Drona

GSoC '26 @ProjectHoneynet | CS '28 Undergrad
This is where I dump my thoughts and rants.

[Drona @dornaoffical]

Claiming it now, 2083 is going to be my year 🤞

[Drona @dornaoffical]

Today was a bad day. Blew an easy math paper for no reason at all. Looks like I’ll be reappearing for this one… lol

[Drona @dornaoffical]

Passed the gsoc final eval. past 3 months were full of learning

Wrap-up blog https://t.co/9QwjRbrjqT

[Drona @dornaoffical]

What looks impossible from today's perspective may become straightforward under different circumstances.

[Drona @dornaoffical]

I haven’t been able to concentrate on work these past few days. I keep scrolling social media for flood updates, but seeing the destruction is deeply unsettling, and I am seeing misinformation is spreading rapidly on SM . Please verify sources before sharing..

[Volcaholic 🌋 @volcaholic1]

🚨 China's Ministry of Water Resources has issued an urgent alert warning that a newly formed barrier lake near the Nepal-Tibet border poses a high risk of breaching within 72 hours.

Formed at the confluence of the Chhochen Khola and Purepu Tsangpo rivers by yesterdays massive rock-and-mudslide, the overflowing lake already holds roughly 2 million cubic metres of water.

The lake faces an additional 3 million cubic metres of inflow from expected severe weather. A collapse would trigger a catastrophic downstream surge along the Bhote Koshi-Trishuli-Narayani river system.

[Nepali Army @NepaliArmyHQ]

जलविद्युत् सुरूङमा फसेकाहरुको उद्धारमा #नेपालीसेना....
जहाँ संकट, त्यहाँ #नेपालीसेना !!!
#NepaliArmy
#BhotekoshiFlood

[Speedy Updates @SpeedUpdates1]

IShowSpeed meets an Argentina fan from Nepal at the FIFA World Cup Quarterfinal 🇦🇷🇳🇵❤️

"Chat, we gotta go to Nepal!" 🇳🇵🔥

[Drona @dornaoffical]

midterm evaluation is complete, and the learning experience so far has been immense. now ready for the next phase

[Drona @dornaoffical]

I need to dive deep into distributed systems and system design. Also learning Go and building some interesting projects. Recently, I've been spending a good chunk of my time reading PDFs and books rather than coding... and I'm loving the journey

[Drona @dornaoffical]

After a long break, I've decided to take DSA seriously again. After all, having an algorithmic mindset is always a huge advantage in tech

[Drona @dornaoffical]

someone out there is silently rooting for you, praying for your success, and watching you grow from afar. They don’t need the credit, they just want to see you win :)

[Drona @dornaoffical]

Built a simple sandboxed execution runner

It securely isolates multi-stage workflows inside an ephemeral containerized environment, using a custom DAG scheduler to parse tasks, and gVisor (runsc) to intercept syscalls and shield the host kernel.

code : https://t.co/VrNWFTJeZK

[Drona @dornaoffical]

monotonous feeling can become a real bottleneck. I go through it too quite often. Whenever it hits for too long, I try to take a break, go somewhere, reset a bit, and then come back with a clearer mind. That’s usually how I tackle it.

[Drona @dornaoffical]

@axi_master @aadityasubedi_ @architectlabs Congrats 🎉

[Ebrahim H @axi_master]

Today, @aadityasubedi_ and I are excited to introduce @architectlabs.

We are building the AI system to design and provably verify chips for the world's most demanding workloads.
AI scaling is fundamentally changing the economics of hardware infrastructure. As models scale, become more capable, and more widely deployed, the bottleneck is shifting from software and models alone to the physical infrastructure that runs it: specialized compute, memory, networking, interconnects, and full-stack system design.

General-purpose hardware is no longer enough, and the world is racing to spin up new chip programs. This is not just true across datacenter training and inference, but everywhere AI enters the physical world: robotics, autonomous systems, spatial computing, defense, personal devices, industrial automation, and scientific instruments.

But designing a chip today remains one of the most gated efforts in modern technology.

A modern chip program takes years, costs hundreds of millions of dollars, and depends on a shrinking pool of expertise concentrated inside a small number of companies.

Architect Labs is a foundational lab building an AI system that designs and provably verifies chips end-to-end. We partner with semiconductor and workload companies, AI labs, and nations to turn demanding workloads into purpose-built chips, on demand at scale. We aim to drastically accelerate chip design, so that the models, software and chip designs can co-evolve together, accelerating the industry’s path to superintelligence.

Two decades ago, the fabless revolution made it possible to build a chip company without owning a fab. TSMC made world-class manufacturing available to anyone with a design. We aim to do the same for chip design itself: enable any organization with a workload, or specification, to get a purpose-built chip design that unlocks scale and distribution of intelligence impossible with current hardware paradigms.

Our founding team collectively has taped out 80+ production chips, led $10B datacenter product lines, been core contributors to Meta’s AI silicon, architected and designed one of the first neuromorphic chips out of Intel, led research teams at Anthropic, xAI, and Google DeepMind, and contributed to fundamental AI research across nearly every frontier lab.

We have already partnered with semiconductor companies to accelerate their chip programs, and some of our AI-generated chip designs are going to tape out on leading-edge foundry nodes later this year.

We’ve also raised a $24M seed round led by @stevejang from @KindredVentures , with participation from @TQVentures , @RaceCapital , @scaletogether , @ora, and Link Ventures.

We are grateful for the support of our angels and advisors, including @snsf , @lukaszkaiser , @AravSrinivas , Kunle Olukotun, @tlbtlbtlb, @alexwg, Siddharth Nath, Thierry Tambe, @arashf , @ekaurghar, @CHHubbell, Selene Casabal, @semiDL , and engineering leaders from OpenAI, NVIDIA, Google DeepMind, Intel and more.

The next great scaling law may not come from the models alone. It will come from making the physical substrate of intelligence programmable. We exist to bring this future to life.

[Balen Shah @ShahBalen]

MAKALU BASE CAMP TREK

Makalu Barun Valley is a remote Himalayan paradise in eastern Nepal, known for its breathtaking landscapes, rich biodiversity, and spectacular mountain scenery.

Trek Overview

Duration: 17-18 Days
Maximum Altitude: 4,870m (Makalu Base Camp)
Region: Eastern Nepal
Difficulty: Challenging
Best Seasons: Spring (March–May) & Autumn (September–November)

Day-by-Day Itinerary:

Day 1 | Kathmandu → Tumlingtar → Num

Option 1: Flight from Kathmandu to Tumlingtar
Option 2: Drive from Kathmandu to Tumlingtar (approximately 14–16 hours).

Day 2 | Num → Seduwa
Trek through terraced fields and local villages. (5 hrs)

Day 3 | Seduwa → Tashigaon
Enter Makalu Barun National Park. (5 hrs)

Day 4 | Tashigaon → Khongma Danda(6 hrs )

Day 5 | Khongma Danda → Dobato
Cross scenic ridges with mountain views. (6 hrs)

Day 6 | Dobato → Yangle Kharka
Descend into the beautiful Barun Valley. (7 hrs)

Day 7 | Acclimatization Day
Rest and explore the surrounding area

Day 8 | Yangle Kharka → Shiva Dhara → Yangle Kharka
Excursion to the sacred Shiva Dhara waterfall. (5–6 hrs)

Day 9 | Yangle Kharka → Langmale Kharka
(6-7 hrs)

Day 10 | Langmale Kharka → Makalu Base Camp (5-6 hrs )

Day 11 | Explore Makalu Base Camp
Enjoy panoramic Himalayan views.

Day 12 | Makalu Base Camp → Yangle Kharka
Begin the return trek. (7 hrs)

Day 13 | Yangle Kharka → Dobato (6-7 hrs)

Day 14 | Dobato → Khongma Danda (6-7 hrs )

Day 15 | Khongma Danda → Seduwa
Descend through forests and villages.

Day 16 | Seduwa → Tumlingtar

Day 17 | Tumlingtar → Kathmandu

[Drona @dornaoffical]

Done for today, did a bit of debugging and built a proper framework. What feature should I add next?

From Monday, exam week starts, so I’ll be solely focused on that :(

[Drona @dornaoffical]

gVisor was breaking my Ubuntu containers.

Turns out "apt" needs root + kernel capabilities, which get restricted for isolation.

Instead of forcing it, I switched to prebaked images.

Simple fix, though learned a lot with this errors...

[Drona @dornaoffical]

lately, doing some math stuff. It is good, though sometimes you spend your time with a pen and paper rather than a keyboard :)

[Drona @dornaoffical]

using Docker API, and gVisor, it now spins up isolated containers, handles auto-pulls, streams real-time logs, and deletes everything cleanly afterward.

next I will build an application-layer cmd injector to automatically set up base environments dynamically on the fly :)