Local AI
169 articles
-
What Can You Run on a DGX Station GB300? Workload Guide (SG)
What an NVIDIA DGX Station GB300 runs in a Singapore office: FP16 70B fine-tuning, 400B+ FP4 serving, agent stacks, private RAG, MIG sharing, rack limits.
Wei Fei
·
-
Supermicro ARS-511GD-NB-LCC: DGX Station GB300 Guide SG
Supermicro ARS-511GD-NB-LCC guide for Singapore: liquid-cooled GB300 DGX Station, 748 GB coherent memory, 1,600 W Titanium PSU, tower or 5U rack, Redfish BMC.
Wei Fei
·
-
GIGABYTE W775-V10-L01 DGX Station GB300 Guide for Singapore
GIGABYTE W775-V10-L01 guide for Singapore: GB300 Superchip, 748 GB coherent memory, 1,600 W Platinum PSU, three PCIe 5.0 slots, RTX PRO 6000 support.
Wei Fei
·
-
Local AI vs Cloud API Cost: 12-Month 70B Model (Singapore)
Local AI vs cloud cost in Singapore: a transparent 12-month worked example comparing an ASUS Ascent GX10 with hosted 70B and frontier API prices.
Wei Fei
·
-
Best PC to Run LLMs Locally in Singapore: 7B to 550B Ladder
Best PC for running Llama, DeepSeek or Qwen locally: a model-size ladder from ASUS NUC 16 Pro and ROG NUC to Ascent GX10 and ET900N G3, with Singapore...
Wei Fei
·
-
Ollama on Intel Arc B390: Local LLMs on the ASUS NUC 16 Pro
Run Ollama or LM Studio on the ASUS NUC 16 Pro (Core Ultra X9 388H, Arc B390): Vulkan setup on Windows 11, models that fit 32 GB vs...
Wei Fei
·
-
OpenClaw on DGX Spark: NemoClaw Setup on ASUS Ascent GX10
Step-by-step NemoClaw install on the ASUS Ascent GX10 (DGX Spark class): sandboxed OpenClaw agents, local vLLM or Ollama models, policies, PAIR. SG guide.
Wei Fei
·
Wei Fei
