Title: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: stts510 on April 14, 2014, 11:07:03 AM I am new here and I don't know where to post it. If it's wrong section, please move it to right section.
My rig: amd sempron 145 processor (45 watts) GA970A-DS3P motherboard (70 watts) 1 toxic r9 280x ghaphics card (250 watts) 2gb ddr3 1600mhz adata RAM (an old one. didn't bought any ram because everyone said RAM doesn't matter for mining) 160 gb sata hdd(old one) RAM and HDD shouldn't be using more than 20 watts. 1 intex psu. model no: TECHNO 450 20+4 PIN (I am not sure. but the seller says,"it's 450 watt psu". Thinking my psu may be the issue, I added another psu to give more wattage. still same problem. I am on windows 7 32 bit. I was getting same error on 64 bit too. I got cgminer 3.7.2 from here http://www.minedogecoin.com/cgminer-3.7.2-windows.zip don't see a windows executable on http://ck.kolivas.org/apps/cgminer/3.7/ and I don't know how to compile one. However I am sure that's not the issue. My batch file Code: timeout /t 2 after I launch it, I get this: https://i.imgur.com/WqjHFrj.png Code: error -4: Enqueueing kernel onto command queue. <clEnqueueNDRangeKernel> If someone could help me, I would appreciate that. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: Wipeout2097 on April 14, 2014, 09:34:27 PM Try cgminer.exe --scrypt -o stratum+tcp://url:port -u username.workername -p pass -g 2 --thread-concurrency 8192
Regarding: Quote amd sempron 145 processor (45 watts) Unfortunately the scene is full of memes and idiots... 2gb ddr3 1600mhz adata RAM (an old one. didn't bought any ram because everyone said RAM doesn't matter for mining) Those 2GB may or may not be a problem. Can depend on OS, mining algo, the videocards, thread-concurrency, motherboard and it's bios config, etc... Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: stts510 on April 15, 2014, 01:38:02 AM already ordered 8gb corsair ram
but with your advice, I got it working with a low hash. screenshot: https://i.imgur.com/T8M5KHq.png guide me in the right direction to get more hash. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: Wipeout2097 on April 15, 2014, 02:43:15 AM Oh, sorry I forgot
cgminer.exe --scrypt -o stratum+tcp://url:port -u username.workername -p pass -g 2 --thread-concurrency 8192 -I 13 --gpu-memclock 1500 --gpu-engine 1000 Then on the miner window, press 'G' and play around with parameters. You may be able to increase Memory and core a bit more. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: stts510 on April 16, 2014, 03:22:57 AM Oh, sorry I forgot cgminer.exe --scrypt -o stratum+tcp://url:port -u username.workername -p pass -g 2 --thread-concurrency 8192 -I 13 --gpu-memclock 1500 --gpu-engine 1000 Then on the miner window, press 'G' and play around with parameters. You may be able to increase Memory and core a bit more. my screen went black with those settings or pc reboots. however I used these settings and I got 140kh/s cgminer.exe --scrypt -o stratum+tcp://url:port -u username.workername -p pass -g 2 --thread-concurrency 8192 -I 10 -w 256 If I use Intensity more than 10, it's black or reboots. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: xingqiaoyin on April 16, 2014, 05:27:00 AM I believe yor PSU is crap
To check Is there any 80+ logo in your PSU ? check the spec for 12v rail in your PSU POS PSU generally lie on ratings but usually they're honest with the voltage spec rails i.e Check the voltage/ampere table on the sides The 12v should be at least 12V 30A. Or if it's multirail 12v1,12v2 the combined ampere should exceed 30A. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: Mirkic7 on April 16, 2014, 06:08:47 AM Try reducing thread concurrency, and changing work size.
I had these issues occasionally with some algo's too, until I upgraded my RAM to 4GB (from 2GB). Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: xingqiaoyin on April 16, 2014, 11:25:26 AM You may need to switch drivers.
Sometimes a driver/sdk combo that works flawlessly in one mobo..when you use the same GPU in another mobo gave me the error you pointed out. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: Bawb3 on April 16, 2014, 04:07:03 PM Check your environment variables. (In win7 just type "environment" in the start menu, and click "edit environment variables").
If GPU_MAX_ALLOC_PERCENT 100 and GPU_USE_SYNC_OBJECTS 1 are in lowercase the AMD drivers will not actually take them, the commands need to be in ALL CAPS (Yes, AMD drivers are terrible, and I cant wait to mine with Maxwell cards). Oh and I assume you have the AMD APP SDK installed too? You need that for the setx GPU commands to take as well. Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: stts510 on April 26, 2014, 01:02:07 PM I believe yor PSU is crap To check Is there any 80+ logo in your PSU ? check the spec for 12v rail in your PSU POS PSU generally lie on ratings but usually they're honest with the voltage spec rails i.e Check the voltage/ampere table on the sides The 12v should be at least 12V 30A. Or if it's multirail 12v1,12v2 the combined ampere should exceed 30A. just got 8gb ram and iam still facing same issue. I don't see any 80+ logo. And here's what is printed on the cover https://i.imgur.com/T8f7FDA.png Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: xingqiaoyin on April 27, 2014, 07:10:20 AM I believe yor PSU is crap To check Is there any 80+ logo in your PSU ? check the spec for 12v rail in your PSU POS PSU generally lie on ratings but usually they're honest with the voltage spec rails i.e Check the voltage/ampere table on the sides The 12v should be at least 12V 30A. Or if it's multirail 12v1,12v2 the combined ampere should exceed 30A. just got 8gb ram and iam still facing same issue. I don't see any 80+ logo. And here's what is printed on the cover https://i.imgur.com/T8f7FDA.png Have you tried switching drivers ? You said you added another PSU for it so i suspect power may not be the whole issue. If wored correctly two 450w psus should make it run. If you do that do you still get bsod or reboot ? That PSU pictured has 12x18 almost 250W not enough to pwer everything (mobo+cpu+gpu+etc) but should be ok to power just a single aux pcie connector. If you power one aux pcie from that psu and the other aux pcie from the other psu it should be enough power given both PSUs have the same spec and you know what your doing. This however i dont recommend to do as the voltage difference from the two PSU may cause problems. I.e one psu may give 12.2 and the other 11.9 to the other connector that plus ripple is not healthy in the long run. Your PSU also has a high 5v amperage it looks old school design... As old mobos require are 5v. Generally you want a healthy new PSU with 80+ certification to run these brands like seasonic corsair are recommended you'd want 550W rating at least, gold rated 80+ psus will give you better electricity bill running the rig 24/7 Title: Re: error -4: Enqueueing kernel onto command queue and gpu 0 failure. Post by: stts510 on April 28, 2014, 11:05:54 AM I believe yor PSU is crap To check Is there any 80+ logo in your PSU ? check the spec for 12v rail in your PSU POS PSU generally lie on ratings but usually they're honest with the voltage spec rails i.e Check the voltage/ampere table on the sides The 12v should be at least 12V 30A. Or if it's multirail 12v1,12v2 the combined ampere should exceed 30A. just got 8gb ram and iam still facing same issue. I don't see any 80+ logo. And here's what is printed on the cover https://i.imgur.com/T8f7FDA.png Have you tried switching drivers ? You said you added another PSU for it so i suspect power may not be the whole issue. If wored correctly two 450w psus should make it run. If you do that do you still get bsod or reboot ? That PSU pictured has 12x18 almost 250W not enough to pwer everything (mobo+cpu+gpu+etc) but should be ok to power just a single aux pcie connector. If you power one aux pcie from that psu and the other aux pcie from the other psu it should be enough power given both PSUs have the same spec and you know what your doing. This however i dont recommend to do as the voltage difference from the two PSU may cause problems. I.e one psu may give 12.2 and the other 11.9 to the other connector that plus ripple is not healthy in the long run. Your PSU also has a high 5v amperage it looks old school design... As old mobos require are 5v. Generally you want a healthy new PSU with 80+ certification to run these brands like seasonic corsair are recommended you'd want 550W rating at least, gold rated 80+ psus will give you better electricity bill running the rig 24/7 I am not worried about electricity bill because where I live, bills are paid on a monthly basis which is a fixed amount. Anyway, when it didn't work with two PSUs at first day, I removed one. But after reading your post today, I added it again and it's giving me 561 kh/s. Here's what I got: https://i.imgur.com/395DH4u.png and when I hit g, i got this: https://i.imgur.com/zjheIzs.png it's a good rate, but I would like to get at least 720-730 kh/s because toxic r9 280x can give 800 kh/s I did change settings a bit. My settings are: Code: timeout /t 60 |