| Mobile Web

Search results for NVLink

Industry

GPU standalone era may end as integrated design determines AI chip performance

The focus of competition in AI chips is shifting from standalone GPUs to integrated CPU-GPU designs that package processors and memory to optimise data flow. U.S.-listed CPU makers Intel and AMD have surged, reflecting rising demand for CPUs and DPUs as inference and agentic AI workloads expand. Nvidia is also pushing CPU offerings and tighter CPU-GPU integration, while memory is becoming a core component. The trend could benefit South Korean memory makers SK Hynix and Samsung Electronics.

Industry

Nvidia steps up push into South Korea\'s AI ecosystem with CUDA-style lock-in strategy

Nvidia has released 7 million synthetic Korean-language personas for free and then added the multimodal model Nemotron3 Nano Omni, signalling the rollout of a four-step lock-in package linking models, data, frameworks and hardware in South Korea. The Nemotron-Personas-Korea dataset is available on Hugging Face under a CC BY 4.0 licence. Nvidia is also expanding the Nemotron3 lineup and open-sourcing post-training tools in its Nemo framework, while tying performance to its Blackwell hardware features.

AI & Enterprise

Microsoft AI chief says AI limits still far off, computing power to jump 1,000-fold by 2028

Mustafa Suleyman, chief executive of Microsoft’s AI group, said AI development is unlikely to hit limits any time soon, citing a surge in computing capacity. He wrote that compute used to train leading models has expanded sharply since 2010 and that gains come from chips, memory, networking and software efficiency. He said training times have fallen dramatically and forecast effective computing could rise 1,000-fold by 2028, while power remains a key constraint.