HackerNews Summary (2023-05-05 02PM)

Source: Runtime Engineer (Zig or C/C++)

Summary
Oven, the company behind Bun, is looking for experienced engineers in Zig or C/C++ to contribute to building a fast and efficient JavaScript runtime. The company offers competitive compensation, remote work options, healthcare benefits, and a supportive workplace environment.

Fact

👨‍💻 Oven is looking for engineers experienced in Zig or C/C++ to help develop Bun, the all-in-one JavaScript runtime, bundler, and package manager.

📈 Oven has raised $7M from top Silicon Valley investors, including YCombinator and Kleiner Perkins, and plans to offer a serverless platform to make deploying applications easier.

🌎 The company has an office in downtown San Francisco but offers remote work options for candidates who cannot relocate.

💼 Oven offers competitive compensation, healthcare benefits, and four weeks of paid time off, along with stipends for commuting, fitness routine, office setup, and computer hardware.

---

Source: A kernel without buffer heads [LWN.net]

Summary
Buffer heads, an ancient part of the kernel’s block I/O subsystem, have been identified as a target for removal, but they have persisted despite causing problems. Christoph Hellwig’s patch series enables the building of a kernel without buffer heads, but removing them at this point would be costly.

Fact
💾 The buffer_head structure was part of the kernel’s block I/O subsystem from the beginning of Linux, holding copies of blocks found on the system’s disks in its cache, marked with its unique identification.

---

Source: eBay’s Blazingly Fast Billion-Scale Vector Similarity Engine

Summary
eBay CoreAI team launched an “Approximate Nearest Neighbor” (ANN) vector similarity engine that provides tooling to match semantically similar items and personalize recommendations. This engine uses an ANN algorithm called Hierarchical Navigable Small World Graphs (HNSW) and Scalable Nearest Neighbors (ScaNN) to speed up the search, making it lightning-fast and massively scalable.

Facts

💡 eBay’s Similarity Engine is a billion-scale vector similarity matching engine, deployed on eBay’s private cloud.

💡 It can handle thousands of requests per second and returns similarity results in less than 25 milliseconds for 95% of the responses.

💡 The engine supports attribute-based filtering, allowing the user to specify attributes such as item size, color, and brand.

💡 ANN techniques speed up the search by preprocessing the data into an efficient index and use data structure like inverted files, trees and neighborhood graphs.

💡 The similarity engine uses a Hierarchical Navigable Small World Graphs (HNSW) and Scalable Nearest Neighbors (ScaNN) algorithm to provide lightning-fast and massively scalable nearest neighbor search.

---

Source: At Musk’s brain-chip startup, animal-testing panel is rife with potential conflicts

Summary
Elon Musk’s brain implant firm, Neuralink, has appointed insiders to its animal-research oversight board who may financially benefit as the firm reaches development goals, according to company documents reviewed by Reuters. The membership of the 22-person panel, which features 19 Neuralink employees, has raised questions about conflict of interest violations, say animal-research and bioethics experts. Its makeup was unveiled in the wake of federal investigations into Neuralink’s animal-welfare and improper pathogen transportation claims. Musk and his firm did not respond to Reuters’ requests for comment.

Fact

🐒 Neuralink has filled its animal-research oversight board with insiders who potentially stand to financially benefit as the firm reaches development goals.

---