CoFeed

Meta

Sequence learning: A paradigm shift for personalized ads recommendations

AI plays a fundamental role in creating valuable connections between people and advertisers within Meta’s family of apps.

Meta

How Meta built large-scale cryptographic monitoring

Cryptographic monitoring at scale has been instrumental in helping our engineers understand how cryptography is used at Meta.

Meta

Diff Authoring Time: Measuring developer productivity at Meta

At Meta, we’re always looking for ways to enhance the productivity of our engineers and developers.

Meta

IPLS: Privacy-preserving storage for your WhatsApp contacts

Your contact list is fundamental to the experiences you love and enjoy on WhatsApp.

Meta

OCP Summit 2024: The open future of networking hardware for AI

At Open Compute Project Summit (OCP) 2024, we’re sharing details about our next-generation network fabric for our AI training clusters.

Meta

Meta’s open AI hardware vision

At the Open Compute Project (OCP) Global Summit 2024, we’re showcasing our latest open AI hardware designs with the OCP community.

Meta

How open source AI can improve population estimates, sustainable energy, and the delivery of climate change interventions

Data for Good at Meta is open-sourcing the data used to train our AI-powered population maps.

Meta

React at Meta Connect 2024

At Meta, React and React Native are more than just tools; they are integral to our product development and innovation.

Meta

Inside Bento: Jupyter Notebooks at Meta

This episode of the Meta Tech Podcast is all about Bento, Meta’s internal distribution of Jupyter Notebooks, an open-source web-based comput...

Meta

Simulator-based reinforcement learning for data center cooling optimization

We’re sharing more about the role that reinforcement learning plays in helping us optimize our data centers’ environmental controls.

Meta

Read Meta’s 2024 Sustainability Report

[.

Meta

Meta is getting ready for post-quantum cryptography

The Quantum Apocalypse is coming.

Meta

How Meta enforces purpose limitation via Privacy Aware Infrastructure at scale

At Meta, we’ve been diligently working to incorporate privacy into different systems of our software stack over the past few years.

Meta

RETINAS: Real-Time Infrastructure Accounting for Sustainability

We are introducing a new metric— real-time server fleet utilization effectiveness —as part of the RETINAS initiative to help reduce emission...

Meta

How PyTorch powers AI training and inference

Learn about new PyTorch advancements for LLMs and how PyTorch is enhancing every aspect of the LLM lifecycle.

Meta

Inside the hardware and co-design of MTIA

In this talk from AI Infra @ Scale 2024, Joel Colburn, a software engineer at Meta, technical lead Junqiang Lan, and software engineer Jack ...

Meta

Bringing Llama 3 to life

Llama 3 is Meta’s most capable openly-available LLM to date and the recently-released Llama 3.

Meta

Aparna Ramani discusses the future of AI infrastructure

Delivering new AI technologies at scale also means rethinking every layer of our infrastructure – from silicon and software systems and even...

Meta

How Meta animates AI-generated images at scale

We launched Meta AI with the goal of giving people new ways to be more productive and unlock their creativity with generative AI (GenAI).

Meta

RoCE networks for distributed AI training at scale

AI networks play an important role in interconnecting tens of thousands of GPUs together, forming the foundational infrastructure for traini...

Meta

DCPerf: An open source benchmark suite for hyperscale compute applications

We are open-sourcing DCPerf, a collection of benchmarks that represents the diverse categories of workloads that run in data center cloud de...

Meta

Meet Caddy – Meta’s next-gen mixed reality CAD software

What happens when a team of mechanical engineers get tired of looking at flat images of 3D models over Zoom? Meet the team behind Caddy, a n...

Meta

AI Lab: The secrets to keeping machine learning engineers moving fast

The key to developer velocity across AI lies in minimizing time to first batch (TTFB) for machine learning (ML) engineers.

Meta

Taming the tail utilization of ads inference at Meta scale

Tail utilization is a significant system issue and a major factor in overload-related failures and low compute utilization.

Meta

Meta’s approach to machine learning prediction robustness

Meta’s advertising business leverages large-scale machine learning (ML) recommendation models that power millions of ads recommendations per...

Meta

The key to a happy Rust/C++ relationship

The history of Rust at Meta goes all the way back to 2016, when we first started using it for source control.

Meta

Leveraging AI for efficient incident response

We’re sharing how we streamline system reliability investigations using a new AI-assisted root cause analysis system.

Meta

PVF: A novel metric for understanding AI systems’ vulnerability against SDCs in model parameters

We’re introducing parameter vulnerability factor (PVF), a novel metric for understanding and measuring AI systems’ vulnerability against sil...

Meta

MLow: Meta’s low bitrate audio codec

At Meta, we support real-time communication (RTC) for billions of people through our apps, including WhatsApp, Instagram, and Messenger.

Meta

How Meta trains large language models at scale

As we continue to focus our AI research and development on solving increasingly complex problems, one of the most significant and challengin...

Meta

Maintaining large-scale AI capacity at Meta

Meta is currently operating many data centers with GPU training clusters across the world.

Meta

Unlocking the power of mixed reality devices with MobileConfig

MobileConfig enables developers to centrally manage a mobile app’s configuration parameters in our data centers.

Meta

Serverless Jupyter Notebooks at Meta

At Meta, Bento, our internal Jupyter notebooks platform, is a popular tool that allows our engineers to mix code, text, and multimedia in a ...

Meta

Composable data management at Meta

In recent years, Meta’s data management systems have evolved into a composable architecture that creates interoperability, promotes reusabil...

Meta

Post-quantum readiness for TLS at Meta

Today, the internet (like most digital infrastructure in general) relies heavily on the security offered by public-key cryptosystems such as...

Meta

Behind the scenes of Threads for web

When Threads first launched one of the top feature requests was for a web client.

Meta

Building new custom silicon for Meta’s AI workloads

[.

Meta

Building new custom silicon for Meta’s AI workloads

[.

Meta

Building new custom silicon for Meta’s AI workloads

[.

Meta

Building an infrastructure for AI’s future

[.

Meta

Building an infrastructure for AI’s future

[.

Meta

Introducing the next-gen Meta Training and Inference Accelerator

[.

Meta

Bringing HDR photo support to Instagram and Threads

Meta’s family of apps serves trillions of image download requests every day.

Meta

Threads has entered the fediverse

Threads has entered the fediverse! As part of our beta experience, now available in a few countries, Threads users aged 18+ with public prof...

Meta

Optimizing RTC bandwidth estimation with machine learning

Bandwidth estimation (BWE) and congestion control play an important role in delivering high-quality real-time communication (RTC) across Met...

Meta

Better video for mobile RTC with AV1 and HD

At Meta, we support real-time communication (RTC) for billions of people through our apps, including Messenger, Instagram, and WhatsApp.

Meta

Logarithm: A logging engine for AI training workflows and services

Systems and application logs play a key role in operations, observability, and debugging workflows at Meta.

Meta

Building Meta’s GenAI Infrastructure

Marking a major investment in Meta’s AI future, we are announcing two 24k GPU clusters.

Meta

Making messaging interoperability with third parties safe for users in Europe

To comply with a new EU law, the Digital Markets Act (DMA), which comes into force on March 7th, we’ve made major changes to WhatsApp and Me...

Meta

How DotSlash makes executable deployment simpler

Andres Suarez and Michael Bolin, two software engineers at Meta, join Pascal Hartig (@passy) on the Meta Tech Podcast to discuss the ins and...

Meta

Aligning Velox and Apache Arrow: Towards composable data management

We’ve partnered with Voltron Data and the Arrow community to align and converge Apache Arrow with Velox, Meta’s open source execution engine...

Meta

Meta loves Python

By now you’re already aware that Python 3.

Meta

Simple Precision Time Protocol at Meta

While deploying Precision Time Protocol (PTP) at Meta, we’ve developed a simplified version of the protocol (Simple Precision Time Protocol ...

Meta

DotSlash: Simplified executable deployment

We’ve open sourced DotSlash, a tool that makes large executables available in source control with a negligible impact on repository size, th...

Meta

Improving machine learning iteration speed with faster application build and packaging

Slow build times and inefficiencies in packaging and distributing execution files were costing our ML/AI engineers a significant amount of t...

Meta

Lazy is the new fast: How Lazy Imports and Cinder accelerate machine learning at Meta

At Meta, the quest for faster model training has yielded an exciting milestone: the adoption of Lazy Imports and the Python Cinder runtime.

Meta

How Meta is advancing GenAI

What’s going on with generative AI (GenAI) at Meta? And what does the future have in store? In this episode of the Meta Tech Podcast, Meta e...

Meta

How Meta built the infrastructure for Threads

On July 5, 2023, Meta launched Threads, the newest product in our family of apps, to an unprecedented success that saw it garner over 100 mi...

Meta

AI debugging at Meta with HawkEye

HawkEye is the powerful toolkit used internally at Meta for monitoring, observability, and debuggability of the end-to-end machine learning ...

Meta

Building end-to-end security for Messenger

We are beginning to upgrade people’s personal conversations on Messenger to use end-to-end encryption (E2EE) by default.

Meta

Writing and linting Python at scale

Python plays a big part at Meta.

Meta

Watch: Meta’s engineers on building network infrastructure for AI

Meta is building for the future of AI at every level – from hardware like MTIA v1, Meta’s first-generation AI inference accelerator to publi...

Meta

Enhancing the security of WhatsApp calls

New optional features in WhatsApp have helped make calling on WhatsApp more secure.

Meta

How Meta built Threads in 5 months

In about five short months, a small team of engineers at Meta took Threads, the new text-based conversations app, from from an idea to the m...

Meta

Automating data removal

Meta’s Systematic Code and Asset Removal Framework (SCARF) has a subsystem for identifying and removing unused data types.

Meta

Automating dead code cleanup

Meta’s Systematic Code and Asset Removal Framework (SCARF) has a subsystem for identifying and removing dead code.

Meta

5 Things you didn’t know about Buck2

Meta has a very large monorepo, with many  different programming languages.

Meta

How Meta is creating custom silicon for AI

Olivia Wu, Meta’s Technical Lead for Infra Silicon, discusses the design and development of Meta’s first-generation AI inference accelerator...

Meta

Automating product deprecation

Systematic Code and Asset Removal Framework (SCARF) is Meta’s unused code and data deletion framework.

Meta

Meta contributes new features to Python 3.12

Python 3.

Meta

Meta Quest 2: Defense through offense

Meta’s Native Assurance team regularly performs manual code reviews as part of our ongoing commitment to improve the security posture of Met...

Meta

Using Chakra execution traces for benchmarking and network performance optimization

Meta presents Chakra execution traces, an open graph-based representation of AI/ML workload execution, laying the foundation for benchmarkin...

Meta

Arcadia: An end-to-end AI system performance simulator

We’re introducing Arcadia, Meta’s unified system that simulates the compute, memory, and network performance of AI training clusters.

Meta

Threads: The inside story of Meta’s newest social app

Earlier this year, a small team of engineers at Meta started working on an idea for a new app.

Meta

What’s it like to write code at Meta?

Ever wonder what it’s like to write code at Meta’s scale? On the latest episode of the Meta Tech Podcast, Meta engineer Pascal Hartig (@pass...

Meta

Scheduling Jupyter Notebooks at Meta

At Meta, Bento is our internal Jupyter notebooks platform that is leveraged by many internal users.

Meta

Code Llama: Meta’s state-of-the-art LLM for coding

[.

Meta

Introducing Immortal Objects for Python

Instagram has introduced Immortal Objects – PEP-683 – to Python.

Meta

Meta Connect 2023: September 27 – 28

[.

Meta

Scaling the Instagram Explore recommendations system

Explore is one of the largest recommendation systems on Instagram.

Meta

How Meta is improving password security and preserving privacy

Meta is developing new privacy-enhancing technologies (PETs) to innovate and solve problems with less data.

Meta

Fixit 2: Meta’s next-generation auto-fixing linter

Fixit is dead! Long live Fixit 2 – the latest version of our open-source auto-fixing linter.

Meta

Using short-lived certificates to protect TLS secrets

Short-lived certificates (SLCs) are part of our latest efforts to further secure our Transport Layer Security (TLS) private keys on our edge...

Meta

Bringing HDR video to Reels

Meta has made it possible for people to upload high dynamic range (HDR) videos from their phone’s camera roll to Reels on Facebook and Insta...

Meta

Meta’s Evenstar is transitioning to OCP to accelerate open RAN adoption

Meta is transferring its IP for Evenstar, a program to accelerate the adoption of open RAN technologies, to the Open Compute Project (OCP).

Meta

Meta developer tools: Working at scale

Every day, thousands of developers at Meta are working in repositories with millions of files.

Meta

Bombyx is being licensed for product development

When we first conceived of our aerial fiber deployment solution, Bombyx (the Latin name for a silk moth), we imagined a robot weaving strand...

Meta

MSVP is Meta’s first video processing ASIC

[.

Meta

Meta introduces its first-generation AI inference accelerator

[.

Meta

Building and deploying MySQL Raft at Meta

We’re rolling out MySQL Raft with the aim to eventually replace our current MySQL semisynchronous databases.

Meta

The malware threat landscape: NodeStealer, DuckTail, and more

We’re sharing our latest threat research and technical analysis into persistent malware campaigns targeting businesses across the internet, ...

Meta

A fine-grained network traffic analysis with Millisampler

What the research is:  Millisampler is one of Meta’s latest characterization tools and allows us to observe, characterize, and debug network...

Meta

Deploying key transparency at WhatsApp

WhatsApp has launched a new cryptographic security feature to automatically verify a secured connection based on key transparency.

Meta

How Device Verification protects your WhatsApp account

WhatsApp has launched a new security feature that further helps prevent attackers from using vectors like on-device malware.

Meta

Why xHE-AAC is being embraced at Meta

We’re sharing how Meta delivers high-quality audio at scale with the xHE-AAC audio codec.