Infrastructure & Runtime
vLLM Configs Shift Energy, Latency, and Accuracy: A 9,000-Run Study
A 9,000-run study shows vLLM attention kernels and prefix caching shift energy, latency, and accuracy. Benchmark configs per workload before production.
The Groundy archive · Page 19 of 43
All articles
Browse Groundy's complete archive of 1027 articles on AI, developer tools and infrastructure. Page 19 of 43.
The archive
433–456 of 1027 articles · Newest first
- Ethics, Policy & Safety
HuggingFace vs GitHub Models vs Replicate: Policy Compliance for Uploaders
HuggingFace, GitHub Models, and Replicate enforce independent content policies. Uploaders must verify each hub's rules at release time because acceptance on one platform does.
Models & ResearchKimi K3: 2.8T Parameters, MoE Routing, and Self-Hosting Reality
Kimi K3 pairs 2.8T parameters with 16-of-896 expert routing and a 1M context window. Hosted API is live, but full weights arrive July 27, 2026. Self-hosting requires.
- Models & Research
Kimi K3 vs Qwen3.8 Max: Routing Strategy for July 2026
Kimi K3 offers concrete pricing for production bake-offs while Qwen3.8 Max remains a shadow evaluation. Compare both against Fable 5, GPT-5.6 Sol, and GLM-5.2 by workload.
- Agents & Frameworks
Cloudflare's Agent Stack: Edge Trust, Identity, and Metering
Cloudflare positions its edge network as the trust boundary for agents. This analysis maps Temporary Accounts, x402 metering, and detection patterns for framework authors.
- Models & Research
Qwen3.8 Max Release Audit: API, Open Weights, and the License Catch
Qwen3.8 Max now has a stable API, $2/$6 pricing, 1M context, and downloadable 2.4T weights. The release is real, but the API and open checkpoint are not the same product.
- Ethics, Policy & Safety
SAMark Text Watermarking: Paraphrase Robustness and the Policy Gap
SAMark reports 90.2% detection under paragraph paraphrase, yet no mandate specifies a robustness threshold. Without a survival metric, disclosure rules reduce to suggestions.
- Industry & Business
HuggingFace's $100M Series C Locks Teams Into Deployment
HuggingFace's $100M Series C shifts focus from free model hosting to paid inference. Open weights no longer prevent vendor lock-in, as switching costs now accumulate at the.
Models & ResearchHuggingFace 100x Inference: Generalizable vs Platform-Locked Optimizations
HuggingFace claims 100x inference speedup, but independent benchmarks show its TGI engine trails vLLM by 24x on throughput. This article decomposes the claim to separate.
- Agents & Frameworks
LM Studio Bionic vs Claude Code: Local-First vs Cloud Agent Tradeoffs
LM Studio Bionic and Claude Code split on a data-egress line. Local agents erase per-run cost but cap you at open-model tool-use reliability. Cloud agents charge $17 to $100.
- Infrastructure & Runtime
AWS Estimated Billing Was Off by $1.7B: Reconciling Actual Cloud Spend
A $1.7 billion AWS estimated billing error exposes why console projections are direction indicators, not bookable numbers. Teams must reconcile against the Cost and Usage.
- Models & Research
Kimi K3 Code Arena Rank: Self-Hosting Cost Math for Coding Agents
Kimi K3 is frontier-class, but its 2.8T MXFP4 checkpoint needs 64-plus accelerators. The real deployment math reveals what an arena rank cannot.
- Infrastructure & Runtime
Cloudflare Attribution vs Custom Logs: The Per-Path AI Crawler Decision
Cloudflare Attribution packages AI crawler traffic as business insights, but rollups hide the per-path decisions operators actually need. Compare vendor dashboards against.
- Developer Tools
Grok CLI uploads entire workspace to GCS by default, independent of model reads
Wire-level analysis shows Grok CLI 0.2.93 uploads the entire workspace to GCS by default, regardless of model reads. The settings toggle fails to disable this side channel,.
- Infrastructure & Runtime
Spectral Compute CUDA Translation: vLLM Procurement vs Porting Cost
Spectral Compute proposes CUDA binary translation for AMD and Intel GPUs, but vLLM already supports native HIP. The real question is whether translation overhead beats the.
- Infrastructure & Runtime
Running MiniCPM-V-4.6 on Fermi: What 6 GB of VRAM Forces
MiniCPM-V-4.6 runs on a 2011 Fermi GPU with 6 GB VRAM. The study shows software staging recovers multimodal inference, shifting the constraint from hardware procurement to.
- Agents & Frameworks
Can a Malicious AGENTS.md File Compromise Your Coding Agent? A Threat Model
AGENTS.md and .cursorrules are a write-once, compromise-many attack surface. arXiv 2607.15143 shows a deterministic pre-install check, not sharper models, stops it.
- Infrastructure & Runtime
LLM Inference Without a GPU: Pure CPU vs Hybrid CPU-GPU Scheduling
ATSInfer schedules LLM tensors across CPU and integrated GPU at tensor granularity, reporting up to 3.29× decode speedup where pure-CPU inference is memory-bandwidth-bound.
- Culture & Society
When Cultural LLM Alignment Gets a Positive Target, Who Writes the Spec?
A Korean alignment paper argues LLM cultural work leans on suppression lists and needs positive specs. Whoever writes the spec owns the model's definition of a culture.
- Agents & Frameworks
How GitHub Projects Actually Adopt Coding Agents: New Empirical Data
A July 2026 arXiv study of 25,264 agentic PRs finds most GitHub repos generate only one to two agent PRs per quarter, contradicting vendor hype about intensive adoption.
- Infrastructure & Runtime
RL-Found CUDA Kernels Beat cuBLAS: Kernel Tuning Shifts to Reward Design
CUDA-L2 used reinforcement learning over 1,000 kernel configurations to beat cuBLAS by 19.2% on HGEMM, shifting the scarce kernel skill from hand-tuning to reward design.
- Ethics, Policy & Safety
A Digital Twin Can Validate AV Safety, but No Regulator Accepts the Evidence
A July 2026 risk-field digital twin replays rare AV hazards no road fleet can accumulate, but no regulator defines when simulated miles count as certification evidence.
- Open Source
62.7% of Linux Foundation Repos Still Carry Non-Inclusive Terms, and LLMs Are Learning Them
A NISCAN audit of 461 Linux Foundation repos finds non-inclusive terms down 47% since 2020, yet 62.7% still carry Tier-1 terms, and LLMs keep resurrecting the old names.
- Ethics, Policy & Safety
Why EU AI Act Monitoring Will Miss Discontinuous LLM Alignment Failures
Hair-Trigger Alignment shows an LLM can pass every black-box probe, then flip after one benign update, so EU AI Act periodic monitoring certifies less than it assumes.
