Hi,
I use a mini PC for travelling by motorhome - also for editing images
with darktable.
Today I managed to activate openCL - but after that dt is so slow I
cannot use it anymore.
Changing exposure has a delay of seconds until the GUI reacts - other
modules I could not test really.
I then started my benchmark scipt which uses dt-cli on 6 sample-images
to export them and report the time.
Here is what I got:
Image Nr.
times in seconds
without openCL
with openCL
2016_N8R0008.NEF
49,0 160,0
2017_N7X0049.nef
18,0 45,0
2017_N7X0218.nef
18,0 43,0
2017_N8T2476.nef
55,0 83,0
2017_OL00020.ORF
22,0 50,0
2017_OL00162.orf
20,0 45,0
This is a sample report:
2016_N8R0008.NEF
Belichtung (ein)
Drehung (ein)
Entrauschen (nicht-lokales Mittel) (ein)
Basiskurve (ein)
Schärfen (ein)
[dev] took 1,270 secs (0,704 CPU) to load the image.
[export] creating pixelpipe took 0,139 secs (0,204 CPU)
[dev_pixelpipe] took 0,043 secs (0,072 CPU) initing base buffer [export]
[dev_pixelpipe] took 0,740 secs (0,060 CPU) processed
`Raw-Schwarz-/Weißpunkt' on GPU, blended on GPU [export]
[dev_pixelpipe] took 0,669 secs (0,004 CPU) processed `Weißabgleich' on
GPU, blended on GPU [export]
[dev_pixelpipe] took 0,122 secs (0,000 CPU) processed
`Spitzlicht-Rekonstruktion' on GPU, blended on GPU [export]
[dev_pixelpipe] took 1,337 secs (0,000 CPU) processed `Entrastern' on
GPU, blended on GPU [export]
[dev_pixelpipe] took 0,501 secs (0,000 CPU) processed `Belichtung' on
GPU, blended on GPU [export]
[dev_pixelpipe] took 0,366 secs (0,000 CPU) processed `Drehung' on GPU,
blended on GPU [export]
[dev_pixelpipe] took 0,782 secs (0,000 CPU) processed `Basiskurve' on
GPU, blended on GPU [export]
[dev_pixelpipe] took 1,219 secs (0,000 CPU) processed
`Eingabefarbprofil' on GPU, blended on GPU [export]
[dev_pixelpipe] took 136,833 secs (0,380 CPU) processed `Entrauschen
(nicht-lokales Mittel)' on GPU, blended on GPU [export]
[dev_pixelpipe] took 7,458 secs (0,004 CPU) processed `Schärfen' on GPU,
blended on GPU [export]
[dev_pixelpipe] took 1,346 secs (0,004 CPU) processed
`Ausgabefarbprofil' on GPU, blended on GPU [export]
[dev_pixelpipe] took 5,995 secs (0,008 CPU) processed `abschließendes
Skalieren' on GPU, blended on GPU [export]
[dev_process_export] pixel pipeline processing took 157,507 secs (0,532 CPU)
[export_job] exported to `2016_N8R0008.jpg'
time used: 160 sec.
compared to - openCL off:
2016_N8R0008.NEF
Belichtung (ein)
Drehung (ein)
Entrauschen (nicht-lokales Mittel) (ein)
Basiskurve (ein)
Schärfen (ein)
[dev] took 1,285 secs (0,692 CPU) to load the image.
[export] creating pixelpipe took 0,199 secs (0,392 CPU)
[dev_pixelpipe] took 0,043 secs (0,076 CPU) initing base buffer [export]
[dev_pixelpipe] took 0,052 secs (0,172 CPU) processed
`Raw-Schwarz-/Weißpunkt' on CPU, blended on CPU [export]
[dev_pixelpipe] took 0,035 secs (0,096 CPU) processed `Weißabgleich' on
CPU, blended on CPU [export]
[dev_pixelpipe] took 0,020 secs (0,080 CPU) processed
`Spitzlicht-Rekonstruktion' on CPU, blended on CPU [export]
[dev_pixelpipe] took 0,993 secs (3,044 CPU) processed `Entrastern' on
CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,194 secs (0,500 CPU) processed `Belichtung' on
CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,339 secs (1,356 CPU) processed `Drehung' on CPU
with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,135 secs (0,540 CPU) processed `Basiskurve' on
CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,209 secs (0,816 CPU) processed
`Eingabefarbprofil' on CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 41,705 secs (165,384 CPU) processed `Entrauschen
(nicht-lokales Mittel)' on CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,766 secs (2,944 CPU) processed `Schärfen' on CPU
with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,371 secs (1,472 CPU) processed
`Ausgabefarbprofil' on CPU with tiling, blended on CPU [export]
[dev_pixelpipe] took 0,708 secs (2,828 CPU) processed `abschließendes
Skalieren' on CPU with tiling, blended on CPU [export]
[dev_process_export] pixel pipeline processing took 45,572 secs (179,316
CPU)
[export_job] exported to `2016_N8R0008.jpg'
time used: 49 sec.
so you see that denoising e. g. takes 3 times as long when openCL is
ON, but also all the other modules take longer, downscaling at the end
is slowed down by a factor of more than 8, sharpening by almost 10.
darktable-cltest reports:
[opencl_init] opencl related configuration options:
[opencl_init]
[opencl_init] opencl: 1
[opencl_init] opencl_library: ''
[opencl_init] opencl_memory_requirement: 768
[opencl_init] opencl_memory_headroom: 300
[opencl_init] opencl_device_priority: '*/!0,*/*/*'
[opencl_init] opencl_size_roundup: 16
[opencl_init] opencl_async_pixelpipe: 0
[opencl_init] opencl_synch_cache: 0
[opencl_init] opencl_number_event_handles: 25
[opencl_init] opencl_micro_nap: 1000
[opencl_init] opencl_use_pinned_memory: 0
[opencl_init] opencl_use_cpu_devices: 0
[opencl_init] opencl_avoid_atomics: 0
[opencl_init] opencl_enable_markesteijn: 1
[opencl_init]
[opencl_init] found opencl runtime library 'libOpenCL'
[opencl_init] opencl library 'libOpenCL' found on your system and loaded
[opencl_init] found 1 platform
[opencl_init] found 2 devices
[opencl_init] device 0 `Intel(R) HD Graphics' supports image sizes of
16384 x 16384
[opencl_init] device 0 `Intel(R) HD Graphics' allows GPU memory
allocations of up to 1634MB
[opencl_init] device 0: Intel(R) HD Graphics
GLOBAL_MEM_SIZE: 3269MB
MAX_WORK_GROUP_SIZE: 256
MAX_WORK_ITEM_DIMENSIONS: 3
MAX_WORK_ITEM_SIZES: [ 256 256 256 ]
DRIVER_VERSION: r5.0.63503
DEVICE_VERSION: OpenCL 1.2
[opencl_init] compiling program `demosaic_ppg.cl' ..
[opencl_load_program] loaded cached binary program from file
`/home/benutzer/.cache/darktable/cached_kernels_for_IntelRHDGraphics/demosaic_ppg.cl.bin'
[opencl_load_program] successfully loaded program from
`/usr/share/darktable/kernels/demosaic_ppg.cl'
[opencl_build_program] successfully built program
[opencl_build_program] BUILD STATUS: 0
BUILD LOG:
and here is the config of the machine:
System: Host: carPC Kernel: 4.10.0-32-generic x86_64 (64 bit gcc: 5.4.0)
Desktop: Cinnamon 3.4.6 (Gtk 3.18.9-1ubuntu3.3)
Distro: Linux Mint 18.2 Sonya
Machine: Mobo: ASRock model: J4205-ITX
Bios: American Megatrends v: P1.30 date: 04/18/2017
CPU: Quad core Intel Pentium J4205 (-MCP-) cache: 1024 KB
flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips:
11980
clock speeds: max: 2600 MHz 1: 1137 MHz 2: 1003 MHz 3: 1189 MHz
4: 1200 MHz
Graphics: Card: Intel Device 5a84 bus-ID: 00:02.0
Display Server: X.Org 1.18.4 driver: intel
Resolution: [email protected]
GLX Renderer: Mesa DRI Intel HD Graphics 505 (Broxton)
GLX Version: 3.0 Mesa 17.0.7 Direct Rendering: Yes
Audio: Card Intel Device 5a98 driver: snd_hda_intel bus-ID: 00:0e.0
Sound: Advanced Linux Sound Architecture v: k4.10.0-32-generic
Network: Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit
Ethernet Controller
driver: r8169 v: 2.3LK-NAPI port: e000 bus-ID: 01:00.0
IF: enp1s0 state: up speed: 1000 Mbps duplex: full mac: <filter>
Card-2: Intel Wireless 8260 driver: iwlwifi bus-ID: 02:00.0
IF: wlp2s0 state: down mac: <filter>
Drives: HDD Total Size: 2000.4GB (43.4% used)
ID-1: /dev/sda model: WDC_WD20NPVZ size: 2000.4GB
Partition: ID-1: / size: 69G used: 13G (19%) fs: ext4 dev: /dev/sda2
ID-2: /home size: 1.8T used: 782G (48%) fs: ext4 dev: /dev/sda4
ID-3: swap-1 size: 17.18GB used: 0.00GB (0%) fs: swap dev:
/dev/sda3
RAID: No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors: System Temperatures: cpu: 37.0C mobo: N/A
Fan Speeds (in rpm): cpu: N/A
Info: Processes: 208 Uptime: 17 min Memory: 719.9/7639.8MB
Init: systemd runlevel: 5 Gcc sys: 5.4.0
Client: Shell (bash 4.3.481) inxi: 2.2.35
I used the Intel drivers:
https://software.intel.com/en-us/articles/opencl-drivers#latest_linux_driver
Why is this?
--
regards
Bernhard
http://www.bilddateien.de
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
