This is a nice overview, thanks. Your drawing where you show a quad of GPUs connected by point-to-point NVLinks replaced by switches is incorrect.
That said, the point you make that point-to-point links statically divide node bandwidth in proportion to the number links, and that using switches instead allows use of the full bandwidth of the link bandwidth to be used for any connection, is correct. It’s just the drawing that needs fixing.
This is a nice overview, thanks. Your drawing where you show a quad of GPUs connected by point-to-point NVLinks replaced by switches is incorrect.
That said, the point you make that point-to-point links statically divide node bandwidth in proportion to the number links, and that using switches instead allows use of the full bandwidth of the link bandwidth to be used for any connection, is correct. It’s just the drawing that needs fixing.
Thanks, Sharad. I've fixed it!
Thanks man basic and simple
Thank you! Glad you found it helpful :)
Great way of simplifying things. Enjoyed reading
Thank you!
The latest and greatest going to: https://substack.com/profile/325616327-firewall/note/c-305382037?r=5dv35z&utm_source=notes-share-action&utm_medium=web