"OpenCL benchMark"

NNN635158767546269381

• February 10, 2015 14:24

In the Home/Library/Logs directory, one can find the CaptureOne logs.

I just noticed this:
2015-02-10 15:16:19> OpenCL initialization...
2015-02-10 15:16:19> OpenCL : found platform Apple, OpenCL Version : OpenCL 1.2 (Jul 29 2014 21:24:39)
2015-02-10 15:16:19> OpenCL Device : HD Graphics 4000
2015-02-10 15:16:19> OpenCL Driver Version : 1.2(Dec 17 2014 19:37:58)
2015-02-10 15:16:19> OpenCL Compute Units : 16
2015-02-10 15:16:19> OpenCL : Loading kernels
2015-02-10 15:16:19> OpenCL : Loading kernels finished
2015-02-10 15:16:19> OpenCL : Benchmarking
2015-02-10 15:16:21> OpenCL : Initialization completed
2015-02-10 15:16:21> OpenCL benchMark : 2.664000


Pretty interesting is the OpenCL benchmark. my question: Higher number means faster or slower performances ? Could anyone post its values on another machine ?
Regards

0

• 

  Permanently deleted user

  • December 16, 2015 10:16

  Thanks, that helps. The file is included in package of log files.

  0
• 

  Jeffrey Jakucyk

  • January 07, 2016 02:23

  For what it's worth, I've never seen any computer max out all cores to 100% when exporting RAW files to JPEG or anything else. Whether it's Aperture, Capture One, Lightroom, or Bridge, whether it's OS X or Windows, Apple or Dell, Xeon, i7, or i5, hyperthreading or no, the best I've seen is an average of about 60% utilization of all cores (yes it bounces up and down a fair bit, but it's never 100% saturation). In all cases they do use all the cores pretty equally and at the same time, so you still get huge benefits from having 8 cores versus 4 for instance. I'm not sure where the bottleneck is, if it's RAM speed, floating point vs. integer, or what, but it just seems to be inherent.
  
  In my case, with an older 2008 Mac Pro, I have to leave OpenCL off in C1. It's pretty painful all around with it turned on. It's not too surprising really, even with an "upgraded" Radeon HD 5770 that it just can't compete with 8 physical processor cores. Otherwise performance is actually pretty good. It's not quite Aperture good, and I'll admit launching C1 is painfully slow even on an SSD, but otherwise it's quite capable.

  0
• 

  BeO

  • January 07, 2016 10:56

  

  Lionel from Phase One (a development lead) indicated in the capture one blog that there is some benchmarking information on its way, I hope it will be soon. However, I can't wait much longer and have to decide for a new machine.
  
  I currently run a Dell Precision Mobile Workstation M4500 with i5-520M, 8GB DDR3 1333 Mhz and a graphics card below memory specs so it is not used (FX 880M, imgcore.log benchmark figure around 3), slow SSD (Sata II). And guess, my 2 cores (4threads) are utilized 100% when processing 😊
  8 GB is not enough, it mostly swaps out memory to disk and slowing down everything.
  
  Loading a preview, working with sliders and especially with LA is very slow.
  Cataloging i.e. opening the browser with "all imagaes" (4000) is a pain (5 minutues). The raws are on a NAS and the slowliness is apparent whether ot not the NAS is on or off (catalog on internal SSD).
  
  I am about to order a workstation 1x E5-1650v3 6 cores (12 threads), 32GB DDR4 2133 Mhz, SSD Sata III. I hope this will satify me for the next 5 years. I don't rely on a gfx card, as quite a few users (like you) have to switch off opencl but I hope I will find a gfx card that supersedes and works together with the processor without any glitches, it should have reasonable power consumption (watt) and price though.
  
  As I can't upgrade the processor or RAM speed in this workstation later, I hope gfx cards and gfx card support from C1 is a potential upgrade path in a couple of years from now.
  
  I am struggeling with this decision for a few weeks now but in the absence of concrete understanding when a gfx card will be faster than the CPU it will be like shooting the dark.
  
  cheers,
  BeO

  0

Post is closed for comments.