Mind-bending new programming language for GPUs just dropped...

preview_player
Показать описание
What is the Bend programming language for parallel computing? Let's take a first look at Bend and how it uses a Python-like syntax to write high performance code that can run on the GPU.

#programming #tech #thecodereport

💬 Chat with Me on Discord

🔗 Resources

🔥 Get More Content - Upgrade to PRO

Use code YT25 for 25% off PRO access

🎨 My Editor Settings

- Atom One Dark
- vscode-icons
- Fira Code Font

🔖 Topics Covered

- What is Bend language?
- Parallelism vs Concurrency
- Rust programming language projects
- CUDA alternatives
- How to run code on a GPU
Рекомендации по теме
Комментарии
Автор

Time to add 10 years of experience to my resume/CV

hamadaelwarky
Автор

That video was *extremely* well done. Seems like someone has read the docs! Thanks for making it. We're long-term fans of your content here at HOC. If you or any other content creator wants to reach out, we're available to talk about the technology and answer any questions.

We know there's a lot of work to do, but we're excited about it. You can expect Bend to become faster with every release. There are also many missing features (64-bit numbers, larger memory limit, etc.) that will be added very soon!

VictorTaelin
Автор

Folds have been a mainstay of the functional programming world for a very long time.

I remember teaching students how to implement folds in Python a few years back and it blowing their minds. The basic concept that all loops are simply a subset of a wider class of iterative constructs called folds is transformational - I love the simplicity and inherent understanding of a fold.

This new concept of a bend is incredibly interesting, as like the fold it appears to be able to emulate any kind of loop or iterative function, but instead of going from many to one, goes from one to many. I will be playing around with this over the next few weeks I am sure!

samwalker
Автор

Watched this video on 8000 Cores in parallel, added 500h of bend experience to my resume

noname-qlfn
Автор

“1 week of problem, instead 7 days with 7 computers” 🤣🤣🤣

jonathanjeshualaniba
Автор

We needed this 17 years ago.
It is happening 17 years too late.
Imagine what we could have done with those core2quads and Phenoms !!

pamus
Автор

People are rediscovering functional programming without knowing it.

EllGeeLabs
Автор

someone tomorrow will start recruiting people with 10 years of experience in this

farouk_bloncko
Автор

Saw this on Twitter/X and was like 'no way, this is a scam'. Then I saw their pfp was an anime avatar and ngl kinda trusted them immediately lmao

_ptoni_
Автор

That diagram showing the combinators being untangled is just so awesome.

richtigmann
Автор

Yoo!! The guy who wrote that paper you mentioned at 2:15 was one of my teachers at my first year of my maths bachelor, that's sick

ChrisMazzerbo
Автор

But GPU guy said we will never have to code again.

cherubinth
Автор

Yoooo new programming thing dropped that's not a new js library finally letsgoo

neposis
Автор

I'm not in tech but I've been following this one after a few people on Haskell Twitter mentioned it, the first of whom exclaimed something along the lines of "this guy is turning GPUs into real modern LISP machines!" which is exciting in and of itself, and then I think Ed Kmett engaged with something indicating that this does reduction of something akin to the lambda calculus really efficiently with GPU acceleration which is REALLY exciting

MorganEarlJones
Автор

7 days 7 computers joke was hilarious 🤣🤣🤣

komeelali
Автор

Parallel Language: Check.
Supports GPU: Check.
Built with Rust(has thread safety, fast, beautiful, runs everywhere): What? ALL IN. let's integrate to Python and push to production now.

mephilees
Автор

I love that he still puts the “Hi Mom” references in. Gets me every time.

tylerwalton
Автор

You're in a league of your own bro, these videos are hella informative and have me rolling lmao

KingVoodoo
Автор

1:20 “may even lead to conflicts with demons”😂

Walker-kyvy
Автор

0:30 So THAT'S why my CUDA code wasn't running close to optimally... it's not my sophomoric understanding of algebra, it was that I wasn't leveraging REGEX!!!

douglaskrause