Currently Founder and CTO @dotproduct3d / Before Mobile Computer Vision @Google / Before Founder @ Enkin AR (acq‘d by Google) / Before uncool geek @ school
Mar 20 • 6 tweets • 2 min read
NVIDIA just announced new GPUs @NVIDIAGTC delivering up to 20 petaFLOPS per chip.
Many don't realize what a staggering, absurd amount of compute that is, so I'd like to provide some perspective:
* With a few gigaFLOPS, you can run basic image processing operations on
1/x
every pixel of a VGA (640x480) stream of images, in real-time (30-60 fps)!
This is roughly what a Raspberry Pi has in compute.
Basic image filters, AR, even full visual SLAM / mapping fit entirely in that realm.
* With one teraFLOP, you could run those same operations, in
2/x
Dec 4, 2022 • 25 tweets • 6 min read
NeRFs are getting attention these days!
However widespread adoption is still slow. But why?
It comes down to a) file size and b) rendering.
An engineer's viewpoint:
TLDR: NeRFs are fundamentally a bad fit for today's edge device architectures.
Let's explain that in detail:
1/
a) A quality NeRF can easily take up many GB of (dedicated) memory on the GPU. Most consumer devices today are miles away from this.
But even if devices could handle it (they can't) there is still the open question of delivery / transmission.
While you could, presumably
2/
