This is done by storing only a subset of PathRadiance, and by storing
direct light immediately in the main PathRadiance. Saves about 10% of
CUDA stack memory, and simplifies subsurface indirect ray code.
Also pass by value and don't write back now that it is just a hash for seeding
and no longer an LCG state. Together this makes CUDA a tiny bit faster in my
tests, but mainly simplifies code.
Unlike regular path tracing, branched path tracing is usually used with lower
sample counts, at least for primary rays. This means that are less samples for
the GPU to work on in parallel and rendering is slower. As there is less work
overall there is also more inactive threads during rendering with BPT. This
patch makes use of those inactive rays to render branched samples in parallel
with other samples.
Each thread that is preparing for a branched sample will attempt to find an
inactive thread and if one is found the state for the sample is copied to that
thread. Potentially, if there are enough inactive threads, 100s of branched
samples could be generated from the same originating thread and ran in
parallel giving large speed ups.
Gives 70% faster render for pavillion midday scene. 20-60% faster on BMW
with car paint replaced with SSS/volumes.
This implements branched path tracing for the split kernel.
General approach is to store the ray state at a branch point, trace the
branched ray as normal, then restore the state as necessary before iterating
to the next part of the path. A state machine is used to advance the indirect
loop state, which avoids the need to add any new kernels. Each iteration the
state machine recreates as much state as possible from the stored ray to keep
overall storage down.
Its kind of hard to keep all the different integration loops in sync, so this
needs lots of testing to make sure everything is working correctly. We should
probably start trying to deduplicate the integration loops more now.
Nonbranched BMW is ~2% slower, while classroom is ~2% faster, other scenes
could use more testing still.
Reviewers: sergey, nirved
Reviewed By: nirved
Subscribers: Blendify, bliblubli
Differential Revision: https://developer.blender.org/D2611
Simplifies code quite a bit, making it shorter and easier to extend.
Currently no functional changes for users, but is required for the
upcoming work of shadow catcher support with OpenCL.
This does a few things at once:
- Refactors host side split kernel logic into a new device
agnostic class `DeviceSplitKernel`.
- Removes tile splitting, a new work pool implementation takes its place and
allows as many threads as will fit in memory regardless of tile size, which
can give performance gains.
- Refactors split state buffers into one buffer, as well as reduces the
number of arguments passed to kernels. Means there's less code to deal
with overall.
- Moves kernel logic out of OpenCL kernel files so they can later be used by
other device types.
- Replaced OpenCL specific APIs with new generic versions
- Tiles can now be seen updating during rendering
This commit re-shuffles code in split kernel once again and makes it so common
parts which is in the headers is only responsible to making all the work needed
for specified ray index. Getting ray index, checking for it's validity and
enqueuing tasks are now happening in the device specified part of the kernel.
This actually makes sense because enqueuing is indeed device-specified and i.e.
with CUDA we'll want to enqueue kernels from kernel and avoid CPU roundtrip.
TODO:
- Kernel comments are still placed in the common header files, but since queue
related stuff is not passed to those functions those comments might need to
be split as well.
Just currently read them considering that they're also covering the way how
all devices are invoking the common code path.
- Arguments might need to be wrapped into KernelGlobals, so we don't ened to
pass all them around as function arguments.
This was broken after the kernel file restructure.
Variables allocated in the __local address space can only be defined
inside a __kernel function.
We probably need to solve this a bit differently once we do the CUDA
kernel split, but this fix shoud be good enough until then.
Since the kernel split work we're now having quite a few of new files, majority
of which are related on the kernel entry points. Keeping those files in the
root kernel folder will eventually make it really hard to follow which files are
actual implementation of Cycles kernel.
Those files are now moved to kernel/kernels/<device_type>. This way adding extra
entry points will be less noisy. It is also nice to have all device-specific
files grouped together.
Another change is in the way how split kernel invokes logic. Previously all the
logic was implemented directly in the .cl files, which makes it a bit tricky to
re-use the logic across other devices. Since we'll likely be looking into doing
same split work for CUDA devices eventually it makes sense to move logic from
.cl files to header files. Those files are stored in kernel/split. This does not
mean the header files will not give error messages when tried to be included
from other devices and their arguments will likely be changed, but having such
separation is a good start anyway.
There should be no functional changes.
Reviewers: juicyfruit, dingto
Differential Revision: https://developer.blender.org/D1314
