I want to use cudaMallocManaged, but is it possible force it allocate memory on a specific gpu id (e.g. via cudaSetDevice) on a multiple GPU system?
The reason is that I need allocate several arrays on the GPU, and I know which set of these arrays need to work together, so I want to manually make sure they are on the same GPU.
I searched CUDA documents, but didn't find any info related to this. Can someone help? Thanks!
No you can't do this directly via cudaMallocManaged. The idea behind managed memory is that the allocation migrates to whatever processor it is needed on.
If you want to manually make sure a managed allocation is "present" on (migrated to) a particular GPU, you would typically use cudaMemPrefetchAsync. Some examples are here and here. This is generally recommended for good performance if you know which GPU the data will be needed on, rather than using "on-demand" migration.
Some blogs on managed memory/unified memory usage are here and here, and some recorded training is available here, session 6.
From N.2.1.1. Explicit Allocation Using cudaMallocManaged() (emphasis mine):
By default, the devices of compute capability lower than 6.x allocate managed memory directly on the GPU. However, the devices of compute capability 6.x and greater do not allocate physical memory when calling cudaMallocManaged(): in this case physical memory is populated on first touch and may be resident on the CPU or the GPU.
So for any recent architecture it works like NUMA nodes on the CPU: Allocation says nothing about where the memory will be physically allocated. This instead is decided on "first touch", i.e. initialization. So as long as the first write to these locations comes from the GPU where you want it to be resident, you are fine.
Therefore I also don't think a feature request will find support. In this memory model allocation and placement just are completely independent operations.
In addition to explicit prefetching as Robert Crovella described it, you can give more information about which devices will access which memory locations in which way (reading/writing) by using cudaMemAdvise (See N.3.2. Data Usage Hints).
The idea behind all this is that you can start off by just using cudaMallocManaged and not caring about placement, etc. during fast prototyping. Later you profile your code and then optimize the parts that are slow using hints and prefetching to get (almost) the same performance as with explicit memory management and copies. The final code may not be that much easier to read / less complex than with explicit management (e.g. cudaMemcpy gets replaced with cudaMemPrefetchAsync), but the big difference is that you pay for certain mistakes with worse performance instead of a buggy application with e.g. corrupted data that might be overlooked.
In Multi-GPU applications this idea of not caring about placement at the start is probably not applicable, but NVIDIA seems to want cudaMallocManaged to be as uncomplicated as possible for this type of workflow.
Related
Double posted here, since I did not get a response I will post here as well.
Cuda Version 10.2 (can upgrade if needed)
Device: Jetson Xavier NX/AGX
I have been trying to find the answer to this across this forum, stack overflow, etc.
So far what I have seen is that there is no need for a atomicRead in cuda because:
“A properly aligned load of a 64-bit type cannot be “torn” or partially modified by an “intervening” write. I think this whole question is silly. All memory transactions are performed with respect to the L2 cache. The L2 cache serves up 32-byte cachelines only. There is no other transaction possible. A properly aligned 64-bit type will always fall into a single L2 cacheline, and the servicing of that cacheline cannot consist of some data prior to an extraneous write (that would have been modified by the extraneous write), and some data after the same extraneous write.” - Robert Crovella
However I have not found anything about cache flushing/loading for the iGPU on a tegra device. Is this also on “32-byte cachelines”?
My use case is to have one kernel writing to various parts of a chunk of memory (not atomically i.e. not using atomic* functions), but also have a second kernel only reading those same bytes in a non-tearing manner. I am okay with slightly stale data in my read (given the writing kernel flushes/updates the memory such that proceeding read kernels/processes get the update within a few milliseconds). The write kernel launches and completes after 4-8 ms or so.
At what point in the life cycle of the kernel does the iGPU update the DRAM with the cached values (given we are NOT using atomic writes)? Is it simply always at the end of the kernel execution, or at some other point?
Can/should pinned memory be used for this use case, or would unified be more appropriate such that I can take advantage of the cache safety within the iGPU?
According to the Memory Management section here we see that the iGPU access to pinned memory is Uncached. Does this mean we cannot trust the iGPU to still have safe access like Robert said above?
If using pinned, and a non-atomic write and read occur at the same time, what is the outcome? Is this undefined/segfault territory?
Additionally if using pinned and an atomic write and read occur at the same time, what is the outcome?
My goal is to remove the use of cpu side mutexing around the memory being used by my various kernels since this is causing a coupling/slow-down of two parts of my system.
Any advice is much appreciated. TIA.
I am writing a simplistic raytracer. The idea is that for every pixel there is a thread that traverses a certain structure (geometry) that resides in global memory.
I invoke my kernel like so:
trace<<<gridDim, blockDim>>>(width, height, frameBuffer, scene)
Where scene is a structure that was previously allocated with cudaMalloc. Every thread has to start traversing this structure starting from the same node, and chances are that many concurrent threads will attempt to read the same nodes many times. Does that mean that when such reads take place, it cripples the degree of parallelism?
Given that geometry is large, I would assume that replicating it is not an option. I mean the whole processing still happens fairly fast, but I was wondering whether it is something that has to be dealt with, or simply left flung to the breeze.
First of all I think you got the wrong idea when you say concurrent reads may or may not cripple the degree of parallelism. Because that is what it means to be parallel. Each thread is reading concurrently. Instead you should be thinking if it affects the performance due to more memory accesses when each thread basically wants the same thing i.e. the same node.
Well according to the article here, Memory accesses can be coalesced if data locality is present and within warps only.
Which means if threads within a warp are trying to access memory locations near each other they can be coalesced. In your case each thread is trying to access the "same" node until it meets an endpoint where they branch.
This means the memory accesses will be coalesced within the warp till the threads branch off.
Efficient access to global memory from each thread depends on both your device architecture and your code. Arrays allocated on global memory are aligned to 256-byte memory segments by the CUDA driver. The device can access global memory via 32-, 64-, or 128-byte transactions that are aligned to their size. The device coalesces global memory loads and stores issued by threads of a warp into as few transactions as possible to minimize DRAM bandwidth. A misaligned data access for devices with a compute capability of less than 2.0 affects the effective bandwidth of accessing data. This is not a serious issue when working with a device that has a compute capability of > 2.0. That being said, pretty much regardless of your device generation, when accessing global memory with large strides, the effective bandwidth becomes poor (Reference). I would assume that for random access the the same behavior is likely.
Unless you are not changing the structure while reading, which I assume you do (if it's a scene you probably render each frame?) then yes, it cripples performance and may cause undefined behaviour. This is called a race condition. You can use atomic operations to overcome this type of problem. Using atomic operations guarantees that the race conditions don't happen.
You can try, stuffing the 'scene' to the shared memory if you can fit it.
You can also try using streams to increase concurrency which also brings some sort of synchronization to the kernels that are run in the same stream.
I've tested empirically for several values of block and of thread, and the execution time can be greatly reduced with specific values.
I don't see what are the differences between blocks and thread. I figure that it may be that thread in a block have specific cache memory but it's quite fuzzy for me. For the moment, I parallelize my functions in N parts, which are allocated on blocks/threads.
My goal could be to automaticaly adjust the number of blocks and thread regarding to the size of the memory that I've to use. Could it be possible? Thank you.
Hong Zhou's answer is good, so far. Here are some more details:
When using shared memory you might want to consider it first, because it's a very much limited resource and it's not unlikely for kernels to have very specific needs that constrain
those many variables controlling parallelism.
You either have blocks with many threads sharing larger regions or blocks with fewer
threads sharing smaller regions (under constant occupancy).
If your code can live with as little as 16KB of shared memory per multiprocessor
you might want to opt for larger (48KB) L1-caches calling
cudaDeviceSetCacheConfig(cudaFuncCachePreferL1);
Further, L1-caches can be disabled for non-local global access using the compiler option -Xptxas=-dlcm=cg to avoid pollution when the kernel accesses global memory carefully.
Before worrying about optimal performance based on occupancy you might also want to check
that device debugging support is turned off for CUDA >= 4.1 (or appropriate optimization options are given, read my post in this thread for a suitable compiler
configuration).
Now that we have a memory configuration and registers are actually used aggressively,
we can analyze the performance under varying occupancy:
The higher the occupancy (warps per multiprocessor) the less likely the multiprocessor will have to wait (for memory transactions or data dependencies) but the more threads must share the same L1 caches, shared memory area and register file (see CUDA Optimization Guide and also this presentation).
The ABI can generate code for a variable number of registers (more details can be found in the thread I cited). At some point, however, register spilling occurs. That is register values get temporarily stored on the (relatively slow, off-chip) local memory stack.
Watching stall reasons, memory statistics and arithmetic throughput in the profiler while
varying the launch bounds and parameters will help you find a suitable configuration.
It's theoretically possible to find optimal values from within an application, however,
having the client code adjust optimally to both different device and launch parameters
can be nontrivial and will require recompilation or different variants of the kernel to be deployed for every target device architecture.
I believe to automatically adjust the blocks and thread size is a highly difficult problem. If it is easy, CUDA would most probably have this feature for you.
The reason is because the optimal configuration is dependent of implementation and the kind of algorithm you are implementing. It requires profiling and experimenting to get the best performance.
Here are some limitations which you can consider.
Register usage in your kernel.
Occupancy of your current implementation.
Note: having more threads does not equate to best performance. Best performance is obtained by getting the right occupancy in your application and keeping the GPU cores busy all the time.
I've a quite good answer here, in a word, this is a difficult problem to compute the optimal distribution on blocks and threads.
TL;DR version: "What's the best way to round-robin kernel calls to multiple GPUs with Python/PyCUDA such that CPU and GPU work can happen in parallel?" with a side of "I can't have been the first person to ask this; anything I should read up on?"
Full version:
I would like to know the best way to design context, etc. handling in an application that uses CUDA on a system with multiple GPUs. I've been trying to find literature that talks about guidelines for when context reuse vs. recreation is appropriate, but so far haven't found anything that outlines best practices, rules of thumb, etc.
The general overview of what we're needing to do is:
Requests come in to a central process.
That process forks to handle a single request.
Data is loaded from the DB (relatively expensive).
The the following is repeated an arbitrary number of times based on the request (dozens):
A few quick kernel calls to compute data that is needed for later kernels.
One slow kernel call (10 sec).
Finally:
Results from the kernel calls are collected and processed on the CPU, then stored.
At the moment, each kernel call creates and then destroys a context, which seems wasteful. Setup is taking about 0.1 sec per context and kernel load, and while that's not huge, it is precluding us from moving other quicker tasks to the GPU.
I am trying to figure out the best way to manage contexts, etc. so that we can use the machine efficiently. I think that in the single-gpu case, it's relatively simple:
Create a context before starting any of the GPU work.
Launch the kernels for the first set of data.
Record an event for after the final kernel call in the series.
Prepare the second set of data on the CPU while the first is computing on the GPU.
Launch the second set, repeat.
Insure that each event gets synchronized before collecting the results and storing them.
That seems like it should do the trick, assuming proper use of overlapped memory copies.
However, I'm unsure what I should do when wanting to round-robin each of the dozens of items to process over multiple GPUs.
The host program is Python 2.7, using PyCUDA to access the GPU. Currently it's not multi-threaded, and while I'd rather keep it that way ("now you have two problems" etc.), if the answer means threads, it means threads. Similarly, it would be nice to just be able to call event.synchronize() in the main thread when it's time to block on data, but for our needs efficient use of the hardware is more important. Since we'll potentially be servicing multiple requests at a time, letting other processes use the GPU when this process isn't using it is important.
I don't think that we have any explicit reason to use Exclusive compute modes (ie. we're not filling up the memory of the card with one work item), so I don't think that solutions that involve long-standing contexts are off the table.
Note that answers in the form of links to other content that covers my questions are completely acceptable (encouraged, even), provided they go into enough detail about the why, not just the API. Thanks for reading!
Caveat: I'm not a PyCUDA user (yet).
With CUDA 4.0+ you don't even need an explicit context per GPU. You can just call cudaSetDevice (or the PyCUDA equivalent) before doing per-device stuff (cudaMalloc, cudaMemcpy, launch kernels, etc.).
If you need to synchronize between GPUs, you will need to potentially create streams and/or events and use cudaEventSynchronize (or the PyCUDA equivalent). You can even have one stream wait on an event inserted in another stream to do sophisticated dependencies.
So I suspect the answer to day is quite a lot simpler than talonmies' excellent pre-CUDA-4.0 answer.
You might also find this answer useful.
(Re)Edit by OP: Per my understanding, PyCUDA supports versions of CUDA prior to 4.0, and so still uses the old API/semantics (the driver API?), so talonmies' answer is still relevant.
I'm working on data prefetching in nVidia CUDA. I read some documents on prefetching on device itself i.e. Prefetching from shared memory to cache.
But I'm interested in data prefetching between CPU and GPU. Can anyone connect me with some documents or something regarding this matter. Any help would be appreciated.
Answer based on your comment:
when we to want perform computation on large data ideally we'll send max data to GPU,perform computation,send it back to CPU i.e SEND,COMPUTE,SEND(back to CPU) now whn it sends back to CPU GPU has to stall,now my plan is given CU program,say it runs in entire global mem,i'll compel it to run it in half of the global mem so that rest of the half i can use for data prefetching,so while computation is being performed in one half simultaneously i cn prefetch data in otherhalf.so no stalls will be there..now tell me is it feasible to do?performance will be degraded or upgraded?should enhance..
CUDA streams were introduced to enable exactly this approach.
If your compoutation is rather intensive, then yes --- it can greatly speed up your performance. On the other hand, if data transfers take, say, 90% of your time, you will save only on computation time - that is - 10% tops...
The details, including examples, on how to use streams is provided in CUDA Programming Guide.
For version 4.0, that will be section "3.2.5.5 Streams", and in particular "3.2.5.5.5 Overlapping Behavior" --- there, they launch another, asynchronous memory copy, while a kernel is still running.
Perhaps you would be interested in the asynchronous host/device memory transfer capabilities of CUDA 4.0? You can overlap host/device memory transfers and kernels by using page-locked host memory. You could use this to...
Copy working set #1 & #2 from host to device.
Process #i, promote #i+1, and load #i+2 - concurrently.
So you could be streaming data in and out of the GPU and computing on it all at once (!). Please refer to the CUDA 4.0 Programming Guide and CUDA 4.0 Best Practices Guide for more detailed information. Good luck!
Cuda 6 will eliminate the need to copy, ie the copying will be automatic.
however you may still benefit from prefetching.
In a nutshell you want the data for the "next" computation transferring while you complete the current computation. to achieve that you need to have at least two threads on the CPU, and some kind of signalling scheme (to know when to send the next data). Chunking will of course play a big role and affect performance.
The above may be easier on an APU (CPU+GPU on the same die) as the need to copy is eliminated as both processors can access the same memory.
If you want to find some papers on GPU prefetching just use google scholar.