How-To
157 articles
-
From DGX Station to GB300 NVL72: How NVIDIA Scales AI From Your Desk to the Data Centre
How to build and debug AI on a DGX Station GB300 and scale to GB300 NVL72: what carries over, what needs testing in the rack and a step-by-step...
Wei Fei
·
-
Can NVIDIA GB300 Replace Your Company's Multi-GPU AI Server?
An 8-step checklist for replacing a 4-8 GPU RTX PRO 6000 or HGX H100 server with GB300: model size, concurrency, parallelism, Arm, uptime, power.
Wei Fei
·
-
Can RTX 5090 Run Your Own AI Video Generation Studio?
Can an RTX 5090 run a local AI video studio? Wan 2.2, HunyuanVideo and LTX-2 VRAM needs, measured render times and ComfyUI workflows.
Wei Fei
·
-
RTX 5090 Power Connector Controversy: Should Buyers Be Worried?
RTX 5090 12V-2x6 connector melts explained: dated incidents, why per-pin current matters, PSU and cable choice, seating and routing tips.
Wei Fei
·
-
Can RTX PRO 6000 Blackwell Be the Perfect Companion to NVIDIA DGX Station?
Pair an RTX PRO 6000 with a DGX Station GB300: physical AI, digital twins, rendering and CFD workflows, power caps, and inside-the-box vs separate workstation.
Wei Fei
·
-
NVIDIA's Most Powerful Workstation GPU Has a 600W Appetite. Is Your PC Ready?
Is your PC ready for a 600W RTX PRO 6000? PSU sizing, the 16-pin connector, cooling, chassis fit, Singapore 13A sockets and UPS sizing explained.
Wei Fei
·
-
No NVLink Bridge on RTX PRO 6000 Blackwell: Does That Make Multi-GPU AI Harder?
No NVLink bridge on RTX PRO 6000 Blackwell: measured PCIe P2P speeds, which AI workloads it hurts, how to tune a multi-GPU build and when to move to...
Wei Fei
·
-
Why Your RTX 5090 Might Never Reach Its Full Potential
Why an RTX 5090 underperforms: CPU limits, PCIe lanes, power cabling, heat, VRAM spill in AI work and software, with published test data.
Wei Fei
·
-
A 96GB GPU Is Only Half the Story: Why Your RTX PRO 6000 Needs the Right Workstation
What an RTX PRO 6000 Blackwell needs from its workstation: CPU lanes, slots, system memory, PSU at 230 V, cooling and ISV certification explained.
Wei Fei
·
-
RTX 5090 PCIe 5.0 vs PCIe 4.0: Are You Losing Performance on an Older Motherboard?
RTX 5090 PCIe 5.0 vs 4.0 vs 3.0: measured gaming loss, DaVinci Resolve and AI effects, x8 lane sharing and whether to upgrade your board.
Wei Fei
·
-
Undervolting an RTX 5090: Can Less Power Deliver Almost the Same Performance?
RTX 5090 undervolting and power limits: published results, frames per watt, AI throughput at 400 W, Singapore running-cost savings and risks.
Wei Fei
·
-
Building Your Own RTX PRO 6000 Workstation? Here's What Could Go Wrong.
DIY RTX PRO 6000 Blackwell build risks: card fit, 600 W power, heat, stacked GPUs, BIOS boot errors and who handles warranty when parts fail.
Wei Fei
·
-
From 10,000 Documents to One Private AI Brain: What Can DGX Station Really Deliver?
What a DGX Station GB300 delivers for private AI search across 10,000 documents: corpus sizing, RAG pipeline, users served, permissions and pitfalls.
Wei Fei
·
-
Can NVIDIA DGX Station Become the Foundation of Your Company's Private AI Cloud?
Turn a DGX Station GB300 into a private AI cloud: OpenAI-compatible API, gateway budgets, SSO, audit logs, MIG quotas and the limits of a single node.
Wei Fei
·
-
NVIDIA DGX Station's 748GB Memory: Can It Really Handle a Trillion-Parameter AI Model?
A 1T model needs ~500GB at 4-bit, 1,000GB at 8-bit. KV cache, context length, HBM3e vs LPDDR5X and what DGX Station GB300 can really run.
Wei Fei
·
-
How Many Tokens Must a DGX Station Generate Before It Pays for Itself?
Break-even maths for a DGX Station: tokens per month needed at S$120k-S$180k hardware against hosted open models, GPT, Sonnet and Opus API prices.
Wei Fei
·
-
A S$150,000 AI Workstation: How Much Productivity Must It Generate to Justify the Investment?
A transparent three-year ROI model for an NVIDIA DGX Station in Singapore: costs, hours to save per employee, sensitivity and how to test the case.
Wei Fei
·
-
The 24/7 AI Factory: Can DGX Station Turn Idle Hours Into Business Value?
A DGX Station idles 73% of the week. How much overnight batch work it can do, what the same jobs cost via API, and how to schedule day...
Wei Fei
·
Wei Fei
