Overclock.net › Forums › Overclockers Care › Overclock.net BOINC Team › Hardware Optimization › DistrRTgen project support thread
New Posts  All Forums:Forum Nav:

DistrRTgen project support thread - Page 18

post #171 of 173
i wish they had like a percentage bar showing completion percentage of the current hash we are working on..
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
post #172 of 173
Next batch of work flowing for GPUs
First, GPU work is now flowing for NV CUDA and ATI/AMD OpenCL alike.

While, I can make excuses for delays, lack of communication, and so on: everything related to this falls short of the project's goals. First, when work is running low and we do not have the next set queued, we should communicate this. Second, we need to be more flexible for unexpected issues, whether for instance it is lack of developer/admin time or a new set explodes due to errors, we need to not just be ready for our next set but the set after that.

Communication is something we need to have at the forefront. On the admin side we have graphs, monitoring, and predictive stats for when work will run out *but* this seems to fail and the reason is simple: we are overly optimistic about the times to test and push a code fix *or* are struggling to use existing deployed code to create useful sets that are not so large and slow as to be unusable.

In this particular case there is a bug for the AMD/ATI OpenCL code for ntlm which resulted in the matching mysqlsha1 set that just wrapped up. I believed I could get this code fixed, tested, and deployed in a timely fashion, but should have just released the md5 set that is now in progress. I *thought* we had already done the md5 set as well as being optimistic on fixing the OpenCL bug on ntlm. This batch *should* have and *could* have been released to prevent any interruption in the work flow.

I made a placeholder thread, which I will expound on this, including making excuses in the primary forums https://www.freerainbowtables.com/phpBB3/viewtopic.php?f=4&t=4658
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
post #173 of 173
ATI OpenCL issues
It appears similar to the disaster of ntlm, that the md5 OpenCL may have the same issues. I *thought* we had run this algorithm before so it may not be the algorithm but issues with the hybrid2 code. In any case I have halted the production of new work for the GPU side, though there is still work already in the buffer.

I think in the immediate case we have the following to do ASAP:
1) find a way to continue Nvidia CUDA even if we have to disable OpenCL AMD/ATI (or vice/versa) so it isn't a 100% GPU shutdown

2) fix the OpenCL code
I am about to go pick up a new GPU that will let me directly attach vms in my server to the card, such as windows x64. On the GPU side PowerBlade has had the burden of all things windows related, and it is past time that I have a dev environment for both sides. I do have notes in the code for the ntlm side that has a very similar design on the md5 side that might be the issue in both, but I will know more as soon as I can test that directly.

*If* we could come up with a table construction for mysqlsha1 to keep work flowing in the meantime that would be great for crunchers, but we have exhausted the algorithm to a large extent with the code currently deployed. In the mean time work is likely to:
1) dry up
2) erroneously validate in some cases 2 OpenCL results agree (we still do limited checks on the server even with matching WUs)
3) result in a lot of inconclusive results

In any case further updates will be made in as timely a manner as possible. I do not see the time-frame to fix the OpenCL code bugs and get that code out to clients to be more than days
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel I7-5820k@4.2ghz Gigabyte X99-UD3P AMD 295x2 16GB G.Skill DDR4-2400 
Hard DriveOptical DriveCoolingOS
Corsair Force GT 120gb (OS), 2x 1tb WD Blacks R... LG Bluray CoolerMaster MasterLiquid Pro 120 Win 10 x64 
MonitorKeyboardPowerCase
26" Emprex Logitech G15 CoolerMaster SPH-1300 Coolermaster HAF922 
MouseMouse Pad
CM Sentinel II CoolerMaster HSM Battle Pad SSK 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware Optimization
Overclock.net › Forums › Overclockers Care › Overclock.net BOINC Team › Hardware Optimization › DistrRTgen project support thread