 14 leading zeros in the hash to solve a block, you are trying to find hashes with say at least 3 leading zeros.
you'll find and submit hashes with 3,4,5,6,7,8,9..... leading zeros (which all get you a share in the pool) and every once in a while someone finds a hash with the required 14 leading zeros and bang, pool gets the block, everyones happy.

Question: Can we run out of difficulty?

Theoretically, we could reach a maximum difficulty. The hardest possible difficulty would mean to solve a block you need to find a hash that is ALL zeroes. Since hashes can be any number up to 2 to the power of 256, this would take longer than the lifetime of the universe to find using current methods. The only way we would ever reach this is if someone found a vastly superior method of finding hashes to the current brute force method. Such a person would probably keep their method to themselves and make a fortune off of Bitcoin

[friendsofkim]

@Kiv

Thanks for the GUI. I've sent 0.5 BTC your way.

[BitterTea]

I'm happy with the UI as well. Actually, I like it better than the one I created. I've sent a small donation your way.

[smgoller]

Trying to get this to work. I've got a Radeon 5870, and I was able to successfully run demos using GPU Caps Viewer, and it definitely says it's capable. But when I run guiminer, I get this error in guiminer.exe.log:

Code:

Exception TypeErrorERROR:root:Exception:
Traceback (most recent call last):
  File "guiminer.py", line 836, in <module>
  File "guiminer.py", line 518, in __init__
SystemExit: 1

It thinks I have no OpenCL devices.

I have catalyst 10.11 and stream sdk 2.1 installed. Is there something I'm missing?

[Kiv]

I haven't tried it with stream sdk 2.1. Maybe m0mchil can confirm this, but perhaps stream sdk 2.2 or higher is required?

Trying to get this to work. I've got a Radeon 5870, and I was able to successfully run demos using GPU Caps Viewer, and it definitely says it's capable. But when I run guiminer, I get this error in guiminer.exe.log:

Code:

Exception TypeErrorERROR:root:Exception:
Traceback (most recent call last):
  File "guiminer.py", line 836, in <module>
  File "guiminer.py", line 518, in __init__
SystemExit: 1

It thinks I have no OpenCL devices.

I have catalyst 10.11 and stream sdk 2.1 installed. Is there something I'm missing?

[buller]

Tested it for 1 day. Working perfectly, easy to use, same performance

XFX Ati HD5870 no overclocking 315 mhash average

[Grinder]

Just installed it, and now I have a bunch of feature requests.
- A summary page which shows all the miners with a stop/start button and the accepted/stale statistics.
- Stale/invalid the last 30 minutes (so I can switch to a different pool if the one I'm using seems broken).
- An option to have it automatically switch to a different pool if there have been too many stale/invalid the last 30 mins.
- Statistics which shows how much time has been used doing nothing because of server down time.

It seems really nice already, though. Perhaps I'll finally stop wasting time looking at the miner window now that I can minimize it to tray. I'll donate if I continue using it.

[Kiv]

Glad you liked it and I appreciate the feedback. Some thoughts:

- A summary page which shows all the miners with a stop/start button and the accepted/stale statistics.

Sure, this is a good idea. I will make this a toggleable feature (like the console) since I suspect that the average user only has one or two miners going. Maybe this information should also be available in the tray icon, I don't know.

- An option to have it automatically switch to a different pool if there have been too many stale/invalid the last 30 mins.

The problem with switching based on stale is that currently deepbit doesn't report when a stale share is submitted. So even though slush might report 1% of stale, it's possible that deepbit has a much higher stale percentage. If deepbit starts to report stale shares, I'll revisit this.

I also know that some people are working on "push" mining where you don't have to poll the server so much. I wonder if that will make stale shares a non-issue; maybe someone more knowledgable can comment on that.

- Statistics which shows how much time has been used doing nothing because of server down time.

What I do is run two tabs at once (slush and deepbit right now) on a single device. Then if either server is down for any amount of time, the other one will still be able to use the full GPU.

The only issue with this method is there's no way to control the priority of each miner - each gets a random amount of time. I would like to look into this in the future but right now I have no idea how OpenCL tasks are scheduled on the GPU.

My hope is that as pools grow more sophisticated, server down time will be reduced enough that we won't have to worry about it.

I'll donate if I continue using it.

Great, any amount makes me feel appreciated I'm hoping to release a new version in the next couple days, so stay tuned.

[Kiv]

Hey everyone, a new version of the GUI is available:

    poclbm-gui-20110306.7z

Changes since the previous version:

- New summary tab lists information about all your miners on one page as well as total hash rate.
- Miners can be started automatically when the GUI starts; this option can be checked/unchecked from the summary page.
- Smaller download (removed useless Tk files).
- Miner tabs show on the status bar the last time a share was accepted.
- Update poclbm to m0mchil's latest version 201103.beta.

Someone had asked about an option to start the GUI when Windows starts; this can be done easily from within Windows by making a shortcut to guiminer.exe in the "Startup" folder of the Start Menu.

Enjoy!

[JWU42]

Excellent Kiv - Great Update!

[Beremat]

Nice, the only thing I was waiting for was the autostarting of miners. Great work on this GUI!

[purpleeggguy]

I have a problem with the gui, it only seems to max out the GPU at around 30% even though the non-gui version maxes out at 100%
I'm using an Nvidia card with the latest drivers and windows 7.

[hazek]

I have a problem with the gui, it only seems to max out the GPU at around 30% even though the non-gui version maxes out at 100%
I'm using an Nvidia card with the latest drivers and windows 7.

Is there actually a way to set what % of power GPU should use for hashing?

[purpleeggguy]

Is there actually a way to set what % of power GPU should use for hashing?

I supposed you could limit it by setting -w to something ridiculously low

[Kiv]

That's odd... you're using the same flags in both cases? If so, can you try running the poclbm.exe that comes with the GUI and see if that does the same thing? I don't have an Nvidia card to test on but I haven't heard of anyone else with your problem.

I have a problem with the gui, it only seems to max out the GPU at around 30% even though the non-gui version maxes out at 100%
I'm using an Nvidia card with the latest drivers and windows 7.

[purpleeggguy]

That's odd... you're using the same flags in both cases? If so, can you try running the poclbm.exe that comes with the GUI and see if that does the same thing? I don't have an Nvidia card to test on but I haven't heard of anyone else with your problem.

Ah, I figured it out. For some reason running a cpu miner at the same time caused the speed to drop, but only when the console window running the GPU miner was in the background.
Since the gui didn't have the console window it got treated as always being in the background and wouldn't run at full capacity.

I didn't notice it before because I always had the GPU miner in the foreground.

[Kiv]

Ah, that makes sense. The GPU miner needs a percent or two of CPU in order to prepare work and send it to the GPU, and if a CPU miner is hogging all the resources then it won't be able to keep the GPU busy.

That's odd... you're using the same flags in both cases? If so, can you try running the poclbm.exe that comes with the GUI and see if that does the same thing? I don't have an Nvidia card to test on but I haven't heard of anyone else with your problem.

Ah, I figured it out. For some reason running a cpu miner at the same time caused the speed to drop, but only when the console window running the GPU miner was in the background.
Since the gui didn't have the console window it got treated as always being in the background and wouldn't run at full capacity.

I didn't notice it before because I always had the GPU miner in the foreground.

[Transisto]

Does it support multiple card ? or am I doing something wrong ?

I have a 4770 and 5870 it only show one "cypress" (the 4770)