P: Slow performance on Xeon CPUs

[Dimizu]

I noticed that LR clone and brush tool on my XEON E5-1650 0 3.20GHz (Attention: E5-1650 0 and not E5-1650 v4) can not stress my CPU and after x minutes of working LR slow down, until I have to restart it.

Please see the full diskussion with the problem here: https://forums.adobe.com/thread/2210245 (test with different Lightroom versions, confrontation with a weak laptop, that works fine, tests wit other graphic grafic card, test with other bios settings etc. No results. Only restart LR or minor display resolution helps.)

Can anyone with an XEON E5-1650 0 3.2Ghz confirm this?

[excorp11]

Simon,

I tried the config.lua patch described above.  On my testing (described below), its a definite improvement over running Lightroom without the patch, but not as good as simply setting Lightroom affinity as described by John Ellis (start /affinity F cmd.exe /c "c:\Program Files\Adobe\Adobe Lightroom\lightroom.exe").

However, I got outstanding performance (the best in my testing), by combining the config.lua patch and also setting the affinity for Lightroom to 4 cores (out of my 8).  Results are all below.

I created a test catalog with 40 raw images.  Starting in the Develop module, I simply apply a preset to the image (highlights 0, shadows 100, clarity 20, camera 'natural'), reset the image, then move to the next one.  I do this for 20 images, measuring how long it takes to process the first 5, then the 2nd 5, etc.  Given all the discussion about Lightroom slow downs, this seemed like the simplest way to get repeatable, comparable data on different versions and CPU affinities.

I'm running on an Alienware laptop, with 4 dual-core CPUs (Intel i7-4800MQ @ 2.7GHz)

As a baseline, this is what 2015.10 does without the config.lua patch, running combinations of different processors.  I tested two configurations of 4 processors (mask F (processors 0-3) and mask 55 (processors 0, 2, 4, 6) to see if that made a difference).

Times are seconds to process the first 5 images, 2nd 5, 3rd 5, and 4th 5
no mask = all 8 processors:     26      38      56      80 
Mask 3F =      6 processors:    18       24      35      49
Mask   F =      4 processors:    16       17      23      28 <-- best early performance, but degrades
Mask 55 =      4 processors:    16       16      21      27 <-- best early performance, but degrades
Mask   3 =      2 processors:     24      29      28      29
Mask   1 =      1 processors:     23      25      24      26 <-- no degradation

Now here's what you get with the config.lua patch added in:

no mask  = all 8 processors:    17       20     27      38 <-- Better than before, but not great
Mask   F =      4 processors:    14      15      18      20  
Mask 55 =      4 processors:    15       16     18      15  !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I can't say that the difference between Mask 55 and Mask F are real, or just my variation at moving the mouse.  But there's no question that masking from 8 processors to 4, AND adding your patch, produced the fastest performance I've seen.

Elsewhere in the thread, you also asked if anyone knew when this problem started.  It was the upgrade to 2015.5.  I ran 2015.4 until the last week, and never saw any degradation in performance.  I also tested it in the same way as above, and got these results:

no mask = all 8 processors:     22      24      22      21 <-- no degradation

I never upgraded past 2015.4 because, when running all 8 processors, results looked like this (for 2015.5):

no mask = all 8 processors:     26      46      80      117 

Hopefully, these trials can give you some added info to get this fixed permanently.  It would be nice to run Lightroom without having to start with a cmd prompt and a hexadecimal conversion table. 

My system info, from the Config.lua trial, without any affinity mask

Lightroom version: CC 2015.10 [ 1111918 ]
License: Creative Cloud
Operating system: Windows 7 
Version: 6.1
Application architecture: x64
System architecture: x64
Logical processor count: 8
Processor speed: 2.6 GHz
Built-in memory: 16311.1 MB
Real memory available to Lightroom: 16311.1 MB
Real memory used by Lightroom: 1084.5 MB (6.6%)
Virtual memory used by Lightroom: 1245.0 MB
GDI objects count: 667
USER objects count: 2159
Process handles count: 980
Memory cache size: 809.8MB / 3821.7MB (21.2%)
Maximum thread count used by Camera Raw: 5
Camera Raw SIMD optimization: SSE2,AVX,AVX2
Camera Raw virtual memory: 663MB / 8155MB (8%)
System DPI setting: 120 DPI
Desktop composition enabled: Yes
Displays: 1) 1920x1200, 2) 1920x1080
Input types: Multitouch: No, Integrated touch: No, Integrated pen: No, External touch: No, External pen: No, Keyboard: No

Graphics Processor Info: 
GeForce GTX 780M/PCIe/SSE2

Check OpenGL support: Passed
Vendor: NVIDIA Corporation
Version: 3.3.0 NVIDIA 372.90
Renderer: GeForce GTX 780M/PCIe/SSE2
LanguageVersion: 3.30 NVIDIA via Cg compiler


Application folder: C:\Program Files\Adobe\Adobe Lightroom
Library Path: C:\Lightroom Catalogs\Test Cat\Test Cat.lrcat
Settings Folder: C:\Users\Mike\AppData\Roaming\Adobe\Lightroom

Installed Plugins: 
1) AdobeStock
2) ColorChecker Passport
3) HDR Efex Pro 2
4) Helicon Focus Export
5) Merge to 32-bit
6) ON1 Effects Standalone 10

Config.lua flags: 
Develop.AdjustMaximumThreadCount = 0.51

Adapter #1: Vendor : 10de
Device : 119f
Subsystem : 5ae1028
Revision : a1
Video Memory : 4026
AudioDeviceIOBlockSize: 1024
AudioDeviceName: Speakers (Realtek High Definition Audio)
AudioDeviceNumberOfChannels: 2
AudioDeviceSampleRate: 48000
Build: LR5x8
Direct2DEnabled: false
GL_ACCUM_ALPHA_BITS: 16
GL_ACCUM_BLUE_BITS: 16
GL_ACCUM_GREEN_BITS: 16
GL_ACCUM_RED_BITS: 16
GL_ALPHA_BITS: 0
GL_BLUE_BITS: 8
GL_DEPTH_BITS: 24
GL_GREEN_BITS: 8
GL_MAX_3D_TEXTURE_SIZE: 2048
GL_MAX_TEXTURE_SIZE: 16384
GL_MAX_TEXTURE_UNITS: 4
GL_MAX_VIEWPORT_DIMS: 16384,16384
GL_RED_BITS: 8
GL_RENDERER: GeForce GTX 780M/PCIe/SSE2
GL_SHADING_LANGUAGE_VERSION: 4.50 NVIDIA
GL_STENCIL_BITS: 8
GL_VENDOR: NVIDIA Corporation
GL_VERSION: 4.5.0 NVIDIA 372.90
GPUDeviceEnabled: false
OGLEnabled: true
GL_EXTENSIONS: GL_AMD_multi_draw_indirect GL_AMD_seamless_cubemap_per_texture GL_ARB_arrays_of_arrays GL_ARB_base_instance GL_ARB_bindless_texture GL_ARB_blend_func_extended GL_ARB_buffer_storage GL_ARB_clear_buffer_object GL_ARB_clear_texture GL_ARB_clip_control GL_ARB_color_buffer_float GL_ARB_compatibility GL_ARB_compressed_texture_pixel_storage GL_ARB_conservative_depth GL_ARB_compute_shader GL_ARB_compute_variable_group_size GL_ARB_conditional_render_inverted GL_ARB_copy_buffer GL_ARB_copy_image GL_ARB_cull_distance GL_ARB_debug_output GL_ARB_depth_buffer_float GL_ARB_depth_clamp GL_ARB_depth_texture GL_ARB_derivative_control GL_ARB_direct_state_access GL_ARB_draw_buffers GL_ARB_draw_buffers_blend GL_ARB_draw_indirect GL_ARB_draw_elements_base_vertex GL_ARB_draw_instanced GL_ARB_enhanced_layouts GL_ARB_ES2_compatibility GL_ARB_ES3_compatibility GL_ARB_ES3_1_compatibility GL_ARB_ES3_2_compatibility GL_ARB_explicit_attrib_location GL_ARB_explicit_uniform_location GL_ARB_fragment_coord_conventions GL_ARB_fragment_layer_viewport GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_framebuffer_no_attachments GL_ARB_framebuffer_object GL_ARB_framebuffer_sRGB GL_ARB_geometry_shader4 GL_ARB_get_program_binary GL_ARB_get_texture_sub_image GL_ARB_gl_spirv GL_ARB_gpu_shader5 GL_ARB_gpu_shader_fp64 GL_ARB_gpu_shader_int64 GL_ARB_half_float_pixel GL_ARB_half_float_vertex GL_ARB_imaging GL_ARB_indirect_parameters GL_ARB_instanced_arrays GL_ARB_internalformat_query GL_ARB_internalformat_query2 GL_ARB_invalidate_subdata GL_ARB_map_buffer_alignment GL_ARB_map_buffer_range GL_ARB_multi_bind GL_ARB_multi_draw_indirect GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_occlusion_query2 GL_ARB_parallel_shader_compile GL_ARB_pipeline_statistics_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_program_interface_query GL_ARB_provoking_vertex GL_ARB_query_buffer_object GL_ARB_robust_buffer_access_behavior GL_ARB_robustness GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_seamless_cubemap_per_texture GL_ARB_separate_shader_objects GL_ARB_shader_atomic_counter_ops GL_ARB_shader_atomic_counters GL_ARB_shader_ballot GL_ARB_shader_bit_encoding GL_ARB_shader_clock GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shader_image_load_store GL_ARB_shader_image_size GL_ARB_shader_objects GL_ARB_shader_precision GL_ARB_shader_storage_buffer_object GL_ARB_shader_subroutine GL_ARB_shader_texture_image_samples GL_ARB_shader_texture_lod GL_ARB_shading_language_100 GL_ARB_shading_language_420pack GL_ARB_shading_language_include GL_ARB_shading_language_packing GL_ARB_shadow GL_ARB_sparse_buffer GL_ARB_sparse_texture GL_ARB_stencil_texturing GL_ARB_sync GL_ARB_tessellation_shader GL_ARB_texture_barrier GL_ARB_texture_border_clamp GL_ARB_texture_buffer_object GL_ARB_texture_buffer_object_rgb32 GL_ARB_texture_buffer_range GL_ARB_texture_compression GL_ARB_texture_compression_bptc GL_ARB_texture_compression_rgtc GL_ARB_texture_cube_map GL_ARB_texture_cube_map_array GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_float GL_ARB_texture_gather GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_mirrored_repeat GL_ARB_texture_multisample GL_ARB_texture_non_power_of_two GL_ARB_texture_query_levels GL_ARB_texture_query_lod GL_ARB_texture_rectangle GL_ARB_texture_rg GL_ARB_texture_rgb10_a2ui GL_ARB_texture_stencil8 GL_ARB_texture_storage GL_ARB_texture_storage_multisample GL_ARB_texture_swizzle GL_ARB_texture_view GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_transform_feedback_instanced GL_ARB_transform_feedback_overflow_query GL_ARB_transpose_matrix GL_ARB_uniform_buffer_object GL_ARB_vertex_array_bgra GL_ARB_vertex_array_object GL_ARB_vertex_attrib_64bit GL_ARB_vertex_attrib_binding GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_vertex_type_10f_11f_11f_rev GL_ARB_vertex_type_2_10_10_10_rev GL_ARB_viewport_array GL_ARB_window_pos GL_ATI_draw_buffers GL_ATI_texture_float GL_ATI_texture_mirror_once GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_bindable_uniform GL_EXT_blend_color GL_EXT_blend_equation_separate GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_depth_bounds_test GL_EXT_direct_state_access GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXTX_framebuffer_mixed_formats GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_framebuffer_object GL_EXT_framebuffer_sRGB GL_EXT_geometry_shader4 GL_EXT_gpu_program_parameters GL_EXT_gpu_shader4 GL_EXT_multi_draw_arrays GL_EXT_packed_depth_stencil GL_EXT_packed_float GL_EXT_packed_pixels GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_polygon_offset_clamp GL_EXT_provoking_vertex GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_shader_objects GL_EXT_separate_specular_color GL_EXT_shader_image_load_store GL_EXT_shader_integer_mix GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_array GL_EXT_texture_buffer_object GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_latc GL_EXT_texture_compression_rgtc GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_integer GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_mirror_clamp GL_EXT_texture_object GL_EXT_texture_shared_exponent GL_EXT_texture_sRGB GL_EXT_texture_sRGB_decode GL_EXT_texture_storage GL_EXT_texture_swizzle GL_EXT_timer_query GL_EXT_transform_feedback2 GL_EXT_vertex_array GL_EXT_vertex_array_bgra GL_EXT_vertex_attrib_64bit GL_EXT_window_rectangles GL_EXT_import_sync_object GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KHR_context_flush_control GL_KHR_debug GL_KHR_no_error GL_KHR_robust_buffer_access_behavior GL_KHR_robustness GL_KTX_buffer_region GL_NV_alpha_to_coverage_dither_control GL_NV_bindless_multi_draw_indirect GL_NV_bindless_multi_draw_indirect_count GL_NV_bindless_texture GL_NV_blend_equation_advanced GL_NV_blend_square GL_NV_command_list GL_NV_compute_program5 GL_NV_conditional_render GL_NV_copy_depth_to_color GL_NV_copy_image GL_NV_depth_buffer_float GL_NV_depth_clamp GL_NV_draw_texture GL_NV_draw_vulkan_image GL_NV_ES1_1_compatibility GL_NV_ES3_1_compatibility GL_NV_explicit_multisample GL_NV_fence GL_NV_float_buffer GL_NV_fog_distance GL_NV_fragment_program GL_NV_fragment_program_option GL_NV_fragment_program2 GL_NV_framebuffer_multisample_coverage GL_NV_geometry_shader4 GL_NV_gpu_program4 GL_NV_internalformat_sample_query GL_NV_gpu_program4_1 GL_NV_gpu_program5 GL_NV_gpu_program5_mem_extended GL_NV_gpu_program_fp64 GL_NV_gpu_shader5 GL_NV_half_float GL_NV_light_max_exponent GL_NV_multisample_coverage GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_parameter_buffer_object GL_NV_parameter_buffer_object2 GL_NV_path_rendering GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_primitive_restart GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_shader_atomic_counters GL_NV_shader_atomic_float GL_NV_shader_buffer_load GL_NV_shader_storage_buffer_object GL_NV_texgen_reflection GL_NV_texture_barrier GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_multisample GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_transform_feedback GL_NV_transform_feedback2 GL_NV_uniform_buffer_unified_memory GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_attrib_integer_64bit GL_NV_vertex_buffer_unified_memory GL_NV_vertex_program GL_NV_vertex_program1_1 GL_NV_vertex_program2 GL_NV_vertex_program2_option GL_NV_vertex_program3 GL_NVX_conditional_render GL_NVX_gpu_memory_info GL_NVX_multigpu_info GL_NVX_nvenc_interop GL_NV_shader_thread_group GL_NV_shader_thread_shuffle GL_KHR_blend_equation_advanced GL_SGIS_generate_mipmap GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_SUN_slice_accum GL_WIN_swap_hint WGL_EXT_swap_control

[FeedbackCommunityMember]

I have experience with C1 v9. Noise reduction is mixed bag compared to Lr. It is worse on large clean areas like arhitecture, but is better on detailed things. I like Color NR in Lr, but I dislike that Lr hardly fix stubborn white dots. For ultimate NR I have external app anyway. So C1 vs Lr noise reduction is irrelavant. I mostly do not apply NR on both. I just add grain. Which is more natural in C1 for my eyes.

[Denyerec]

Heh, yes this amused me as well. If I forget to start an import before I go to bed, I lose half a working day waiting for it to complete 😕

[naestyoh]

So detect mouse and keyboard inputs and shift the UX. If I'm here, dial it back. If the mouse isn't moving nor the keyboard clicking, jack up the import. This is how FCPx works. Video rendering pauses when I move my mouse over any UI elements. If I stop for 5+ seconds, the render pegs my CPU. 

[matthewg51914714]

Hey Simon, I've been directed here to try this fix by another user. I must be doing something wrong, because I'm seeing "Config.lua flags: None" in my System Info window after copying the file to the folder that opens up when I click "Show Lightroom Presets Folder" (User > AppData > Roaming > Adobe)

I'm running windows 10, most up to date Lightroom, and an i7-4800MQ CPU

[Arnie47]

You have to copy the file into the Lightroom subfolder of the User > AppData > Roaming > Adobe directory

[robv69577667]

I see some improvement editing several photo's. I'll have to do some more editing this week. I'll get back to you.

[bob frost]

I have three ssds in my i7-5930K machine and no slowdown in LR, except in rendering previews and exporting over long periods of time.

Keep experimenting, turn everything off you can think of including plugins, navigator, histogram, savetoxmp, etc, etc. Make new prefs file, make new previews, update Win10, disable any strange software, try a clean version of Windows10 without any other software.

Since we don't all have this problem, there must be some incompatibility between LR and your computer, try LR in Win8/7 compatibility mode, look for windows errors, try using MS Resource monitor or Process monitor to see what is happening when it slows down.

Bob Frost

[excorp11]

Bob,
Would you mind posting what your system is / where you got it?  I've been holding off on buying anything new until I was confident that I'd get something that actually eliminated these "Lightroom gets less and less responsive and eventually must be shut down" problems.  There are too many reports on the feedback site of people who've bought new machines, or built them, expecting to see great performance improvements only to be disappointed.

[simonsaith]

Thanks Mark for sharing the results of your experiments. Can you try tweak the config.lua in an editor and set the value of Develop.AdjustMaximumThreadCount to 0.5. So in Lr's system info, you should see

Develop.AdjustMaximumThreadCount = 0.5

and check the your experience without the affinity mask.

[simonsaith]

Copy the config.lua file into User > AppData > Roaming > Adobe > Lightroom.

Sorry for the late response. I just got back from a leave.

[excorp11]

Simon,

Very slight improvement, but nothing I'd label as significant:

at .51 (copied from above) no mask  =     17       20     27      38 
at .50                                  no mask  =     15       19     27      33

I also tried to do some editing sessions this weekend, both with and without the patch and with various affinity masks.  Nothing keeps 2015.10 running smoothly for more than 5-10 images.  Limiting the cores is a definite improvement, but it still becomes unusable in very short order. 

[FeedbackCommunityMember]

Affinity has no point whatsoever. On my 6/12-core at least not. This is complete nonsense. 
Lightroom uses all cores (all 12 logic cores) when exporting, creating 1:1 previews etc. While editing it uses for 1 core of power, maybe 2. I can set affinity to only 1 core and performance would not suffer mauch. But it would be huge difference on export. If U se GPU there is boost in performance regardless of number of cores. But for brush, I need to switch off GPU in settings. Which is annoying, and I do not do it always. Brushes doing ok under GPU, I just do not see what I'm, painting until I stop and area get overlay. For simple things I paint on "air". If I need to be precise, I turn GPU off. Annoying. SInce I use Lr only for "preset/auto" stff latelly I don't care anymoore. Adobe will lost one customer when my 1 year CC runs off. No point wasting my creative life here. 100's of replyes and no solution. It can't be...

[Jerry Syder]

Hey Simon, I would just like to pitch in here and say that I'm gobsmacked! I never even thought that Lightroom was capable of doing what I'm seeing it do now that I've taken the steps suggested. I'm actually trying to keep up with the speed(someone please pinch me now, I am dreaming right!?!). I may be counting chickens but so far I've just been going through image after image and doing my general edits for an event and it's all snappy with the config.lua doing it's thing. I was a bit adamant to give it a go because I, like some others here, didn't quite get what it would do(in fact, assumed that it would use less cores). After reading your metaphor of the traffic scenario, I thought ahhh lets try. Jerry happy 🙂

[Assaf Frank]

Hi Jerry,

can you let me know the what is the line in the config.lua
and do you notice any degradation in performance after a while
and what is your spec (CPU/GPU/OS)

Thank you in advance.
Assaf

[Jerry Syder]

Oh dear!!! I may have been a bit hasty in my response ;-(!! Ashamed of myself, I stayed quiet until someone called me out. I guess when something appears too good to be true, they are in fact, too good to be true. Two things 1. I was using smart objects so this was not a fair test as I was not working on the full RAW file. 2. Things did degrade after about 10mins or so. n.b I switched off Smart Previews to see what it would be like - result, laggy but I must admit that after testing both cases(with or without smart previews) things are marginally better than before but nothing to jump and scream about. Also, in both cases, things did degrade after minutes of a LR restart.

Config.lua flags:
Develop.AdjustMaximumThreadCount = 0.51

Lightroom version: CC 2015.10 [ 1111918 ]
License: Creative Cloud
Operating system: Windows 10
Version: 10.0
Application architecture: x64
System architecture: x64
Logical processor count: 8
Processor speed: 3.4 GHz
Built-in memory: 16325.8 MB

Graphics Processor Info:
GeForce GTX 1060 6GB

..............................

Now I go bury my head 

[Assaf Frank]

No worries mate, I have also started using smart previews and it makes a big difference, but it is not ideal when you need to check focus and zoom in, but it is defiantly more responsive.

The main issue I want adobe to concentrate is the performance degradation.

after restart everything is always much faster (without smart previews) and also creating panoramas are half the time if not less.
my Config.lua flags is the same: 
Develop.AdjustMaximumThreadCount = 0.51

which I think makes a some difference in performance.
but hey fix the performance degradation...

we had enough restarting lightroom every time we want to speed up things.

[Jerry Syder]

Haha! I know the feeling buddy. If only it continued working as when we first launched the program. I think they are working on something to nail this issue finally so fingers crossed. 

[bob frost]

Hi Simon,

I tried your config.lua a while ago and found no difference in the slowdown in rendering with time, BUT have now realised I had it in the wrong LR folder!!! Copying the .51 config.lua to the correct folder seems to have cured the slowdown in rendering D810 nefs. Early days, but I have just rendered about 1200 nefs without any slowdown!

I haven't tried the config.lua with 0.5 but will do later.

Some months ago, I had a look at what LR is doing during rendering with MS Process Monitor (filtered on lightroom.exe). I could see that when rendering became slow after an hour or more, it seemed to be spending more time creating and closing threads without anything happening (the RH side of the data lines were full of zero times). Is that the logjam; too many threads being created which can't get onto the 'motorway' and thus wasting LR time? Just had a look again at it running with 0.51 after 1200 renders and there seem to be far fewer threads being created and closed than before. I'm not a programmer, but this did seem to be a clue to me.

I'll let this rendering continue for another hour or so, and then try the other setting, but you might, just might, have solved my slowdown problem. Never had a slowdown with editing, but then I don't sit editing for hours - it's my hobby, not work 😉

Bob Frost

[FeedbackCommunityMember]

Didn't know about those feedbacks threads a few hours ago.
Just to add my experience about that config.lua on my 6 years old workstation  ( see below )
The perf boost on 2015.10 seems reel here, I can't believe my eyes, I'll be back in some days to confirm that but it looks very very promising. 

Btw, don't think what I didn't say : it looks better, but lightroom still is too slow 🙂
My hardware do not help I guess 🙂

[robv69577667]

Today went crazy on a photo with all things LR has to offer on ajusting a photo. The Config.lua allready installed. It was better, didn't get the "LR not responding" and the processor never went to 100% on al cores. But still the brush and the gragradueerd filter laked behind. Then I set the value of Develop.AdjustMaximumThreadCount to 0.5 and went crazy againg. Then LR went crazy! I've never seen it so fast!!! Processor load was between very low and 90%, when I paused it droped in an instand to below 10%. If I wanted to make panoramas with a lot of photos I had always time for a coffee, now made several panos. Just with one (10 photos) I got one moment the "not responding" but after a second or 5 LR continued. It seems to me it now also made better use of the 24Gb memory. Normaly it uses around 12 Gb, now with the panos it went up to 19Gb.  

So for me it seems okay at this moment 🙂

Lightroom-versie: 6.10 [ 1111918 ]Licentie: Onbeperkt
Besturingssysteem: Windows 7 
Versie: 6.1
Toepassingsarchitectuur: x64
Systeemarchitectuur: x64
Aantal logische processors: 12
Processorsnelheid: 2,6 GHz
Ingebouwd geheugen: 24125,5 MB
Beschikbaar fysiek geheugen voor Lightroom: 24125,5 MB
Fysiek geheugen gebruikt door Lightroom: 5973,1 MB (24,7%)
Virtueel geheugen gebruikt door Lightroom: 6503,2 MB
Aantal GDI-objecten: 901
Aantal gebruikersobjecten: 3501
Aantal proceshandelingen: 1937
Cachegeheugengrootte: 2857,2MB / 5775,3MB (49,5%)
Maximumaantal door Camera Raw gebruikte threads: 6
Camera Raw SIMD-optimalisatie: SSE2
Virtueel geheugen Camera Raw: 4745MB / 12062MB (39%)
Dpi-instelling systeem: 96 dpi
Bureaubladcompositie ingeschakeld: Ja
Schermen: 1) 2560x1440
Invoertypen: Multi-aanraking: Nee, Geïntegreerde aanraking: Nee, Geïntegreerde pen: Ja, Externe aanraking: Nee, Externe pen: Ja, Toetsenbord: Nee

GPU-info: 
GeForce GT 610/PCIe/SSE2

Check OpenGL support: Passed
Vendor: NVIDIA Corporation
Version: 3.3.0 NVIDIA 378.92
Renderer: GeForce GT 610/PCIe/SSE2
LanguageVersion: 3.30 NVIDIA via Cg compiler


Toepassingsmap: C:\Program Files\Adobe\Adobe Lightroom
Bibliotheekpad: C:\Lightroom\Algemeen\Algemeen\Algemeen.lrcat
Map Instellingen: C:\Users\T5650\AppData\Roaming\Adobe\Lightroom

Geïnstalleerde plug-ins: 
1) Canon-plug-in voor tethering
2) Facebook
3) Flickr
4) HDR Efex Pro 2
5) Helicon Focus Export
6) Leica Tether-plug-in
7) ON1 Photo RAW 2017
8) ON1 Resize 2017
9) Tetheringplug-in voor Nikon

Config.lua-vlaggen: 
Develop.AdjustMaximumThreadCount = 0.5

Adapter 1: Leverancier : 10de
Apparaat : 104a
Subsysteem : 84601043
Revisie : a1
Videogeheugen : 2000
AudioDeviceIOBlockSize: 1024
AudioDeviceName: Luidsprekers (SoundMAX Integrated Digital High Definition Audio)
AudioDeviceNumberOfChannels: 2
AudioDeviceSampleRate: 48000
Build: LR5x8
Direct2DEnabled: false
GL_ACCUM_ALPHA_BITS: 16
GL_ACCUM_BLUE_BITS: 16
GL_ACCUM_GREEN_BITS: 16
GL_ACCUM_RED_BITS: 16
GL_ALPHA_BITS: 0
GL_BLUE_BITS: 8
GL_DEPTH_BITS: 24
GL_GREEN_BITS: 8
GL_MAX_3D_TEXTURE_SIZE: 2048
GL_MAX_TEXTURE_SIZE: 16384
GL_MAX_TEXTURE_UNITS: 4
GL_MAX_VIEWPORT_DIMS: 16384,16384
GL_RED_BITS: 8
GL_RENDERER: GeForce GT 610/PCIe/SSE2
GL_SHADING_LANGUAGE_VERSION: 4.50 NVIDIA
GL_STENCIL_BITS: 8
GL_VENDOR: NVIDIA Corporation
GL_VERSION: 4.5.0 NVIDIA 378.92
GPUDeviceEnabled: false
OGLEnabled: true
GL_EXTENSIONS: GL_AMD_multi_draw_indirect GL_ARB_arrays_of_arrays GL_ARB_base_instance GL_ARB_blend_func_extended GL_ARB_buffer_storage GL_ARB_clear_buffer_object GL_ARB_clear_texture GL_ARB_clip_control GL_ARB_color_buffer_float GL_ARB_compatibility GL_ARB_compressed_texture_pixel_storage GL_ARB_conservative_depth GL_ARB_compute_shader GL_ARB_compute_variable_group_size GL_ARB_conditional_render_inverted GL_ARB_copy_buffer GL_ARB_copy_image GL_ARB_cull_distance GL_ARB_debug_output GL_ARB_depth_buffer_float GL_ARB_depth_clamp GL_ARB_depth_texture GL_ARB_derivative_control GL_ARB_direct_state_access GL_ARB_draw_buffers GL_ARB_draw_buffers_blend GL_ARB_draw_indirect GL_ARB_draw_elements_base_vertex GL_ARB_draw_instanced GL_ARB_enhanced_layouts GL_ARB_ES2_compatibility GL_ARB_ES3_compatibility GL_ARB_ES3_1_compatibility GL_ARB_ES3_2_compatibility GL_ARB_explicit_attrib_location GL_ARB_explicit_uniform_location GL_ARB_fragment_coord_conventions GL_ARB_fragment_layer_viewport GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_framebuffer_no_attachments GL_ARB_framebuffer_object GL_ARB_framebuffer_sRGB GL_ARB_geometry_shader4 GL_ARB_get_program_binary GL_ARB_get_texture_sub_image GL_ARB_gl_spirv GL_ARB_gpu_shader5 GL_ARB_gpu_shader_fp64 GL_ARB_gpu_shader_int64 GL_ARB_half_float_pixel GL_ARB_half_float_vertex GL_ARB_imaging GL_ARB_indirect_parameters GL_ARB_instanced_arrays GL_ARB_internalformat_query GL_ARB_internalformat_query2 GL_ARB_invalidate_subdata GL_ARB_map_buffer_alignment GL_ARB_map_buffer_range GL_ARB_multi_bind GL_ARB_multi_draw_indirect GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_occlusion_query2 GL_ARB_parallel_shader_compile GL_ARB_pipeline_statistics_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_program_interface_query GL_ARB_provoking_vertex GL_ARB_query_buffer_object GL_ARB_robust_buffer_access_behavior GL_ARB_robustness GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_separate_shader_objects GL_ARB_shader_atomic_counter_ops GL_ARB_shader_atomic_counters GL_ARB_shader_bit_encoding GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shader_image_load_store GL_ARB_shader_image_size GL_ARB_shader_objects GL_ARB_shader_precision GL_ARB_shader_storage_buffer_object GL_ARB_shader_subroutine GL_ARB_shader_texture_image_samples GL_ARB_shader_texture_lod GL_ARB_shading_language_100 GL_ARB_shading_language_420pack GL_ARB_shading_language_include GL_ARB_shading_language_packing GL_ARB_shadow GL_ARB_sparse_buffer GL_ARB_sparse_texture GL_ARB_stencil_texturing GL_ARB_sync GL_ARB_tessellation_shader GL_ARB_texture_barrier GL_ARB_texture_border_clamp GL_ARB_texture_buffer_object GL_ARB_texture_buffer_object_rgb32 GL_ARB_texture_buffer_range GL_ARB_texture_compression GL_ARB_texture_compression_bptc GL_ARB_texture_compression_rgtc GL_ARB_texture_cube_map GL_ARB_texture_cube_map_array GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_float GL_ARB_texture_gather GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_mirrored_repeat GL_ARB_texture_multisample GL_ARB_texture_non_power_of_two GL_ARB_texture_query_levels GL_ARB_texture_query_lod GL_ARB_texture_rectangle GL_ARB_texture_rg GL_ARB_texture_rgb10_a2ui GL_ARB_texture_stencil8 GL_ARB_texture_storage GL_ARB_texture_storage_multisample GL_ARB_texture_swizzle GL_ARB_texture_view GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_transform_feedback_instanced GL_ARB_transform_feedback_overflow_query GL_ARB_transpose_matrix GL_ARB_uniform_buffer_object GL_ARB_vertex_array_bgra GL_ARB_vertex_array_object GL_ARB_vertex_attrib_64bit GL_ARB_vertex_attrib_binding GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_vertex_type_10f_11f_11f_rev GL_ARB_vertex_type_2_10_10_10_rev GL_ARB_viewport_array GL_ARB_window_pos GL_ATI_draw_buffers GL_ATI_texture_float GL_ATI_texture_mirror_once GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_bindable_uniform GL_EXT_blend_color GL_EXT_blend_equation_separate GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_depth_bounds_test GL_EXT_direct_state_access GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXTX_framebuffer_mixed_formats GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_framebuffer_object GL_EXT_framebuffer_sRGB GL_EXT_geometry_shader4 GL_EXT_gpu_program_parameters GL_EXT_gpu_shader4 GL_EXT_multi_draw_arrays GL_EXT_packed_depth_stencil GL_EXT_packed_float GL_EXT_packed_pixels GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_polygon_offset_clamp GL_EXT_provoking_vertex GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_shader_objects GL_EXT_separate_specular_color GL_EXT_shader_image_load_store GL_EXT_shader_integer_mix GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_array GL_EXT_texture_buffer_object GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_latc GL_EXT_texture_compression_rgtc GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_integer GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_mirror_clamp GL_EXT_texture_object GL_EXT_texture_shared_exponent GL_EXT_texture_sRGB GL_EXT_texture_sRGB_decode GL_EXT_texture_storage GL_EXT_texture_swizzle GL_EXT_timer_query GL_EXT_transform_feedback2 GL_EXT_vertex_array GL_EXT_vertex_array_bgra GL_EXT_vertex_attrib_64bit GL_EXT_window_rectangles GL_EXT_import_sync_object GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KHR_context_flush_control GL_KHR_debug GL_KHR_no_error GL_KHR_robust_buffer_access_behavior GL_KHR_robustness GL_KTX_buffer_region GL_NV_alpha_to_coverage_dither_control GL_NV_bindless_multi_draw_indirect GL_NV_bindless_multi_draw_indirect_count GL_NV_blend_equation_advanced GL_NV_blend_square GL_NV_command_list GL_NV_compute_program5 GL_NV_conditional_render GL_NV_copy_depth_to_color GL_NV_copy_image GL_NV_depth_buffer_float GL_NV_depth_clamp GL_NV_draw_texture GL_NV_ES1_1_compatibility GL_NV_ES3_1_compatibility GL_NV_explicit_multisample GL_NV_fence GL_NV_float_buffer GL_NV_fog_distance GL_NV_fragment_program GL_NV_fragment_program_option GL_NV_fragment_program2 GL_NV_framebuffer_multisample_coverage GL_NV_geometry_shader4 GL_NV_gpu_program4 GL_NV_internalformat_sample_query GL_NV_gpu_program4_1 GL_NV_gpu_program5 GL_NV_gpu_program5_mem_extended GL_NV_gpu_program_fp64 GL_NV_gpu_shader5 GL_NV_half_float GL_NV_light_max_exponent GL_NV_multisample_coverage GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_parameter_buffer_object GL_NV_parameter_buffer_object2 GL_NV_path_rendering GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_primitive_restart GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_shader_atomic_counters GL_NV_shader_atomic_float GL_NV_shader_buffer_load GL_NV_shader_storage_buffer_object GL_NV_texgen_reflection GL_NV_texture_barrier GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_multisample GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_transform_feedback GL_NV_transform_feedback2 GL_NV_uniform_buffer_unified_memory GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_attrib_integer_64bit GL_NV_vertex_buffer_unified_memory GL_NV_vertex_program GL_NV_vertex_program1_1 GL_NV_vertex_program2 GL_NV_vertex_program2_option GL_NV_vertex_program3 GL_NVX_conditional_render GL_NVX_gpu_memory_info GL_NVX_multigpu_info GL_NV_shader_thread_group GL_KHR_blend_equation_advanced GL_SGIS_generate_mipmap GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_SUN_slice_accum GL_WIN_swap_hint WGL_EXT_swap_control 

[Assaf Frank]

that's great. Do you find that after working in the development module for few minutes the performance becomes slower and you have to restart to get it back up to speed?

[excorp11]

Simon -- Can you provide any updates on the status of this thread?  I notice that its not "acknowledged", or "in progress".  Has this problem of performance decay been reproduced by Adobe, or is there still a question as to whether its a Lightroom issue?

[robv69577667]

After importing 600+ photos and then the development module gets slower. But still not the "not responding". I restarted and it gets faster again, but not as fast as when I just put the Config.lua file. Next day, everything okay again, but becomes slow again but workeble. On a scale from a 100%, I was at 100% with the config.lua just in place, now I at 80/85%. So it's not completly fixed

[FeedbackCommunityMember]

Haha, that's normal for Lr... U do this, it's faster for a day, than same again. You do that, it's faster for a day, than same again. You "this" again and "that" again... switch GPU on, Switch GPU off,  make smart previews, make 1:1 previews... sneeze, dance a rain-calling dance... everything helps for that moment, but next day (or just hour), it is just same again. 

We are making biggest donkeys out of ourselves chasing our tail with debugging this, since it is unsolvable, until developers will not make proper (huge) changes from their side.