➀ Chinese companies are modifying NVIDIA GeForce RTX 40 series GPUs to increase VRAM for AI workloads, particularly for cloud computing providers. ➁ The GeForce RTX 4090D 48GB and RTX 4080 SUPER 32GB are being upgraded with custom PCBs to double their memory capacity. ➂ These modified GPUs are in high demand among cloud service providers, renting for as low as $0.03 per hour, reflecting the strong market need for AI computing power.
Related Articles
- PAVE360 SDV tech available on AMD CPUs and GPUs on Azure8 months ago
 - ByteDance plans to sidestep U.S. sanctions by renting Nvidia GPUs in the cloud — report says it has set aside $7 billion budget10 months ago
 - Oracle offering 2.4zettaFLOPSabout 1 year ago
 - Alibaba and Tencent Clouds See Demand for CPUs Level Off, GPUs Accelerateabout 1 year ago
 - Nebius Pullback: The Smart Money Entry Point11 days ago
 - October 2025 Issue Of Electronics For Youabout 1 month ago
 - Next-Gen GPU Platform Redefines AI Performanceabout 2 months ago
 - Nvidia launches GPU for disaggregated inferenceabout 2 months ago
 - NVIDIA GeForce RTX 5090 and the Age of Neural Rendering at Hot Chips 20252 months ago
 - Nvidia introduces compact Blackwell professional graphics cards — RTX Pro 4000 SFF and Pro 2000 GPUs arrive at SIGGRAPH 20253 months ago