151
Can GROUP BY still be optimized after all these years?
An optimization that made Pivot significantly faster on GROUP BY queries with a Top-K operator above them, and a walkthrough of how analytical databases run GROUP BY in the first place.
152
Can LLMs model real-world systems in TLA+?
Editors’ note: AI has been actively pushing the frontier of applied formal methods for computing systems. In this article, the Specula team wrote about their experience of evaluating LLMs on modeling system code, the basic capability for agentic model checking, using TLA+, a specification language f
153
Lakebase architecture delivers faster Postgres writes
How Lakebase's storage architecture pushes down full page write generation to cut WAL traffic by 94% and boost Postgres write throughput by up to 5x.
154
Rust Threads on the GPU
At VectorWare , we are building the first GPU-native software company . Today, we are excited to announce that we can successfully use Rust's std::thread on the GPU. This milestone marks a significant step towards our vision of enabling developers to write complex, high-performance applications that leverage the full…
155
How Uber Protects Against Retry Storms
What if retries could know when they’re helping, and when they’re making things worse? Uber Engineering built error ownership to prevent and contain retry storms.
156
Compute-efficient pretraining and scaling to trillion-parameter models
Research update on compute-efficient pretraining and scaling to trillion-parameter models.
157
Hunting Down a Go Runtime Bug on 32-Bit Embedded Systems
A customer's Go application kept crashing on a 32-bit embedded Linux system. We traced the fatal netpoll error to a pointer aliasing bug in the Go runtime.
158
Keeping a Postgres Queue Healthy
Dead tuples from high-churn job queues can silently degrade your Postgres database when vacuum falls behind—especially alongside competing workloads. Traffic Control keeps cleanup on track.
159
Jepsen: MariaDB Galera Cluster 12.1.2
MariaDB is a popular open source SQL database which began as a fork of MySQL . MariaDB Galera Cluster is an active-active replication system for MariaDB which allows reads and writes on every node. In 2015 this author analyzed MariaDB with Galera Cluster and found that although Galera claimed to offer Snapshot…
160
Inference Optimization for MiMo v2.5: Pushing Hybrid SWA Efficiency to the Limit
An end-to-end engineering practice for the MiMo-V2.5 series inference system, covering KVCache management, tiered caching, SWA-aware prefix cache trees, scheduling, Prefill/Decode pipelines, and multimodal optimizations.
161
Software sandboxing: The basics (2025)
Diving into the territory of software sandboxing is diving into mostly uncharted territory. The necessary pieces to implement good sandboxing in your software are scattered all-around and the pioneers haven’t yet gathered enough knowledge into an unified mappa mundi that can guide new sailors through some well…
162
What happens when a GPU reads memory
Following an LDG.E SASS instruction through the hardware units in an RTX 4090.
163
Building Your Own Efficient uint128 in C++
A practical walk-through of a fixed-width uint128 implementation in modern C++.
164
Topcoat is pushing the boundary of server applications with Rust
Tokio is a runtime for writing reliable asynchronous applications with Rust. It provides async I/O, networking, scheduling, timers, and more.
165
166
Async Rust vs RTOS showdown (2022)
It's time for another technical blog post about async Rust on embedded. This time we're going to pitch Embassy/Rust against FreeRTOS/C on an STM32F446 microcontroller.
167
AVX2 is slower than SSE2-4.x under Windows ARM emulation
If you compile your app for AVX2 and it runs on Windows ARM under Prism emulation, is it faster or slower than compiling for SSE2-4.x? I assumed it would be roughly the same — maybe slightly slower due to emulation overhead, but AVX2's wider operations would compensate. The
168
Process-Based Concurrency: Why Beam and OTP Keep Being Right
What makes BEAM and OTP different from async/await, threads, and actors. Process isolation, supervisors, and why Erlang's model keeps winning 40 years later.
169
A 32-year-old bug walks into a Telnet server
A long, long time ago, in a land free of binary exploit mitigations, when Unix still roamed the Earth, there lived a pre-authentication Telnetd vulnerability. In fact, this vulnerability was born so long ago (way back in 1994) that it may even be older than you. To put the
170
Making ast.walk 220x Faster
Why ast.walk when you can ast.sprint? How we made Python AST traversal 220x faster while building a custom linter for Reflex's AI code generation.
171
Don't use musl if you care about performance
I've spent most of my career working inside the JVM, with a sidebar of Python. So the first time I hit incompatible libc with containerized Bifrost, and GPT suggested musl as the solution, not only did I jump on it, I moved my other Rust projects to musl as
172
Postgres SELECT DISTINCT Does Not Scale
Postgres SELECT DISTINCT performs surprisingly poorly. We explain why and how we mitigated it.
173
174
The Road to the WASM Component Model 1.0
WASI P3 is almost here, bringing native async support to the WebAssembly System Interface (WASI) and Component Model. In this post, we’re looking to the next big milestone: a stable, formally specified Component Model 1.0.At February’s B...
175
How Cloudflare responded to the “Copy Fail” Linux vulnerability
When a critical Linux kernel privilege escalation was publicly disclosed, Cloudflare's security and engineering teams detected, investigated, and mitigated the threat across our global fleet, confirming zero customer impact and no malicious exploitation.
176
118M Queries per Second on Neki
We ran a massive, sharded Postgres database at 118.5 million queries per second, with 200k queries per second on each shard across 512 shards.
177
Data-only attacks are easier than you think (2024)
Data-only attacks, those that do not affect a program’s control flow, have long been considered too sophisticated and niche to pose a practical threat. With our research, however, we have built a tool that automatically generates them with surprising ease. We explain how such attacks work, and why our tool, Einstein…
178
