Harsh @itzHarsh08

19 yo | AI undergrad '29 Joined February 2025

Tweets

494
Followers

294
Following

2K
Likes

10K

Harsh @itzHarsh08

19 hours ago

yeh wala que toh easy tha ... abhi kara 15-20 min mai hogaya without hint contest mai proper time nhi diya aur -35 🥀🙂

0 0 1 25 0

View Details

Diksha Kandpal🇮🇳 @DikshaKandpal8

3 days ago

Government to public at the stroke of the midnight hour

Press Trust of India @PTI_News

3 days ago

Domestic cooking gas LPG price hiked by Rs 29 per 14.2-kg cylinder: Sources

364 1K 4K 845K 244

7 428 4K 81K 168

View Details

Harsh @itzHarsh08

3 days ago

Bruh 😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭😭

0 0 3 26 0

View Details

Watcher.Guru @WatcherGuru

5 days ago

JUST IN: Jim Cramer says Michael Saylor "murdered Bitcoin."

920 509 6K 691K 192

0 0 0 20 0

View Details

Elon Musk @elonmusk

8 years ago

@McMurchie @alexmd2 @Techmeme @CatherineShu At this point, I want ETH even if it is a scam

1K 2K 9K 0 384

View Details

r @bekhayalime

5 days ago

This is not suicide, this is murder. This scene from 3 Idiots is profoundly important and remains deeply relevant today.

NDTV @ndtv

6 days ago

'Don't Have Courage To Take Exam Again': NEET Aspirant Dies By Suicide ndtv.com/india-news/nee…

1K 4K 11K 3.5M 1K

37 3K 19K 317K 891

View Details

Shikamaru ☕ @pranavcodes_

5 days ago

"Learn Machine Learning, the money is crazy" Me after spending 1 hour on Linear Algebra:

30 110 1K 39K 218

View Details

BOOM X @CryptoBoomNews

4 days ago

31 108 904 111K 107

View Details

Yami @YamiPNL

5 days ago

🚨 BREAKING: LILY PHILLIPS ACCUSES MICHAEL SAYLOR OF STEALING HER IDEA OF FUCKING 100'000 MEN IN A DAY

856 2K 37K 2.2M 2K

View Details

Watcher.Guru @WatcherGuru

5 days ago

JUST IN: Bitcoin is now down 32% this year. $ETH is down 45% this year

406 590 5K 259K 105

0 0 1 28 0

View Details

Harsh @itzHarsh08

7 days ago

Friendship ended with Crypto Now DSA is my best friend.

0 0 1 36 0

View Details

recently made a project on solana. perps matching & liquidation engines are not fully on-chain. most of the times orderbooks are offline ( dydx ) or uses critbitbook (openbook) but some of the most important operations : - like how many orders need to be liquidated upon price change? - how many levels needed to fill a big order? are left O(n) i provide a sample implementation of those queries in O(logn), completely on-chain. benchmark: a matching order which took more than 200k CU executed in less than 8k CU ! tech behind it: i used a fenwick tree instead of critbitbook which inherently stores sum for O(logn) queries in just 2x size. had to do multiple workarounds like zero-copy implementation, scaling down all u128 to u64, etc to match solanas limits. if you like it you go read more on my github and can star or follow me: github.com/tirth2004/logn… made it for @SuperteamIN's fellowship. CU Maxxing for life :)