ASUS

Local MCP Server with Ollama and ChromaDB

An engineering post‑mortem describes building a local MCP server for codebase memory using Ollama (local LLM hosting) and ChromaDB (vector retrieval) within the open-source zerikai_memory project. The author tested two local models (mistral:7b and ornith:9b) on an 8GB RTX 3050 system and measured latency, synthesis quality, and operational issues. Key findings: mistral:7b is faster and fits 8GB VRAM, while ornith:9b produces denser, more precise briefs given enriched docstrings but has longer cold starts and higher VRAM needs. The team implemented an ollama_semaphore / OLLAMA_MAX_CONCURRENCY gate to avoid GPU saturation during concurrent brief synthesis. The recommended workflow emphasizes docstring enrichment (embedding-docstring → scan_workspace) and advises RTX 3060 12GB as the practical minimum for ornith:9b in production local deployments.

Acer and Asus Resume PC Sales in Germany After Nokia Settlement

Acer and Asus have reached a settlement with Nokia in a patent dispute and are again allowed to sell PCs and notebooks in Germany. The agreement, reached with the help of a private arbitration panel, follows a court-ordered sales ban by the Munich I regional court in mid-February 2026. The dispute centered on patent EP2661892, covering the H.265/HEVC video standard; Nokia and the manufacturers disagreed over how licence fees should be calculated. Neither the settlement terms nor future royalty levels were disclosed. Both vendors say they have resumed distribution and marketing in Germany and their German online shops list devices for immediate order and delivery.

NVIDIA RTX Spark: 128GB Unified CPU–GPU Memory

NVIDIA unveiled the RTX Spark "superchip" around Computex/Build 2026. RTX Spark pairs a 20-core Grace CPU with a Blackwell RTX GPU connected via NVLink‑C2C into a single 128GB coherent memory pool, removing host↔device PCIe copies for zero‑copy CPU–GPU access. The Blackwell GPU is described with 6,144 CUDA cores and fifth‑generation FP4 Tensor Cores, and NVIDIA cites up to 1 petaflop of low‑precision (FP4) peak AI compute. Consumer NVLink‑C2C bandwidth is undisclosed. NVIDIA positions RTX Spark for on‑device AI agents and large context models; laptops and compact desktops from ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI are expected to ship RTX Spark systems in fall 2026. The release highlights reduced PCIe transfer overhead rather than eliminating memory bandwidth limits entirely.