SilverKey Monitor

Pgvectorscale: Cost-Efficient Scaling for High Volume Vector Workloads

published on 2024/06/12

Pgvectorscale is a PostgreSQL extension that builds on pgvector for more performance and scale. pgvectorscale enables developers to build more scalable AI applications with higher-performance embedding search and cost-efficient storage.

The initial release includes two key innovations: (1) StreamingDiskANN, a high-performance, cost-efficient vector search index for pgvector data inspired by research at Microsoft, and (2) Statistical Binary Quantization (SBQ), developed by Timescale’s own researchers to improve upon standard binary quantization techniques.


This is fantastic.

Introducing Apple’s On-Device and Server Foundation Models

published on 2024/06/11

Our foundation models are trained on Apple's AXLearn framework, an open-source project we released in 2023. It builds on top of JAX and XLA, and allows us to train the models with high efficiency and scalability on various training hardware and cloud platforms, including TPUs and both cloud and on-premise GPUs. We used a combination of data parallelism, tensor parallelism, sequence parallelism, and Fully Sharded Data Parallel (FSDP) to scale training along multiple dimensions such as data, model, and sequence length.

MachineLearning - Apple

Exploring Gleam, a type-safe language on the BEAM!

published on 2024/06/11

The thing that always kind of put me off regarding Elixir was the dynamic typing. I’m a static typing person; it solves a LOT of common issues in my day-to-day job and provides cool benefits like:

  • Less runtime errors, since the compiler YELLS at you when something is wrong about your code structure.
  • Trust in what goes to production, since it was approved by the boss (the compiler).
  • Code with typing indications is often (not always, but still, often) clearer and easier to understand. If the documentation is incomplete or non-existent, you still have something to go by, and new engineers can start contributing faster.
  • Coding by thinking about the data structure FIRST, and then, the implementation matches my thinking process.


RazorSlices, a Razor-based templates for ASP.NET Core

published on 2024/06/11

Lightweight Razor-based templates for ASP.NET Core without MVC, Razor Pages, or Blazor, optimized for high-performance, unbuffered rendering with low allocations. Compatible with trimming and native AOT. Great for returning dynamically rendered HTML from Minimal APIs, middleware, etc. Supports .NET 8+


Study finds 268% higher failure rates for Agile software projects

published on 2024/06/06

A study has found that software projects adopting Agile practices are 268 percent more likely to fail than those that do not.

Even though the research commissioned by consultancy Engprax could be seen as a thinly veiled plug for Impact Engineering methodology, it feeds into the suspicion that the Agile Manifesto might not be all it's cracked up to be.

The study's fieldwork was conducted between May 3 and May 7 with 600 software engineers (250 in the UK and 350 in the US) participating. One standout statistic was that projects with clear requirements documented before development started were 97 percent more likely to succeed. In comparison, one of the four pillars of the Agile Manifesto is "Working Software over Comprehensive Documentation."

The Register

There is no silver bullet.