Tutorial: CUDA programming in Python with numba and cupy

Подписаться 915

Просмотров 74 тыс.

50% 1

/Using the GPU can substantially speed up all kinds of numerical problems. Conventional wisdom dictates that for fast numerics you need to be a C/C++ wizz. It turns out that you can get quite far with only python. In this video, I explain how you can use cupy together with numba to perform calculations on NVIDIA GPU's. Production quality is not the best, but I hope you may find it useful.
00:00 Introduction: GPU programming in python, why?
06:52 Cupy intro
08:39 Cupy demonstration in Google colab
19:54 Cupy summary
20:21 Numba.cuda and kernels intro
25:07 Grids, blocks and threads
27:12 Matrix multiplication kernel
29:20 Tiled matrix multiplication kernel and shared memory
34:31 Numba.cuda demonstration in Google colab
44:25 Final remarks
Edit 3/9/2021: the notebook is use for demonstration can be found here colab.research.google.com/dri...
Edit 9/9/2021: at 23:56 one of the grid elements should be labeled 1,3 instead of 1,2. Thanks to _______ for pointing this out.

Наука

Опубликовано:

27 июн 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 73

@ErolErten Год назад

I have been looking into gpu programming using numba and python for a while, this seems to be the best tutorial I was able to find so far.. . thank you

@Omgtired 2 года назад

Thank you so much. Probably the best introdution to CUDA with Python. The example you use, while very basic, touches on usage of blocks, which is usually omitted in other introduction-level tutorials. Great stuff! Hope you return with some more videos. I have subscribed!

@kayakMike1000 Год назад

Cuda is bullshit closed source. Just wait for Tenstorrent, it's gonna be HUGE.

@vallurirajesh 2 года назад

Thank you so very much. This is the exact kind of material I was looking for on this very specific subject. Kudos.

@taj-ulislam6902 2 месяца назад

Definitely a lot of new material not seen else where - not a run-of-the-mill video. Great job on originality.

@prietjepruck Год назад

Really great introduction to GPU programming. I hope you make a new one soon.

@sciencewolf963 2 года назад

Excellent explanation, keep going with this content man ;)

@leaodev 2 года назад

Great video, nick!

@jakob3267 2 года назад

Really nice video, thank you for sharing!

@Zysperro 2 года назад

Just what I needed! Thanks!

@thousandTabs Год назад

this was such an excellent video, thank you so much!

@kineticraft6977 11 месяцев назад

This reminds me a lot of the mindset you need to program in assembly.

@silkworm6861 2 года назад

This is a great video!

@ouaililydia3835 Год назад

thank you so much, it is the best explaination i found. Please keep going and give us more information and examples on that

@andrjo 2 года назад

wanted to comment that the information in this presentation is very well structured and the flow is excellent.

@nickcorn93 2 года назад

Thanks man!

@terriplays1726 2 года назад

Thanks for the video, I found the first half and the wrap up really excellent.

@shaheeng8034 4 месяца назад

Thanks a lot! Still the best guide I could find.

@Shoz_ Год назад

Thank you, this is gold

@LoneXeaglE 2 года назад

Thank you so much sir, you are an amazing human being !

@dfrank5157 2 года назад

This is really helpful for my computing. Thank you.

@PhoenixReflex 7 месяцев назад

Thank you so much. Keep up the hard work. Just hoping that more and more libraries in python will support GPU computations soon.

@srepmub 2 года назад

fantastic video.

@duongkstn Год назад

great tut ! thanks

@timharris72 2 года назад

This was really good. Thanks for posting this!

@mfatihaydogdu7 2 года назад

Very helpful, thank you.

@ArijitBhattacharya971 2 года назад

wold love to see a video on what are a few CUDA programming challenges

@therealbatman664 2 года назад

Thanks a lot really got me started .

@tooniatoonia2830 Год назад

Really learnt a lot here, thanks!💪

@localhost_mds Год назад

thank you. good video!!! it was very helpful

@lfmtube 2 года назад

Perfect Video! Saw was revealing to me to understand how it works. Thank you! I am a new subscriber of your channel. Regards from Buenos Aires, Argentina

@user-tx1we1hw8b Год назад

thank you! super helpful

@______373 2 года назад

wait i tought that this made by some popular channel, done pretty well and then saw, 29 subscribers

@nickcorn93 2 года назад

you would be surprised what powerpoint can do. To be honest I don't enjoy making videos that much, it's a lot of work, it always turns out kind of shit (especially audio and webcam quality), and I get nothing in return. But when I encounter a really niche topic that I struggled with myself and I don't find many resources for it I figure I make it myself hopefully such that it may be useful to someone else.

@______373 2 года назад

@@nickcorn93 "nickcorn93 nickcorn93 2 hours ago you would be surprised what powerpoint can do." not only powerpoint))))))

@nucspartan321 Год назад

Great video

@Khaled_Elsadani 7 месяцев назад

Thanks for sharing INFO

@rezidwipradana495 2 года назад

Thank you very much

@mattiskardell 5 месяцев назад

Thank you so much

@plumberski8854 Год назад

Great intro for me. Waiting for my new GPU (likely 4060 Ti) for me to dig deeper into Python, CUDA, deep learning ...

@Julian-tf8nj 2 года назад

VERY helpful, thank you!!!!

@1Eagler Год назад

Very educational. One thing I've missed: The function matmul is running on the PC or the GPU?

@AngeloHafner 8 месяцев назад

Muito bom...

@zaharkohut7881 Год назад

Thank you for this tutorial, it has been very helpful! But since it is only an introduction could anyone tell me what I should watch or read next on this topic? Thanks in advance for the advice!

@garywilliams4214 10 месяцев назад

Great tutorial, Nick! One minor critique: your pronunciation of ‘array’ was confusing…a more standard pronunciation is “uh-RAY”.

@user-um9sl1kj6u 11 месяцев назад

What about if you want to develop a library for neural net work? A highly specialized library

@glenneric1 2 года назад

You say ARRay, I say arRAY. Let's call the whole thing off. But seriously, good stuff.

@Julian-tf8nj 2 года назад

I kept thinking, "huh? what is he talking about?? Oh, he meant an ARRay!" lol Other than that, awesome vid!

@nickcorn93 2 года назад

Interesting, so I've basically been pronouncing array incorrectly my whole life. Will try to watch out for that in the future.

@glenneric1 Год назад

@@nickcorn93 I've heard other people saying it your way too.

@rweaver6 Год назад

@@nickcorn93 it was very distracting. Work on it google it and use the pronunciation feature. Otherwise outstanding and very useful tutorial.

@niffoxichere8394 2 года назад

is it only me or the cooling fan going brrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr.

@HectorHernandez-ws3el 2 года назад

Thanks for the video, it isn´t very information about, sorry for my english

@0Clappy Год назад

Can you do a tutorial series on how to accelerate things using cuda python?

@nickcorn93 Год назад

I've thought about it but it's a lot of work to make and edit a silly video like this, and at the moment I really don't have the time. I don't get anything for making these videos.

@jakubkahoun8383 2 года назад

Hi, I m trying this on my local computer, but cannot install Cupy, I have NVida geforece RTX 3060. EDIT: Installed CUDA 11.6 toolkit and it works now.

@nickcorn93 2 года назад

What is your OS? You may be having issues if you are using windows and pip. Easiest to install cupy in a conda virtual environment, as it will also install the cuda toolkit.

@jakubkahoun8383 2 года назад

@@nickcorn93 Sorry for bother you, the problem was not installing Cuda Toolkit, srly I hate people who doesnt watch full video closely and ask stupid questions....and now I m one of them :D. Thx alot for this tutorial in 2 months i will try write my own GPU operator for my program, would be interting if this will be faster than CPU. (Btw using normal Visual code in python 3.10 env. on win 11, so far so good. (Altrough i have some code output delay problem when using openCV for some strange reason)

@richardbennett4365 Год назад

Wait. At 12:10, the narrator says the timeit magic function reports a duration of 5 ms, but the number is 0.01 ms from 6 ms. The number us far away from 5 compared to 6. It shoukd be 6 ms if he's rounding, not 5 ms. He's truncating the decimals to arrive at an integer.

@nickcorn93 Год назад

Congratulations, you have invalidated the entire video by spotting this massive mistake ;) !

@richardbennett4365 Год назад

@@nickcorn93 🆗.

@billyblackburn864 2 года назад

hi, I have a program that I want to translate to numba. could you help me?

@nickcorn93 2 года назад

- what should the program do? - who is the program for? - what is it currently written in?

@wrcz 24 дня назад

all these tutorials using light mode while I learn at night... I'm gonna go blind :X

@gauravdeshpande4298 Год назад

I am unable to install cupyx from pip any help

@kayakMike1000 Год назад

GPUs aren't general purpose... sigh... They are really good at specific executing the same operation on many data banks. It just happens to be similair type of needs for graphics an machine learning

@nickcorn93 Год назад

Isn't that what I say in this video? Did you even watch it?

@jesusmtz29 4 месяца назад

Approximate arbitrary function? There are caveats.

@nigmaxus Год назад

Cupy does not install well through the use of pip

@nickcorn93 Год назад

typically it is easier via conda yes.

@TheAIEpiphany 11 месяцев назад

Something is seriously off with your fast matmul implementation, it's 3 orders of magnitude slower than the built-in method (12.5 ms vs 8.82 us)? You probably have some host-device copying going on?

@nickcorn93 11 месяцев назад

The matmul example shown is the example from the numba documentation so I don't think it's wrong. It's (relatively) slow because matrix multiplication is something that is so common, it is insanely optimized in available implementations. You won't write a matrix multiplication implementation with numba that's faster than cupy. But if you have something custom you need to do, a custom kernel can be faster than a combination of cupy operations.

@snapo1750 Год назад

There is a python opencl package (pyopencl) a = pyopencl.array.arange(queue, 400, dtype=numpy.float32) b = pyopencl.array.arange(queue, 400, dtype=numpy.float32) krnl = ReductionKernel(ctx, numpy.float32, neutral="0", reduce_expr="a+b", map_expr="x[i]*y[i]", arguments="__global float *x, __global float *y") my_dot_prod = krnl(a, b).get() 🙂 Benefit is it works on ALL GPU's not only Nvidia, (works on intel built in cpu gpu's and on amd gpus)