Aster Dev (@aster_dev_) header
Aster Dev (@aster_dev_) profile picture

A confused generalist. Wannabe GameDev, forced to write Kernels for Inference

Joined January 2026
Have a j*b? I just built a simple Leave Planner tool which incorporates all of India's national holidays and helps you maximize a window of leaves. You can plan ahead without the worry of no vacancy be it hotels or flights. Check it out: asterdev.vercel.app/tools/ho…
1
2
33
I just put up my own RSS feed on my portfolio website Do give your opinions on it :)
6
Replying to @abzbuilds
Let's connect dude :)
1
1
7
Replying to @AYUSHKUMAR82274
Lets connect dude :)
1
1
7
Replying to @clicknify
Happy to connect :)
4
Replying to @divyansh_1729
Isnt it just thinking with extra steps
194
Replying to @ATaylorFPGA
That would be great! I wanted to dive into fpga programming uk recently AI is writing llm weights directly. I dont get FPGAs but i think weighted chips like Tallas are the future and FPGAs are a great start.
13
Replying to @charliehustler9
which which :0
1
1
4
Replying to @agupta
I hope to replicate this sooner or later as a student. Vibecoding for a expert = Understanding and implementing for a novice in Systems Programming
84
Replying to @ThePrimeagen
They call it "simplicity" in their lingo
12
Replying to @BharatJain
Have been wanting to do the same :) Gonna follow this journey
53
Replying to @parthshr370
Gonna read this sooner or later...
1
9
Replying to @prathoshap
And guess what? Literally a day after their IPO, after almost a $400B crash. How convenient right?
574
From tomorrow, I'll start a series to complete my blog on matrix multiplication on GPUs. Have been procrastinating on it for over a month. Gotta finish it before endsems start.
1
13
Have heard great things of this book. Existentialism is one of the topics I love to dwell on. Hope to have a great read
8
Replying to @yifanzhang_
And here I though I would be the one to solve it (delusionally)
35
Replying to @prathoshap
How have they not run out of business with the advent of AI? Anyone doing research is decently equipped to fake a paper right. I wonder if it is now only the connection with the fake journal/pub that gives this incentive to pay?
41
Replying to @Arnesh_24
So.... no time limit?
1
1
5
The constant learn - implement - forget - repeat loop is exhausting
3
The @Docker logo change just upset me and ruined my day
9