ambesh

Just need to show up daily.

[ambesh @0xAmbesh]

@futarddotio @ordrtrade W @ordrtrade .

[ambesh @0xAmbesh]

@thsottiaux Thanks @thsottiaux 😌

[ambesh @0xAmbesh]

@varadrane17 @sama Anthropic is it then

[ambesh @0xAmbesh]

@sama pls fix need to fill ITR man .

[ambesh @0xAmbesh]

We suffer more often in imagination than in reality ~ Seneca

[ambesh @0xAmbesh]

I had just built a disk manager that does concurrency control the blocking way: coarse latch over the file, per-page shared_mutex, WritePage overwriting in place.
your whole arc is the escape from that. single mutex can't guard a read-modify-write → version everything, never overwrite, just append. nearly the exact opposite move. great read.

[ambesh @0xAmbesh]

@ordrtrade @colosseum @solana @SuperteamIN @superteam Congratulations Team 🔥

[ambesh @0xAmbesh]

Was going through the Iceberg Summit 2026 recorded talks and hit a great one:
how Apache Iceberg moved from v2 merge-on-read deletes → v3 deletion vectors.
https://t.co/g5s19Degex

>In v2, deleting a row meant writing side "delete files" (positions or matching values) that the engine had to merge back in on every read and they pile up into lots of small files that slow reads down.

>v3 swaps that for a deletion vector: one compact Roaring bitmap per data file, bit P set = row P deleted.

That sent me down a rabbit hole into the paper behind it — "Better bitmap performance with Roaring bitmaps" (S. Chambi, D. Lemire, O. Kaser, R. Godin). Here's my deep dive 👇
https://t.co/o8z6ZEhExm

[ambesh @0xAmbesh]

https://t.co/yXk2xNpwwS

[ambesh @0xAmbesh]

https://t.co/yXk2xNpwwS

[ambesh @0xAmbesh]

@ThePrimeagen The idea goes by where every frontend can be replaced by a MCP

[ambesh @0xAmbesh]

Was deep diving on how data gets stored on disk in various DBMS and landed on this 24-year-old paper. So good I had to write an article on it .
https://t.co/eorG8aOBBN

[ambesh @0xAmbesh]

https://t.co/wmIhV0nEML

[ambesh @0xAmbesh]

https://t.co/wmIhV0nEML

[ambesh @0xAmbesh]

@SakanaAILabs How are the responses on chat interface so fast 😳.

[ambesh @0xAmbesh]

@Hi_Mrinal is it better than Database Internals ?

[ambesh @0xAmbesh]

@piyushgarg_dev https://t.co/5jockINmmj

[ambesh @0xAmbesh]

Big moves already
https://t.co/oKfIFra6Az

[SpaceX @SpaceX]

SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world’s most useful AI models.

For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.

We look forward to working closely with the Cursor team to advance our frontier AI capabilities

[ambesh @0xAmbesh]

IBM sat on the idea that changed databases forever
>Codd , an IBM Employee published the relational model in 1970.
>IBM hated it , they'd already sold IMS, their navigational DB, to customers. So they buried his own paper internally.
>Oracle (then a 3-person startup) read it and shipped SQL before IBM did.
>IBM's own researcher invented the thing that killed their product.

[ambesh @0xAmbesh]

Pls wait for sometime need to finish my project 🙃.
https://t.co/PbYsfYCr5x

[ClaudeDevs @ClaudeDevs]

As a result of a US government directive, we are suspending access to Claude Fable 5 for all users. You can continue to use all other Claude models.

Here’s what this means for you:

Across Claude products, new sessions will run on your selected default model or Opus 4.8, and existing Fable 5 sessions will end with an error.

On the Claude Platform, requests to Fable 5 will also return an error. Please update your integrations to other Claude models.

We know this is a disruption to your workflows; we appreciate your patience and support.

[ambesh @0xAmbesh]

@BenjDicken Here’s how Open AI cut down their CPU utilisation by 40% by this small bloom filter technique in click house

[ambesh @0xAmbesh]

Just one change in ClickHouse Bloom filters: replace division with multiplication for position calc.
Boom 40% less CPU on OpenAI's petabyte logs.Simple tweaks, massive wins. 🤯https://t.co/dlz0ooCny1

[ambesh @0xAmbesh]

@BenjDicken Bloom filter are great but deletion is tricky wrote a thread about how cuckoo filter solves this .

[ambesh @0xAmbesh]

Bloom Filter has one major drawback
❌ Deletion is very tricky
You can add items easily,
but removing them risks breaking other entries.

So what do systems use when deletes are required?
🧵 Enter Cuckoo Filter ↓

https://t.co/GlgFMLyzWS