Join us

ContentUpdates and recent posts about vLLM..
 Activity
@derynleigh started using tool OWASP Dependency-Check , 2 weeks, 5 days ago.
 Activity
@derynleigh started using tool Amazon S3 , 2 weeks, 5 days ago.
 Activity
@derynleigh started using tool Amazon ELB , 2 weeks, 5 days ago.
 Activity
@derynleigh started using tool Amazon EC2 , 2 weeks, 5 days ago.
 Activity
@derynleigh started using tool Amazon Cloudfront , 2 weeks, 5 days ago.
 Activity
@derynleigh started using tool Akamai , 2 weeks, 5 days ago.
Story
@laura_garcia shared a post, 2 weeks, 6 days ago
Software Developer, RELIANOID

🔐 RELIANOID & NIST Cybersecurity Framework Alignment

At RELIANOID, security is built into both our Load Balancer and our internal operations. We align our product and organizational practices with the NIST Cybersecurity Framework (CSF) across its five core functions: Identify, Protect, Detect, Respond, and Recover. ✔️ Consistent security controls acro..

NIST Cybersecurity Framework RELIANOID compliance
Link
@varbear shared a link, 2 weeks, 6 days ago
FAUN.dev()

Goodbye Microservices

Twilio Segment collapsed 140+ destination-specific microservices into asingle monolith, one repo, one set of dependencies, one test harness. They leveled out version sprawl and builtTraffic Recorder, a homegrown yakbak-based HTTP playback tool. That killed off hours-long test runs, dropping them to.. read more  

Link
@varbear shared a link, 2 weeks, 6 days ago
FAUN.dev()

Why I Didn’t Sign the Resonant Computing Manifesto: The Foundations Need Work

A sharp critique of theResonant Computing Manifestopushes it past vague ideals. It calls for real governance scaffolding, not just poetic prose. Without that? The manifesto risks becoming just another glossy PDF for entrenched players to wave around while changing nothing. Under the hood:What’s real.. read more  

Why I Didn’t Sign the Resonant Computing Manifesto: The Foundations Need Work
Link
@varbear shared a link, 2 weeks, 6 days ago
FAUN.dev()

Rust unit testing: file writing

To test file writes without hitting the disk, the author swaps in a closure that takes a file handle. That handle’s a test double, so after the code runs, you can crack it open and inspect what got written... read more  

vLLM is an advanced open-source framework for serving and running large language models efficiently at scale. Developed by researchers and engineers from UC Berkeley and adopted widely across the AI industry, vLLM focuses on optimizing inference performance through its innovative PagedAttention mechanism — a memory management system that enables near-zero waste in GPU memory utilization. It supports model parallelism, continuous batching, tensor parallelism, and dynamic batching across GPUs, making it ideal for real-world deployment of foundation models. vLLM integrates seamlessly with Hugging Face Transformers, OpenAI-compatible APIs, and popular orchestration tools like Ray Serve and Kubernetes. Its design allows developers and enterprises to host LLMs with reduced latency, lower hardware costs, and increased throughput, powering everything from chatbots to enterprise-scale AI services.