Herpy Derp
Topic: Look... all these GPU miners waste 100% CPU time for nothing... WHY?! (Read 11954 times)
June 18, 2013, 04:48:51 PM
June 18, 2013, 09:13:52 AM
Can someone answer me this?
All these miners - Phoenix, ufasoft, cgminer, Diablo... every one I've tried so far, with both AMD and nVidia GPUs, always seem to run in a blind "game loop", consuming as much CPU power as is allocated to them. They don't use any intelligent control schemes to loop with less CPU, they just go 100% full-time.
That is HUGE. Power consumption is the #1 problem with Bitcoin hashing; for many people that don't consider it, they actually waste money by mining, getting a few bucks (and a warm fuzzy feeling), but then getting slapped with a huge power bill that eats it all away. I'm running my rig outside on the porch (2nd floor) to offset the 2:3 consumption:A/C cooling ratio problem (for about every 2 watts consumed by electronics, it takes 3 watts to remove the heat produced via air-conditioning). And even with that, it consumes 200 watts at 200MHash/sec on a 6770 and an underclocked (1.2GHz/200MHz/0.95v) Core 2 Quad. The meter reports I've spent $4.50 in electricity to produce ~0.57 Bitcoin over the past week (yuck!). That's not taking into account the losses in refining my mining methods, and it's due to improve, but that's a VERY tiny profit to be made from the amount of environmental resources consumed to get there...
One of the bigger oversights of GPU mining is the CPU factor. Having a miner eat up 100% of a single core makes the PC think there's an important process running that needs additional power to accelerate the process. That's completely untrue in Bitcoin mining - it just checks/updates the GPU's progress more often. Having a higher clock speed does NOTHING to increase the speed of the GPU process! Maybe a 1-2% change at worst, but going from full clock to minimal clock reduces the CPU's power consumption by more than half!
I brought this up once in the Phoenix thread, but got pretty much ignored with just one reply suggesting it was my nVidia drivers at fault. Now that I'm using ATI and seeing exactly the same behavior (and also with an nVidia Quadro driver, which is completely different), I know it's not the driver. It's the miner's loop. I've just got to wonder.... has anyone taken steps to address this in their setups? And are the developers of these miners able to do anything about the "game loop" problem?
===========================================================================================================All these miners - Phoenix, ufasoft, cgminer, Diablo... every one I've tried so far, with both AMD and nVidia GPUs, always seem to run in a blind "game loop", consuming as much CPU power as is allocated to them. They don't use any intelligent control schemes to loop with less CPU, they just go 100% full-time.
That is HUGE. Power consumption is the #1 problem with Bitcoin hashing; for many people that don't consider it, they actually waste money by mining, getting a few bucks (and a warm fuzzy feeling), but then getting slapped with a huge power bill that eats it all away. I'm running my rig outside on the porch (2nd floor) to offset the 2:3 consumption:A/C cooling ratio problem (for about every 2 watts consumed by electronics, it takes 3 watts to remove the heat produced via air-conditioning). And even with that, it consumes 200 watts at 200MHash/sec on a 6770 and an underclocked (1.2GHz/200MHz/0.95v) Core 2 Quad. The meter reports I've spent $4.50 in electricity to produce ~0.57 Bitcoin over the past week (yuck!). That's not taking into account the losses in refining my mining methods, and it's due to improve, but that's a VERY tiny profit to be made from the amount of environmental resources consumed to get there...
One of the bigger oversights of GPU mining is the CPU factor. Having a miner eat up 100% of a single core makes the PC think there's an important process running that needs additional power to accelerate the process. That's completely untrue in Bitcoin mining - it just checks/updates the GPU's progress more often. Having a higher clock speed does NOTHING to increase the speed of the GPU process! Maybe a 1-2% change at worst, but going from full clock to minimal clock reduces the CPU's power consumption by more than half!
I brought this up once in the Phoenix thread, but got pretty much ignored with just one reply suggesting it was my nVidia drivers at fault. Now that I'm using ATI and seeing exactly the same behavior (and also with an nVidia Quadro driver, which is completely different), I know it's not the driver. It's the miner's loop. I've just got to wonder.... has anyone taken steps to address this in their setups? And are the developers of these miners able to do anything about the "game loop" problem?
I have found out the ATI drivers written after version 11.6 are buggy. It appears to "automatically" have use 100% of the core. Even though it appears that some of the "newer" drivers perform "better". It is not necessarily the case.
I have read that it can be remedied by using this variable in the args of your batch file.
'sext GPU_USE_SYNC_OBJECTS"
Hit
September 15, 2011, 11:24:31 AM
I never really experienced the 100% CPU issue with 11.5/2.1. However I just did the full upgrade bundle with 11.8 a few days ago and sure enough my CPU meter is pegged at 100%. So then I completely uninstalled the drivers and re-installed 11.5/2.1... aaaaand my CPU is still at 100% during mining. FML
September 15, 2011, 10:55:39 AM
which sdk? 2.1, 2.4, or 2.5?
I got a very slight increase (0.1%) in performance going from 2.4 to 2.5
September 15, 2011, 03:23:07 AM
which sdk? 2.1, 2.4, or 2.5?
2.1 never gives 100% cpu, but unfortunately that doesn't support AMD 6xxx GPUs. 
September 14, 2011, 04:57:00 PM
Sempron 140 with 3 x 6950 modded to 6970. Cata 11.6. Hashing speed drops when I connect with teamviewer on the computer and ~10% cpu is teamviewer alone.
which sdk? 2.1, 2.4, or 2.5?
September 13, 2011, 03:12:36 PM
The first time I experienced the 100% CPU bug was with an NVidia card (GeForce GT 330M). When I started mining I tried that and other NVidia cards (including a 9800 GX2 and 8800 GT).
Things were mostly fine when I switched to ATI cards. Two 5830s and a 5750. Catalyst 11.6 drivers, 0% CPU.
When I finally upgraded to Catalyst 11.8, then every single system hit 100% CPU. Setting the priority to "Below Normal" prevents it from slowing things down, but my CPUs still generates heat when at 100% load, 24/7.
I use GUIMiner, and it doesn't recognize my ATI cards when I use older drivers and SDK.
Things were mostly fine when I switched to ATI cards. Two 5830s and a 5750. Catalyst 11.6 drivers, 0% CPU.
When I finally upgraded to Catalyst 11.8, then every single system hit 100% CPU. Setting the priority to "Below Normal" prevents it from slowing things down, but my CPUs still generates heat when at 100% load, 24/7.
I use GUIMiner, and it doesn't recognize my ATI cards when I use older drivers and SDK.
September 13, 2011, 02:32:05 PM
Perhaps it's because my Windows machine has an Intel CPU? Anyone with Intel able to avoid 100% CPU usage with more than one AMD 69xx GPU?
September 13, 2011, 11:14:52 AM
Hell, maybe I'll throw a 0.02bcn donation to the first guy that can actually reply in a civilized manner, like the experienced tech I am, not some kind of computer-newbie...
did you try cgminer and did it have the same issue?
if yes: driver fault, if no: you're right and it could be phonenix.
sorry cant't help further
did i qualify for the 0.02btc
September 13, 2011, 10:58:17 AM
https://i.imgur.com/hXPNWl.jpg
Sempron 140 with 3 x 6950 modded to 6970. Cata 11.6. Hashing speed drops when I connect with teamviewer on the computer and ~10% cpu is teamviewer alone.
Sempron 140 with 3 x 6950 modded to 6970. Cata 11.6. Hashing speed drops when I connect with teamviewer on the computer and ~10% cpu is teamviewer alone.
September 12, 2011, 07:41:26 AM
New leaked version of Catalyst drivers 11.9
Build Info:
DriverVer=08/10/2011, 8.890.0.0000
8.89-110810a2-124125C
Is 100% CPU Usage free during OpenCL computations!
You can get it from Guru3D forums
http://forums.guru3d.com/showthread.php?t=350638
This mostly worked, cpu usage reduced to 2-3 percent with phoenix 1.6.2. , WORKSIZE=128 AGGRESSION=8. Oddly, cpu usage goes up with any aggression level that's higher or lower than 8 for me. Build Info:
DriverVer=08/10/2011, 8.890.0.0000
8.89-110810a2-124125C
Is 100% CPU Usage free during OpenCL computations!
You can get it from Guru3D forums
http://forums.guru3d.com/showthread.php?t=350638
September 12, 2011, 07:23:54 AM
Just tested, cgminer happily mining at 1% CPU on a win7 box on dual 6970s, cat 11.6 (driver only) + sdk 2.4
Can you specify what you mean by "driver only"? I have W7 64 bit, 2x6950, cat 11.6, sdk 2.4 and cgminer, and it still uses 100% cpu. 
September 11, 2011, 02:18:39 PM
Trying out 11.9 and still experiencing 100% CPU.
Win7 X64 Dual 6990 rig.
Win7 X64 Dual 6990 rig.
September 11, 2011, 04:38:43 AM
New leaked version of Catalyst drivers 11.9
Build Info:
DriverVer=08/10/2011, 8.890.0.0000
8.89-110810a2-124125C
Is 100% CPU Usage free during OpenCL computations!
You can get it from Guru3D forums
http://forums.guru3d.com/showthread.php?t=350638
Build Info:
DriverVer=08/10/2011, 8.890.0.0000
8.89-110810a2-124125C
Is 100% CPU Usage free during OpenCL computations!
You can get it from Guru3D forums
http://forums.guru3d.com/showthread.php?t=350638
September 09, 2011, 04:12:02 AM
All five of my Win7 boxes have always had Phoenix1.5 and CGminer eat up 100% of the CPUs with various catalyst versions and always SDK2.4+
To the OP - For changing ram clocks, try MSI Afterburner Beta, TriXX, and BarelyClocked. I use combinations of those utilities throughout my Win7 boxes on my LAN with TightVNC and have every card at 300 MHz RAM (except the one my 3 year old uses to play Starcraft2, obviously)
sc2 runs fine for me, even while miningTo the OP - For changing ram clocks, try MSI Afterburner Beta, TriXX, and BarelyClocked. I use combinations of those utilities throughout my Win7 boxes on my LAN with TightVNC and have every card at 300 MHz RAM (except the one my 3 year old uses to play Starcraft2, obviously)
even while underclocking ram
September 09, 2011, 12:10:42 AM
All five of my Win7 boxes have always had Phoenix1.5 and CGminer eat up 100% of the CPUs with various catalyst versions and always SDK2.4+
To the OP - For changing ram clocks, try MSI Afterburner Beta, TriXX, and BarelyClocked. I use combinations of those utilities throughout my Win7 boxes on my LAN with TightVNC and have every card at 300 MHz RAM (except the one my 3 year old uses to play Starcraft2, obviously)
To the OP - For changing ram clocks, try MSI Afterburner Beta, TriXX, and BarelyClocked. I use combinations of those utilities throughout my Win7 boxes on my LAN with TightVNC and have every card at 300 MHz RAM (except the one my 3 year old uses to play Starcraft2, obviously)
September 08, 2011, 02:34:05 PM
Just tested, cgminer happily mining at 1% CPU on a win7 box on dual 6970s, cat 11.6 (driver only) + sdk 2.4
September 08, 2011, 02:19:12 PM
phatk 2.2 was saying that it doesn't support my graphics cards when I had 11.6 and 2.1 SDK installed.
Regardless, I tried diapalo's phatk and the cpu bug was there anyways. Too much of a headache so i went back to 11.8 2.5
Regardless, I tried diapalo's phatk and the cpu bug was there anyways. Too much of a headache so i went back to 11.8 2.5
September 07, 2011, 07:59:08 PM
tl;dr use 11.6 with 2.1 SDK
http://developer.amd.com/sdks/AMDAPPSDK/downloads/pages/AMDAPPSDKDownloadArchive.aspx
http://developer.amd.com/sdks/AMDAPPSDK/downloads/pages/AMDAPPSDKDownloadArchive.aspx
I tried 11.6 with 2.1 sdk and phoenix 1.6.2 with diapolo's phatk kernel (phatk 2.2 doesnt support sdk 2.1 for some reason), still had 100% cpu with 1 graphics card. wouldve tried guiminer but it needs 2.4 sdk or higher.
sapphire 5830, win7 x64 ultimate, above stuff.
September 07, 2011, 07:00:48 PM
I tried 11.6 with 2.1 sdk and phoenix 1.6.2 with diapolo's phatk kernel (phatk 2.2 doesnt support sdk 2.1 for some reason), still had 100% cpu with 1 graphics card. wouldve tried guiminer but it needs 2.4 sdk or higher.
sapphire 5830, win7 x64 ultimate, above stuff.
"durr, hurr, this is noobie stuff, hurr, you're doing it wrong, hurr hurr read the note on your forehead, hurr."sapphire 5830, win7 x64 ultimate, above stuff.
- quoth the beginning of this topic.
(to: everyone else ITT, not at quoted:)
Guess it's not such an easy problem to solve after all, huh. Maybe if people would look past the low post count and actually read what people (e.g. I) actually wrote in their posts, more stuff like this would come to light and get resolved. Are we done durr-hurring that everyone except the little falcon-brain has their stuff working 100% properly?
BTW, I found that the CPU-usage problem occurs any time the client tries to make the GPU work harder than it wants to work, with the "golden" 11.6 drivers all set up properly. Feed too high an "aggression" setting at it, and it'll slow the desktop responsiveness down, but send CPU usage through the roof. Pretty easy to screw up a setting like that, considering there are no warnings against doing so... doesn't sound so "durr hurr" to me.
Jump to: