Comparing results from classical Radiance with Accelerad's

[Alex Katsikogiannis]

Hey everyone,

I've been using Radiance through a Python wrapper called bifacial_radiance, which is tailored to modelling of solar photovoltaic applications.

The simulated scene, as you might have guessed, consists of an elevated PV system in an open-field. A combination of gendaylit with rtrace are being used to determine irradiance on 38 sampling points (irradiance sensors). Given the following Radiance parameters "-ab 2 -aa 0.05 -ar 128 -ad 2048 -as 128", classical Radiance requires around 1 minute per rtrace and around 45 minutes for all 38 sampling points per daylight hour. 

This is a time-consuming process and I hoped that Accelerad could be of use here, regardless of the relatively low -ad setting. However, computational time remained the same and in fact the simulated irradiance was underestimated by 10% in comparison to classical Radiance.

Currently, my script iterates through each sampling point and calls rtrace. Is it possible for Accelerad to simultaneously process multiple points in parallel? Is there something else that I can do to reduce computational time?

Any help would be appreciated,

Cheers,

Alex



Note: whenever rtrace is called the following prints out. How can I mute it?

message: rtrace: OptiX 6.8.6 found display driver 532.10, CUDA driver 12.1.0, and 1 GPU device: rtrace: Device 0: NVIDIA GeForce RTX 4080 Laptop GPU with 58 multiprocessors, 1024 threads per block, 1665000 Hz, 12878086144 bytes global memory, 1048576 hardware textures, compute capability 8.9, timeout enabled, Tesla compute cluster driver disabled, PCI 0000:01:00.0. rtrace: Geometry build time: 470 milliseconds for 313919 objects. rtrace: OptiX kernel 2 time: 130 milliseconds (0 seconds). rtrace: Adaptive sampling: 9 milliseconds. rtrace: Retrieved 1 of 1 potential seeds at level 0. rtrace: Using all 1 seeds at level 0 (4096 needed for k-means). rtrace: OptiX kernel 3 time: 11 milliseconds (0 seconds). rtrace: Retrieved 744 of 4194304 potential seeds at level 1. rtrace: Using all 744 seeds at level 1 (4096 needed for k-means). rtrace: OptiX kernel 4 time: 16 milliseconds (0 seconds). rtrace: OptiX kernel 1 time: 21 milliseconds (0 seconds). rtrace: Retrieved 744 ambient records from 4096 queries at level 1. rtrace: OptiX kernel 4 time: 0 milliseconds (0 seconds). rtrace: OptiX kernel 1 time: 10 milliseconds (0 seconds). rtrace: Retrieved 1 ambient records from 4096 queries at level 0. rtrace: OptiX kernel 0 time: 9 milliseconds (0 seconds). rtrace: ray tracing time: 1470 milliseconds (1 seconds).

[Nathaniel Jones]

Hi Alex,

Whether you use Radiance or Accelerad, calling rtrace separately for each sampling point is a very inefficient strategy. You should provide all the sensor points in a single list and make single call to rtrace. The image below from this paper shows the amount of time it takes Radiance (in dark blue) and Accelerad (in lighter blues) to trace varying numbers of rays. It's an old paper, and there have been a number of optimizations to Accelerad since then. With only 38 sampling points, I do not know if there is any benefit for you to use Accelerad, but you can certainly achieve faster timing than you have now.

You can mute informational messages using the -w argument.

Nathaniel

[Odysseas Alexandros Katsikogiannis]

Hey Nathaniel,

I was naive to think that the cached ambient values are shared amongst sensors regardless of how sensor points are provided to rtrace. Indeed, computational time reduced significantly! Thanks for the treat :P

Just to verify my understanding, if I were to sample 500 sensor points with an -ad value of 2000, would that result in 500x2000 = 1,000,000 primary rays? Based on the plot, through the use of Accelerad computational times would reduce by an order of magnitude (depending on GPU architecture). 

The paper that you forwarded mentions that NVIDIAs OptiX utilizes a bounding volume hierarchy and does not necessitate the creation of an octree. You also discuss the possibility of by-passing the creation of an octree to reduce computational time. Have you or anyone else attempted to resolve this?

Cheers,

Alex

[Nathaniel Jones]

Hi Alex,

If you had 500 sample points with and -ad value of 2000, then you have 500 primary rays aimed toward your sensors. At the first bounce, Accelerad will select all 500 sensor locations (because it is less than the default -ac value of 4096) and launch 1936 rays from each (the largest perfect square less than 2000), assuming your minimum ray weight -lw is less than 1/1936. So for the first bounce, Accelerad would trace 500*1936 = 968,000 rays. The plot shows only primary rays, but again, it's an early paper that doesn't account for some optimizations.

To my knowledge, no one has attempted a version of Radiance that does not require octree formatted input, although you could optimize the oconv parameters for faster octree generation, since that will not penalize the creation of the bounding volume hierarchy.

Nathaniel

[Odysseas Alexandros Katsikogiannis]

Hey Nathaniel,

I've conducted a sensitivity analysis on computational time versus primary rays, similar to the study you detailed in one of your papers. My aim is to determine what speed enhancements are possible based on the scene and my graphics card. The scene features PV modules elevated above ground accompanied by simple geometries that mimic trees. 

I've varied the number of sampling points from 16 to 21,420 and called rtrace with the following parameters:

rtrace -i -ab 2 -aa 0.05 -ar 128 -ad 2048 -as 128 -st 0.053 -g -h -oovs octfile.oct

Low spatial resolution: 16 sampling points (32,400 primary rays)

Classical Radiance required 159, while Accelerad required 61 seconds to complete.

Overall, the GPU-based irradiation (for which I used gencumsky instead of gendaylit) is in agreement with classical Radiance; however, for certain points receiving predominantly diffuse light, there is a great difference between the two raytracing algorithms up to -40%.  I aim to tweak the rtrace simulation parameters to see if these can align more closely. 

High spatial resolution: 21,420 sampling points (43,375,500 primary rays)

Classical Radiance required 4354 seconds, while Accelerad required 68 seconds to complete! This substantial performance gain is misleading as sensors that primarily receive diffuse light registered a simulated irradiation of 0 with Accelerad. I've re-run the simulation with a higher -ad, -ar, and -ac parameters, yet those sensors still show 0 irradiation. 

Here is a sample of these results:
classical Radiance

x,y,z,backZ,mattype,backMat,Wm2Front,Wm2Back
14.26678,-6.39698,0.51,0.49,a0.21.0.0.NP_leaves.6457,a0.21.0.0.NP_leaves.2310,343271.0,68263.03
14.30148,-6.303191,0.51,0.49,a0.20.0.0.NP_leaves.6457,a0.20.0.0.NP_leaves.2310,339293.5,68297.16
14.33617,-6.209402,0.51,0.49,a0.19.0.0.NP_leaves.6457,a0.19.0.0.NP_leaves.2310,335335.9,67990.57
14.37087,-6.115613,0.51,0.49,a0.18.0.0.NP_leaves.6457,a0.18.0.0.NP_leaves.2310,331419.1,67891.91

Accelerad

x,y,z,backZ,mattype,backMat,Wm2Front,Wm2Back
14.26678,-6.396979,0.51,0.49,*,*,334075.2,0.0
14.30148,-6.303191,0.51,0.49,*,*,337704.3,67607.84
14.33617,-6.209402,0.51,0.49,*,*,331600.5,0.0
14.37087,-6.115613,0.51,0.49,*,*,94140.93,0.0

x, y, z, backZ specify the sampling point coordinates which are identical to both raytracing algorithms. However, the last column (Wm2Back) which corresponds to the irradiation received at the back side of the surface is 0 for some sensors. 

Any tips to overcome this issue and minimize the difference between the two algorithms would be greatly appreciated!

Cheers,

Alex

Below, I've copied the output given a -gv 3.

message: rtrace: OptiX 6.8.6 found display driver 532.10, CUDA driver 12.1.0, and 1 GPU device: rtrace: Device 0: NVIDIA GeForce RTX 4080 Laptop GPU with 58 multiprocessors, 1024 threads per block, 1665000 Hz, 12878086144 bytes global memory, 1048576 hardware textures, compute capability 8.9, timeout enabled, Tesla compute cluster driver disabled, PCI 0000:01:00.0. rtrace: 1 SYS INFO: OptiX Version:[7.7.0] Branch:[r530_00] Build Number:[32931673] ABI Version:[84] CUDA Version:[12.1,] 64-bit Display driver: 532.10 Devices available: CUDA device: 0 0000:01:00.0 NVIDIA GeForce RTX 4080 Laptop GPU SM count: 58 SM arch: 89 SM clock: 1665 KHz GPU memory: 12281 MB TCC driver: 0 Compatible devices: 0 RT core version: 30 rtrace: 2 MEM USAGE: CUDA context memory (CUDA device: 0): 197.0 MBytes rtrace: 2 INFO: Program cache HIT : ray_generator rtrace: 2 INFO: Program cache HIT : exception rtrace: 2 INFO: Program cache HIT : miss rtrace: 2 INFO: Program cache HIT : miss_shadow rtrace: 2 INFO: Program cache HIT : closest_hit_normal_radiance rtrace: 2 INFO: Program cache HIT : closest_hit_normal_shadow rtrace: 2 INFO: Program cache HIT : closest_hit_normal_point_cloud rtrace: 2 INFO: Program cache HIT : gencumsky rtrace: 2 INFO: Program cache HIT : any_hit rtrace: 2 INFO: Program cache HIT : closest_hit_radiance rtrace: 2 INFO: Program cache HIT : closest_hit_shadow rtrace: 2 INFO: Program cache HIT : closest_hit_point_cloud rtrace: 2 INFO: Program cache HIT : any_hit_ambient rtrace: 2 INFO: Program cache HIT : closest_hit_ambient rtrace: 2 INFO: Program cache HIT : defaultTriangleAttributeProgram rtrace: 2 INFO: Program cache HIT : mesh_attribute rtrace: 2 INFO: Program cache HIT : aabb_indexed_1_0 rtrace: 2 INFO: Program cache HIT : intersect_indexed_1_0 rtrace: 2 INFO: Program cache HIT : rtc_bvh_dummy_traverser_program rtrace: Geometry build time: 636 milliseconds for 394019 objects. rtrace: 2 INFO: Program cache HIT : cloud_generator rtrace: 2 INFO: Program cache HIT : exception rtrace: 2 INFO: Program cache HIT : ambient_sample_camera rtrace: 2 INFO: Program cache HIT : exception rtrace: 2 INFO: Program cache HIT : hemisphere_camera rtrace: 2 INFO: Program cache HIT : exception rtrace: 2 INFO: Program cache HIT : ambient_cloud_camera rtrace: 2 INFO: Program cache HIT : exception rtrace: 2 INFO: Program cache HIT : ambient_record_bounds rtrace: 2 INFO: Program cache HIT : ambient_record_intersect rtrace: 2 INFO: Program cache HIT : ambient_record_any_hit rtrace: 2 INFO: Program cache HIT : ambient_miss rtrace: 2 INFO: Launch index 0. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 33 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 787992 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 1 TIMING: Time to first launch: 1058.5 ms rtrace: 2 INFO: Program cache HIT : compute_aabb rtrace: 2 INFO: Program cache HIT : compute_aabb_exception rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1349.1 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 61.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-_Z22compute_aabb_exceptionv_ptx0x4731e27b8245aa26-key52b90447ae2e7530a5f0ee4aeac7844b-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z12compute_aabbv_ptx0x4731e27b8245aa26-keyb4e50382975414d8ee6e50061ffe3c60-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z16aabb_indexed_1_0iiPf_ptx0x1e25132a3223f0df-key0de5e7212a2f51fa655bb2463f331e0f-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-bounds_rtcbvh_nomotion_ptx0x3053b7c092f1ab23-key0de5e7212a2f51fa655bb2463f331e0f-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z21ambient_record_boundsiPf_ptx0x84028db6e97007c0-keyebb0e90c45a362e4cc93691c23a534c5-sm_89-drv532.10 rtrace: 2 INFO: Module: exception___Z22compute_aabb_exceptionv_ptx0x4731e27b8245aa26: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: raygen___Z12compute_aabbv_ptx0x4731e27b8245aa26: stack size (bytes): direct=152, continuation=0 rtrace: 2 INFO: Module: directcallable___Z16aabb_indexed_1_0iiPf_ptx0x1e25132a3223f0df: stack size (bytes): direct=56, continuation=0 rtrace: 2 INFO: Module: directcallable__bounds_rtcbvh_nomotion_ptx0x3053b7c092f1ab23: stack size (bytes): direct=224, continuation=0 rtrace: 2 INFO: Module: directcallable___Z21ambient_record_boundsiPf_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=0, continuation=0 rtrace: 1 TIMING: Compilation time: 25.4 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1320.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 61.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1320.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 61.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 TIMING: Acceleration update time: 83.9 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1320.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 61.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-key844b87d98cdc8757ef78f072f1856fb3-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d-key4ea64466a068718bb2774a26f38a6a7c-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4-key0718b872d9845a3dc2d0a4b4511d0eaf-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9gencumsky6float3S__ptx0xb76bfedfdd98548b-key4f29f9802dd8f6aebdd8c20a8f982d66-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df-key3b6aecc9964d21c067f39d66a84ec47e-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d-key7bbe11905408257702295956fdadc303-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9exceptionv_ptx0x9e80777abb91084d-key761b7c762c526b8109d82e5dd22dadb4-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-key0a7b6351d047d0717a903841ed7770af-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z4missv_ptx0x69017fd5e7fc7d5d-key2006ae79e32e75f326062caca3002689-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z11miss_shadowv_ptx0x69017fd5e7fc7d5d-key8c2dad7995c94ddaad92db2e8268ca22-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z12ambient_missv_ptx0x84028db6e97007c0-key0a7b6351d047d0717a903841ed7770af-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-key3ae414c8e30c8b1384b8db559424d721-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z7any_hitv_ptx0xffbfed6b39164451-key67e7361c7e910cd02b7f6b26e92e394a-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z22ambient_record_any_hitv_ptx0x84028db6e97007c0-key01e40132a622e537225bb9b10c9e927f-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z15cloud_generatorv_ptx0x9e80777abb91084d-key35a7bed9d7813526c49893871f208f16-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z24ambient_record_intersecti_ptx0x84028db6e97007c0-key7154a45da6378226cf7ac9a55fd78a1e-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z20closest_hit_radiancev_ptx0xffbfed6b39164451-key64af22d4de275d6ffcc423ed8cad30d5-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z18closest_hit_shadowv_ptx0xffbfed6b39164451-key966b8668b6d905c88aaa529530c6b025-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z23closest_hit_point_cloudv_ptx0xffbfed6b39164451-keyd3eb4c0e49f4b9b58c63fa60d6da3557-sm_89-drv532.10 rtrace: 2 INFO: Module: continuationcallable___Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d: stack size (bytes): direct=8, continuation=576 rtrace: 2 INFO: Module: continuationcallable___Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d: stack size (bytes): direct=0, continuation=96 rtrace: 2 INFO: Module: directcallable___Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: directcallable___Z9gencumsky6float3S__ptx0xb76bfedfdd98548b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: closesthit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=32 rtrace: 2 INFO: Module: closesthit___Z20closest_hit_radiancev_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=320 rtrace: 2 INFO: Module: closesthit___Z18closest_hit_shadowv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: closesthit___Z23closest_hit_point_cloudv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=288 rtrace: 2 INFO: Module: raygen___Z15cloud_generatorv_ptx0x9e80777abb91084d: stack size (bytes): direct=0, continuation=352 rtrace: 2 INFO: Module: intersection___Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df: stack size (bytes): direct=88, continuation=0 rtrace: 2 INFO: Module: intersection___Z24ambient_record_intersecti_ptx0x84028db6e97007c0: stack size (bytes): direct=8, continuation=0 rtrace: 2 INFO: Module: exception___Z9exceptionv_ptx0x9e80777abb91084d: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss___Z4missv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z11miss_shadowv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z12ambient_missv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z7any_hitv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z22ambient_record_any_hitv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: WARNING: Large stack size 18272 computed from maximum trace depth and maximum callable program depth rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=8, continuation=18272 rtrace: 1 TIMING: Compilation time: 27.3 ms rtrace: 1 TIMING: Total launch time: 171.7 ms rtrace: OptiX kernel 2 time: 171 milliseconds (1 seconds). rtrace: Adaptive sampling: 8 milliseconds. rtrace: Retrieved 21420 of 21420 potential seeds at level 0. rtrace: K-means performed 3 loop iterations in 15 milliseconds. rtrace: K-means produced 8192 of 8192 clusters at level 0. rtrace: 2 INFO: Launch index 1. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 787992 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 TIMING: Acceleration update time: 0.0 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1699.5 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 61.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-keyb2c29a6d2900f1ea78158759c8b533e7-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z20closest_hit_radiancev_ptx0xffbfed6b39164451-key6529a987ba3f1c525c61b84c33856b46-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z18closest_hit_shadowv_ptx0xffbfed6b39164451-key05c20e5cbf2fbc0ebe5f018814f430ec-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z23closest_hit_point_cloudv_ptx0xffbfed6b39164451-key86e7c856c94650a11586130eaeb1cce6-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z17hemisphere_camerav_ptx0xaf84a208746ee964-key8cff83d3cdb6822d4a6357950758931c-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d-key5d8ee9f4e6711d592f45564d034d2953-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d-keybc1e8f00b418a03a1cfbd2b50bd9a48f-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4-keyadf80bd53f31efef83f2190cab8d5cfa-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-keyc53d203ac74ce0172c0b7c89bc09d0d7-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z11miss_shadowv_ptx0x69017fd5e7fc7d5d-key5f869824417c01e1920e566500cb928f-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z12ambient_missv_ptx0x84028db6e97007c0-keyc53d203ac74ce0172c0b7c89bc09d0d7-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-null_program_ptx0xce987d5f52ee715b-key985dbeedfd4a15e03d17337105333ceb-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z7any_hitv_ptx0xffbfed6b39164451-key3421d0d0d515a16e2338307ca1f08e95-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z22ambient_record_any_hitv_ptx0x84028db6e97007c0-key9d81a097b56d7532a4bd1ecb73d4247d-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9gencumsky6float3S__ptx0xb76bfedfdd98548b-key145977eef7a42f818fff8cead899ee70-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df-keyf6738e824a4f06800ecaa0cfcdfdcb71-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9exceptionv_ptx0xaf84a208746ee964-keyd27079ee2be879969654caa1f7e5d595-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z24ambient_record_intersecti_ptx0x84028db6e97007c0-keyb3543753d5400011a01caf1ec306532c-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z4missv_ptx0x69017fd5e7fc7d5d-keya83bacb2cfa360cedd05e89bc33af134-sm_89-drv532.10 rtrace: 2 INFO: Module: continuationcallable___Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d: stack size (bytes): direct=8, continuation=576 rtrace: 2 INFO: Module: continuationcallable___Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d: stack size (bytes): direct=0, continuation=96 rtrace: 2 INFO: Module: directcallable___Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: directcallable___Z9gencumsky6float3S__ptx0xb76bfedfdd98548b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: closesthit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=32 rtrace: 2 INFO: Module: closesthit___Z20closest_hit_radiancev_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=320 rtrace: 2 INFO: Module: closesthit___Z18closest_hit_shadowv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: closesthit___Z23closest_hit_point_cloudv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=288 rtrace: 2 INFO: Module: raygen___Z17hemisphere_camerav_ptx0xaf84a208746ee964: stack size (bytes): direct=0, continuation=128 rtrace: 2 INFO: Module: intersection___Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df: stack size (bytes): direct=88, continuation=0 rtrace: 2 INFO: Module: intersection___Z24ambient_record_intersecti_ptx0x84028db6e97007c0: stack size (bytes): direct=8, continuation=0 rtrace: 2 INFO: Module: exception___Z9exceptionv_ptx0xaf84a208746ee964: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss___Z4missv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z11miss_shadowv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z12ambient_missv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z7any_hitv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z22ambient_record_any_hitv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: WARNING: Large stack size 18048 computed from maximum trace depth and maximum callable program depth rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=8, continuation=18048 rtrace: 1 TIMING: Compilation time: 14.7 ms rtrace: 1 TIMING: Total launch time: 37.5 ms rtrace: 1 TIMING: Time from RTcontext creation to first frame: 1290.1 ms rtrace: OptiX kernel 3 time: 37 milliseconds (0 seconds). rtrace: Retrieved 14245802 of 16588800 potential seeds at level 1. rtrace: K-means performed 6 loop iterations in 5293 milliseconds. rtrace: K-means produced 8151 of 8192 clusters at level 1. rtrace: 2 INFO: Launch index 2. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 787992 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 TIMING: Acceleration update time: 0.0 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1699.5 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 441.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-_Z21ambient_sample_camerav_ptx0x7fb01235e266f2a7-keye2ee49628a2ba52d8b1d43cdf8e31ef6-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9exceptionv_ptx0x7fb01235e266f2a7-key97ebc51571ac1c4d06b1c88cc78cfde6-sm_89-drv532.10 rtrace: 2 INFO: Module: continuationcallable___Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d: stack size (bytes): direct=8, continuation=576 rtrace: 2 INFO: Module: continuationcallable___Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d: stack size (bytes): direct=0, continuation=96 rtrace: 2 INFO: Module: directcallable___Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: directcallable___Z9gencumsky6float3S__ptx0xb76bfedfdd98548b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: closesthit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=32 rtrace: 2 INFO: Module: closesthit___Z20closest_hit_radiancev_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=320 rtrace: 2 INFO: Module: closesthit___Z18closest_hit_shadowv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: raygen___Z21ambient_sample_camerav_ptx0x7fb01235e266f2a7: stack size (bytes): direct=0, continuation=192 rtrace: 2 INFO: Module: intersection___Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df: stack size (bytes): direct=88, continuation=0 rtrace: 2 INFO: Module: intersection___Z24ambient_record_intersecti_ptx0x84028db6e97007c0: stack size (bytes): direct=8, continuation=0 rtrace: 2 INFO: Module: exception___Z9exceptionv_ptx0x7fb01235e266f2a7: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss___Z4missv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z11miss_shadowv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z12ambient_missv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z7any_hitv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z22ambient_record_any_hitv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: WARNING: Large stack size 18112 computed from maximum trace depth and maximum callable program depth rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=8, continuation=18112 rtrace: 1 TIMING: Compilation time: 5.7 ms rtrace: 1 TIMING: Total launch time: 644.6 ms rtrace: OptiX kernel 4 time: 644 milliseconds (1 seconds). rtrace: 2 INFO: Launch index 3. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 787992 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 TIMING: Acceleration update time: 0.0 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1699.5 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 441.6 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-_Z9exceptionv_ptx0x088bb3c87bee337e-key87f272938b75d74661b797fdb902507a-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z15any_hit_ambientv_ptx0xb82ce2ea3cab3d74-keya88d33598516ef67efa828d803c9406e-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z20ambient_cloud_camerav_ptx0x088bb3c87bee337e-key48d0f34153a7acf8e6959c0e1cd71733-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z19closest_hit_ambientv_ptx0xb82ce2ea3cab3d74-key9d21db268cd05f56858726a31ad92407-sm_89-drv532.10 rtrace: 2 INFO: Module: continuationcallable___Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d: stack size (bytes): direct=8, continuation=576 rtrace: 2 INFO: Module: continuationcallable___Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d: stack size (bytes): direct=0, continuation=96 rtrace: 2 INFO: Module: directcallable___Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: directcallable___Z9gencumsky6float3S__ptx0xb76bfedfdd98548b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: closesthit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=32 rtrace: 2 INFO: Module: closesthit___Z20closest_hit_radiancev_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=320 rtrace: 2 INFO: Module: closesthit___Z18closest_hit_shadowv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: closesthit___Z19closest_hit_ambientv_ptx0xb82ce2ea3cab3d74: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: raygen___Z20ambient_cloud_camerav_ptx0x088bb3c87bee337e: stack size (bytes): direct=0, continuation=128 rtrace: 2 INFO: Module: intersection___Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df: stack size (bytes): direct=88, continuation=0 rtrace: 2 INFO: Module: intersection___Z24ambient_record_intersecti_ptx0x84028db6e97007c0: stack size (bytes): direct=8, continuation=0 rtrace: 2 INFO: Module: exception___Z9exceptionv_ptx0x088bb3c87bee337e: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss___Z4missv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z11miss_shadowv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z12ambient_missv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z7any_hitv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z15any_hit_ambientv_ptx0xb82ce2ea3cab3d74: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z22ambient_record_any_hitv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: WARNING: Large stack size 18048 computed from maximum trace depth and maximum callable program depth rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=8, continuation=18048 rtrace: 1 TIMING: Compilation time: 9.1 ms rtrace: 1 TIMING: Total launch time: 93.6 ms rtrace: OptiX kernel 1 time: 94 milliseconds (0 seconds). rtrace: warning - Exception code of user exception out of bounds occurred in ambient calculation rtrace: Retrieved 8150 ambient records from 8192 queries at level 1. rtrace: 2 INFO: Launch index 4. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 796142 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 TIMING: Acceleration update time: 3.0 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 TIMING: Total launch time: 1142.2 ms rtrace: OptiX kernel 4 time: 1143 milliseconds (1 seconds). rtrace: 2 INFO: Launch index 5. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 796142 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 TIMING: Acceleration update time: 0.0 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 TIMING: Total launch time: 310.2 ms rtrace: OptiX kernel 1 time: 310 milliseconds (0 seconds). rtrace: Retrieved 8192 ambient records from 8192 queries at level 0. rtrace: 2 INFO: Launch index 6. rtrace: 2 SCENE STAT: Node graph object summary: rtrace: 2 SCENE STAT: RTprogram : 35 rtrace: 2 SCENE STAT: RTbuffer : 23 rtrace: 2 SCENE STAT: RTtexturesampler : 1 rtrace: 2 SCENE STAT: RTacceleration : 2 rtrace: 2 SCENE STAT: RTgroup : 0 rtrace: 2 SCENE STAT: RTgeometrygroup : 2 rtrace: 2 SCENE STAT: RTtransform : 0 rtrace: 2 SCENE STAT: RTselector : 0 rtrace: 2 SCENE STAT: RTgeometryinstance: 2 rtrace: 2 SCENE STAT: RTgeometry : 2 rtrace: 2 SCENE STAT: Total prim: 796184 rtrace: 2 SCENE STAT: RTmaterial : 2 rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.3 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.2 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.2 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 2 TIMING: Acceleration update time: 1.4 ms rtrace: 2 MEM USAGE: Buffer GPU memory usage: rtrace: 2 MEM USAGE: | Category | Count | Total MByte | rtrace: 2 MEM USAGE: | buffer | 31 | 1700.2 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | texture | 1 | 0.0 | rtrace: 2 MEM USAGE: | > p2p | 0 | 0.0 | rtrace: 2 MEM USAGE: | demand load | 0 | 0.0 | rtrace: 2 MEM USAGE: | gfx interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | cuda interop | 0 | 0.0 | rtrace: 2 MEM USAGE: | optix internal | 18 | 0.1 | rtrace: 2 MEM USAGE: Buffer host memory usage: 442.8 Mbytes rtrace: 2 MEM USAGE: WARNING: Raw pointer buffer access detected. rtrace: 1 INFO: Compilation triggered rtrace: 2 INFO: Module cache HIT : rtx-_Z13ray_generatorv_ptx0x7a018fad90163088-keye30e64830c9f00dad609afd6a5037b03-sm_89-drv532.10 rtrace: 2 INFO: Module cache HIT : rtx-_Z9exceptionv_ptx0x7a018fad90163088-key7ff12c77da9875c0d9898a45e08404bd-sm_89-drv532.10 rtrace: 2 INFO: Module: continuationcallable___Z27closest_hit_normal_radianceRK13IntersectData19PerRayData_radiance_ptx0x0f367227c69ad37d: stack size (bytes): direct=8, continuation=576 rtrace: 2 INFO: Module: continuationcallable___Z25closest_hit_normal_shadowRK13IntersectData17PerRayData_shadow_ptx0x0f367227c69ad37d: stack size (bytes): direct=0, continuation=96 rtrace: 2 INFO: Module: directcallable___Z30closest_hit_normal_point_cloudRK13IntersectData22PerRayData_point_cloud_ptx0x14c1b703410ae2c4: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: directcallable___Z9gencumsky6float3S__ptx0xb76bfedfdd98548b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: closesthit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=32 rtrace: 2 INFO: Module: closesthit___Z20closest_hit_radiancev_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=320 rtrace: 2 INFO: Module: closesthit___Z18closest_hit_shadowv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=256 rtrace: 2 INFO: Module: raygen___Z13ray_generatorv_ptx0x7a018fad90163088: stack size (bytes): direct=0, continuation=160 rtrace: 2 INFO: Module: intersection___Z21intersect_indexed_1_0i_ptx0x1e25132a3223f0df: stack size (bytes): direct=88, continuation=0 rtrace: 2 INFO: Module: intersection___Z24ambient_record_intersecti_ptx0x84028db6e97007c0: stack size (bytes): direct=8, continuation=0 rtrace: 2 INFO: Module: exception___Z9exceptionv_ptx0x7a018fad90163088: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: miss___Z4missv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z11miss_shadowv_ptx0x69017fd5e7fc7d5d: stack size (bytes): direct=40, continuation=0 rtrace: 2 INFO: Module: miss___Z12ambient_missv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit__null_program_ptx0xce987d5f52ee715b: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z7any_hitv_ptx0xffbfed6b39164451: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: Module: anyhit___Z22ambient_record_any_hitv_ptx0x84028db6e97007c0: stack size (bytes): direct=0, continuation=0 rtrace: 2 INFO: WARNING: Large stack size 18080 computed from maximum trace depth and maximum callable program depth rtrace: 2 INFO: Traversal pipeline stack (bytes): traversalCallableDirect=0, stateCallableDirect=8, continuation=18080 rtrace: 1 TIMING: Compilation time: 6.3 ms rtrace: 1 TIMING: Total launch time: 457.5 ms rtrace: OptiX kernel 0 time: 457 milliseconds (1 seconds). rtrace: ray tracing time: 10658 milliseconds (11 seconds).

[Nathaniel Jones]

Hi Alex,

For clarity, if you have 16 sample points using -i to calculate irradiance, then you have 16 primary rays, not 32,400. That is, unless you are calculating the irradiance at each point 2048 times, in which case I missed something in your explanation.

What -ac value are you using? If you have 21,420 sampling points, make sure you also have -ac set to 21,420.

Assuming the issue is related to the irradiance cache, have you tried turning irradiance caching off by setting -aa to zero?

If that doesn't solve the issue, then we muse consider issues with the model. What tool did you use to export the geometry to Radiance format? Could there be an issues with the winding direction of holes that would cause the surfaces to triangulate incorrectly?

Nathaniel

[Odysseas Alexandros Katsikogiannis]

Oh, I thought that primary rays could be estimated by multiplying the number of sampling points by the ambient divisions (-ad) setting - fitting a perfect square. I am not running the simulation multiple times. Based on your suggestion, it seems that primary rays are independent of the -ad setting. 

I used the default -ac value, and indeed, setting it equal to the number of sampling points resolved the issue. Thank you Nathaniel! 

I am now exploring ways to decrease the mismatch between Radiance and Accelerad. Currently, for sampling points that receive primarily diffuse light—simulated with Accelerad—the variance ranges between 20% and -30% compared to those simulated with Radiance. Of course, there's no guarantee that the latter are closer to the true values, but this variance should still be minimized. In an attempt to reduce that I:

* doubled -ad and tripled -ar settings

* doubled -an, while halving -at and -ax settings

however, it seems that this variance remains. I'll need to do some reading to get the hang of these. Regardless, any suggestions or tips would be highly appreciated.

Cheers,

Alex

[Nathaniel Jones]

Hi Alex,

Before playing more with the ambient lighting settings, I still suggest that you try running the simulations with -aa 0 to confirm that the irradiance cache is indeed causing the discrepancies.

If it is the case, then I would suggest making -ar smaller rather than larger, and potentially tuning -ac even larger. Keep in mind that Accelerad's irradiance cache is built in parallel, quite different from the way classic Radiance builds its irradiance cache, so it does not necessarily make sense to use the same settings in Accelerad that you would use in Radiance.

Nathaniel

[Odysseas Alexandros Katsikogiannis]

A delayed response—I've been updating my script to simplify the comparison between the classical Radiance and Accelerad methods.

By adjusting the setting of - ac to 1.5 * total_number_of_sampling_points (which amounts to 19,404 in this instance), the discrepancy between the two raytracing methods has diminished, but it is still significant. Of course, Accelerad achieves that x70 faster given my CPU/GPU combination. 

I specifically looked at the minimum, maximum, and mean irradiation values across all sampling points, along with the standard deviation. The findings indicate that Accelerad tends to overestimate irradiation values for the given scene:

Min irradiation: - 46%
Max irradiation: - 14%
Mean irradiation: + 5%
Standard deviation: - 17%

Min and max irradiation are subject to artifacts, while mean irradiation is a more reliable metric to compare the performance.

classical Radiance

Accelerad

As the scene consists of complex geometries (think of photovoltaic modules with small cell spacings ~ 10 cm), I believe that simulation settings must be refined to capture light transmission through these thin gaps given that the maximum scene dimension ~ 30 m.  

Accelerad and classical Radiance utilize the same set of Radiance parameters, other than - ac. Setting - aa to 0 with Accelerad, did not reduce this discrepancy, so I am wondering if you have any further tips for the remaining Radiance parameters, or Accelerad specific parameters - an, - at, -a x.

[Nathaniel Jones]

Hi Alex,

I'm sure what the images represent, but I can see that the values in the 2nd image are slightly higher.

In a geometrically complex scene, increasing -ac will be helpful. You are still better off running your comparisons with -aa 0 so that you can avoid differences caused by Accelerad's parallel irradiance caching algorithm. If you need to use irradiance caching, then run classic Radiance with an overture pass. This will improve the accuracy of classic Radiance and make its results more similar to Accelerad's.

If irradiance caching is not the issues, then the next things to consider are light sources and materials. You could run a test in which all materials are perfectly diffuse plastics, and then a test in which all materials are 100% specular plastics, to see if either the materials or the sky as a light source are causing issues.

Nathaniel