GoKawiil - Tech News

91.

Discovering hard disk physical geometry through microbenchmarking (2019) (news.ycombinator.com)

2026-04-30 | by Leela Kamalesh | get SSD Diagnostic Tool Kit → | tags: hard disk, microbenchmarking, disk geometry

92.

Show HN: A new benchmark for testing LLMs for deterministic outputs (news.ycombinator.com)

2026-04-29 | get LLM Deterministic Output Benchmark → | tags: structured output benchmark, llms, json pass

93.

A Decade of AMD Ryzen: 10 Years of CPUs Tested (techspot.com)

2026-04-29 | get AMD Ryzen Processor → | tags: amd ryzen, cpu, benchmarking

94.

A Decade of AMD Ryzen: 10 Years of CPUs Tested (techspot.com)

2026-04-29 | get AMD Ryzen 7 Processor → | tags: amd ryzen, cpu, benchmarking

95.

A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all (news.ycombinator.com)

2026-04-28 | get AI Model Documentation Guide → | tags: agents.md, code generation, auggiebench

96.

Show HN: OSS Agent I built topped the TerminalBench on Gemini-3-flash-preview (news.ycombinator.com)

2026-04-27 | get TerminalBench Performance Monitor → | tags: dirac, gemini-3-flash-preview, terminalbench

97.

Startup unveils benchtop metal 3D printer that brings industrial tech below $10,000 (techspot.com)

2026-04-27 | get Creality Ender 3 V2 → | tags: metal 3d printer, benchtop, laser melting

98.

The predictable failure of the QDay Prize (news.ycombinator.com)

2026-04-26 | by Craig Gidney | tags: quantum cryptanalysis, shor’s algorithm, quantum error correction

99.

SWE-bench Verified no longer measures frontier coding capabilities (news.ycombinator.com)

2026-04-26 | get Coding Challenge Platform Subscription → | tags: swe-bench, verified, frontier coding

100.

Why SWE-bench Verified no longer measures frontier coding capabilities (news.ycombinator.com)

2026-04-26 | tags: swe-bench, verified, frontier coding

101.

Mowing Down Simulated Elephants Could Help Self-Driving Cars Prepare For the Chaos of Real Life Streets (futurism.com)

2026-04-26 | by Frank Landymore | get Self-Driving Car Simulation Kit → | tags: autonomous vehicles, university of tübingen, simon elon musk

102.

Lambda Calculus Benchmark for AI (news.ycombinator.com)

2026-04-25 | get Lambda Calculus Puzzle Game → | tags: lambda calculus, benchmark, computation

103.

Linux 7.1 Removes Drivers for Bus Mouse Support (news.ycombinator.com)

2026-04-24 | get Linux Kernel USB Mouse Adapter → | tags: linux 7.1, bus mouse, phoronix

104.

OpenAI's GPT-5.5 is here, and it's no potato: narrowly beats Anthropic's Claude Mythos Preview on Terminal-Bench 2.0 (venturebeat.com)

2026-04-23 | get OpenAI GPT-5.5 Mug → | tags: openai, gpt-5.5, anthropic

105.

AMD Ryzen 9 9950X3D2 review: More cache, more cash (tomshardware.com)

2026-04-21 | by Jake Roach | tags: amd ryzen 9 9950x3d2, amd dual-3d v-cache, am5 platform

106.

Kimi vendor verifier – verify accuracy of inference providers (news.ycombinator.com)

2026-04-20 | get Kimi Vendor Verifier Tool → | tags: kimi vendor verifier, inference providers, open-source models

107.

Arc Prize Foundation (YC W26) Is Hiring a Platform Engineer for ARC-AGI-4 (news.ycombinator.com)

2026-04-17 | get Arc AGI Development Kit → | tags: arc-prize foundation, arc-agi-4, platform engineer

108.

Experience vs specs: Our readers have spoken, and benchmarks aren’t everything (androidauthority.com)

2026-04-17 | get Tech News Notebook → | tags: google pixel, apple macbook neo, android authority

109.

Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7 (news.ycombinator.com)

2026-04-16 | by Simon Willison | get Pelican 3D Printer Kit → | tags: qwen3.6-35b-a3b, claude opus 4.7, pelican benchmark

110.

Databricks tested a stronger model against its multi-step agent on hybrid queries. The stronger model still lost by 21%. (venturebeat.com)

2026-04-14 | get AI Model Performance Analyzer → | tags: databricks, rag, karlbench

111.

Databricks research shows multi-step agents consistently outperform single-turn RAG when answers span databases and documents (venturebeat.com)

2026-04-14 | get AI Research Assistant Kit → | tags: databricks, rag, multi-step agents

112.

N-Day-Bench – Can LLMs find real vulnerabilities in real codebases? (news.ycombinator.com)

2026-04-13 | get Cybersecurity Coding Challenge Kit → | tags: n-day-bench, large language models, vulnerability discovery

113.

Android now stops you sharing your location in photos (news.ycombinator.com)

2026-04-13 | by Terence Eden | get Photo Location Privacy Protector → | tags: android, google, geolocation

114.

Why we spent 50+ hours retesting Intel’s Core Ultra 270K Plus and 250K Plus (tomshardware.com)

2026-04-13 | by Jake Roach | get Intel Core Ultra Processor Kit → | tags: intel, core ultra 270k plus, arrow lake refresh

115.

Exploiting the most prominent AI agent benchmarks (news.ycombinator.com)

2026-04-11 | get AI Benchmark Test Kit → | tags: ai agent benchmarks, uc berkeley, sweet-bench

116.

How We Broke Top AI Agent Benchmarks: And What Comes Next (news.ycombinator.com)

2026-04-11 | get AI Benchmark Testing Kit → | tags: ai benchmarks, uc berkeley, trustworthiness

117.

AI models are terrible at betting on soccer—especially xAI Grok (arstechnica.com)

2026-04-11 | tags: xai grok, google gemini, anthropic claude

118.

Nubia defends the ethics of REDMAGIC 11 Pro benchmark manipulation (androidauthority.com)

2026-04-10 | get RedMagic 11 Pro Gaming Phone → | tags: nubia, redmagic 11 pro, benchmark manipulation

119.

Geekbench 6.7 adds Intel BOT detection to spoof out 'unrealistic' CPU scores — Benchmark runs with BOT enabled will be marked as invalid (tomshardware.com)

2026-04-09 | by Hassam Nasir | get Geekbench 6.7 Software → | tags: geekbench, intel, binary optimization tool

120.

Astropad unveils Workbench for Mac: ‘Remote desktop made for the AI era’ (9to5mac.com)

2026-04-08 | by Zac Hall | get Astropad Workbench Software → | tags: astropad, workbench, luna display