WHY! is this so underrated? excellent explanation

*"If (Score > 85) { std::print("pass"); } else { std::print("failed");}"* Ahh Test😮

is a double and floating point number the same thing?

Loving the course so far!

Great series, just setting up SIL harness for my C project this would help!

i want to learn more c++ software development knowlege from you

thanks very much for your great explaination

Short, nice, clear and straight to the point, no IDE no hidden config files, just simple command line & gcc Thanks a lot Ps: it would be nice if you showered us how can we automate this using makefile

Thanks dude, feels like the first time its explained to me from a normal person

For the sum example, it looks more like the form (pack op ...)? Why have you said it's of the form (... op pack) when the fold expression is on the right?

Great video!

What I didn't understand is how integers were 4 bytes apart

Very nice, thank you

The only useful cuda tutorial that I could find to fit my needs in 2025!

Bravo pour la vidéo.;)

Why are they scaled by 0.85

Bro get recommended to me by ChatGPT. Great video, tho!

Great explanation !

twohundredandfiftysix

explanation was great, super easy to follow. also I like your beard

Could you please explain more clearly why this solves bank conflicts? With the sequential code, th 0 accesses integers(32 bit) at addr 0,1 which are in banks 0,1. th1 accesses ints at addr 2,3 and th 255 will access ints at 255*2 and 255*2+1 . I understand that when these addresses %32 are the same, they access the same bank and require serial access. But with the new code, th 0 accesses addr 0,128, th 1 accesses addr 1,129 and so on. addr 0 and 128 access bank 0, so aren't they causing conflict as well?

what is the purpose of the for loop in the kernel? why not just use a[tid] for the single item this thread will handle?

Thank you so much

well explained

May you give an order or sort for playlists in your channel if someone start programming from scratch? Thanks 😊

Superb video structure. you started with introducing valgrind, then essential useful commands and how to use valgrind without going too much in detail like those tutorials hell. amazing 👌🏻

I swear you are probably one of the most underrated content creator ever

In the past, Nearly every time I have used constexpr, eventually, the requirements change, and I can no longer use it. 99% of programming is not mathy, but loaded with business rules.

Awesome! Thank you!

kya baat ha bhai maza agaya

salut , où est ce que tu pourrais nous fournir de la documentation ?

Thank you Nick ! that was helpful and straight to the point, you're amazing!

It helps!

nice video

Hi. I think we should use TB_SIZE * sizeof(int) for the size of our dynamic shared memory in this code. Is n * sizeof(int) necessary? Is not it too big for a shared memory?

What might be the reason for me not seeing cache misses (~2% vs ~0.1%) when compiling and running direct_sharing example on my pc?

Amazing work!

This is so clutch thank you

9:10 I can’t figure out what is non-intuitive about this. Processor 0 writes to x and reads to y, processor 1 writes to y and reads to x. I guess if you expect at least r0 or r1 to be 1, it would be a bug, but idk why that would ever come up…

OMG I am so excited for this new series

Thank you! Very useful and clear to understand.

I think what you explained about NRVO example (the first example) is called mandatory copy Ellison! NRVO is when no temporary created from return statement and value in front of return statement returns not its copy!

Thank you! I was wondering if there is a book or something to explain this topic specifically ?

Great video :-)

You didn't explain what (N)RVO is in the first place! It is a feature that allows us to optimize away the creation of a return value from a named object in a return statement. If we already have a local object that we return in a function by value, the compiler can use the object directly as a return value instead of copying or moving it. Also @8:02 you should have made it clear that NRVO is a form of copy elision so we didn't get NRVO(a form of copy elision)

Love these random videos being helpful and better than other videos.