Darshan @neuronfitting

21 y/o making gpu's go brrrrr dcbaslani.xyz Joined December 2024

Tweets

813
Followers

60
Following

430
Likes

23K

Darshan @neuronfitting

17 hours ago

@ranaharshraj7 @NVIDIAAI congrats! check this out: dcbaslani.xyz/blog/gpu_maste…

0 1 2 136 4

View Details

Darshan @neuronfitting

2 days ago

process for generating knowledge is virtually indistinguishable from process for generating speech

0 0 0 17 0

View Details

@maharshii what is your workflow for writing kernels? i.e. if you wanted to write a GDN kernel and you didn't knew anything about GDN, how would you start? and also what excites you about writing kernels?

0 0 0 71 0

View Details

Naval @naval

2 days ago

Science is not a process, a credential, or an institution. It is the unflinching pursuit of truth, carried out by the few, co-opted by the many.

762 2K 14K 1.0M 1K

View Details

Paras Chopra @paraschopra

2 days ago

I see a lot of enthusiasm about building sovereign models on my timeline. That's great to hear and India needs it, BUT.. building a Fable-class model is a compute and funding game. Last I checked, India had ~50-100k H100 equivalents while frontier labs would have a million each. Unless we have a paradigm shift in how AIs are trained, the conversation ought to be happening about amount of funding available to do what we want to do. Show me an Indian company that's secured funding/compute in the same range as that of Chinese AI labs (let alone American labs). Without compute, what will happen is what has happened before: we'd promise to shake the world and then build models that are a year or two behind the top ones. The path forward for sovereign models that I see is to invest in basic R&D so we have a chance to go beyond the current paradigm, OR the government pooling in several orders of magnitude more compute to seriously commit competing at par.

92 87 1K 72K 179

View Details

Darshan @neuronfitting

3 days ago

wtf?

Anthropic @AnthropicAI

3 days ago

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of

13K 26K 88K 89.9M 24K

0 0 1 13 0

View Details

Darshan @neuronfitting

a week ago

To feel the burning itch of curiosity requires both that you be ignorant, and that you desire to relinquish your ignorance.

0 0 0 28 0

View Details

Darshan @neuronfitting

a week ago

@prathamgrv Where in blr?

0 0 0 549 0

View Details

Darshan @neuronfitting

a week ago

opposite of happiness isn't sadness, its boredom.

0 0 0 31 0

View Details

Darshan @neuronfitting

a week ago

People should get smarter at a rate sufficient to integrate their old experiences, but not so much smarter so fast that they can't integrate their new intelligence. Being smarter means you get bored faster, but you can also tackle new challenges you couldn't understand before.

0 0 1 33 0

View Details

Darshan @neuronfitting

a week ago

@jino_rohit find the gaps, fill the gaps

0 0 0 61 0

View Details

Darshan @neuronfitting

2 weeks ago

Introducing your new modern GPU blueprint. Read the full post here: dcbaslani.xyz/blog/gpu_maste…

0 0 0 14 0

View Details

Darshan @neuronfitting

2 weeks ago

Register files used to be the ultimate bottleneck for Tensor Core accumulators. Introducing Blackwell’s Tensor Memory (TMEM), a completely new address space inside the SM that isolates the accumulator entirely from the register file.

1 0 0 22 0

View Details

Darshan @neuronfitting

2 weeks ago

"What I cannot create, I do not understand." Introducing: The Feynman GPU Lectures. Your H100s and B200s are running at a fraction of their peak utilization because your custom kernels are written with massive hardware bottlenecks. If you don't know what tcgen05. mma does at the wire level, you're lighting compute efficiency on fire.

1 0 2 47 0

View Details

Darshan @neuronfitting

2 weeks ago

@maharshii Whats wrong with that?

0 0 0 142 0

View Details

Darshan @neuronfitting

2 weeks ago

Open source is catching up to fronteir labs!

MiniMax (official) @MiniMax_AI

2 weeks ago

Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities - Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas - MiniMax Sparse Attention scales context to 1M -

563 1K 10K 5.1M 3K

0 0 0 27 0

View Details

Darshan @neuronfitting

3 weeks ago

This is crazy work!!!!

Elon Musk @elonmusk

3 weeks ago

SpaceX has almost finished writing V1.0 of an in-house AI training stack in C that exact-maps to 220k GB300s with 800G NICs, making heavy use of pipeline parallelism and getting as close to bare metal as possible. The potential speed improvement vs JAX for large training runs is