Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124

[ad_1]
NVIDIA Corp’s co-founder and CEO Jensen Huang, Taiwan, Taiwan, talks to Taiwan in Taiwan in Taiwan in Taiwan in Taiwan.
Bloomberg | Bloomberg | Getty pictures
Nvidia CEO Jensen Huang has revealed an announcement, and new products aimed at storing the company’s artificial intelligence in Monday and Monday.
One of the most notable ads was a new New “NVLINK FUSPION” program that allows customers and partners to use NVIDIA products and NVIDIA central processing sections and graphic processing sections.
So far, nvlinks were closed to the plugs made by NVIDIA. NvLink is a technology designed to exchange and exchange information between NVIDIA and GPUS and CPUs.
“NV Link Fusion, not only half special chips, for you can build a half-special AI infrastructure,” Huang said at Huang 2025 at the largest electronics conference “Asia.
According to Huang, NVLink Fusion AI infrastructure allows NVIDIA processors to connect with different CPU and special integrated circuits (ASIC). “In any case, there is benefit from using the NV link infrastructure and NV link ecosystem.”
Nvidia Announced on Monday Includes AI Chipmmiring partners for NvLink Fusion Mediap, Martyr, Alchip, Astera Laboratory, Synops and Cadence. NVLINK FUSION, Like NVIDIA customers Fujitsu and Qualcomm technologies He added that the AI will be able to connect his third party CPU with NVIDIA GPU.
According to the Washington-based semi-tract and technology analyst, NVIDIA’s NVIDIA rivals represent NVIDIA’s plans to share the share of information centers based on Asics.

Although NVIDIA has held a dominant position in GPU, which is used for common AI training, many opponents see Expanding room In chips designed for a more specific application. NVIDIA’s largest competitors, the biggest competitors of NVIDIA – this is one of the largest customers, Google, Microsoft and Amazon build private processors.
Nvlink Fusion “NVIDIA NVIDIA connects NVIDIA as the center of NVIDIA factories, even if these systems are not built with NVIDIA plugs, it allows customers who do not establish NVIDIA-based systems, but are united part of some GPU.
“If the extensive adoption, Nvlink Fusion can expand industrial traces of NVIDIA with special CPU developers and Asik designers in the establishment of the infrastructure of the AI infrastructure in the future,” he said.
However, Nvlink Fusion NVIDIA has reduces NVIDIA’s CPU demand for NVIDIA customers to use alternatives to NVIDIA customers in new street research.
Nevertheless, “System level, additional comfort, increases NVIDIA GPU-based solutions to NVIDIA’s architecture, helps NVIDIA maintain its position in the AI Computing Center,” he said.
NVIDIA’s competitors Broadcom, AMD and Intel have no nvlink fusion ecosystem so far.
HUANG opened NVIDIA with an update for the next generation of NVIDIA GRACE Blackwell systems for AI workload. The company’s GB300 will be released in the third quarter of this year, he said he would provide a higher general system performance.
On Monday, NVIDia, NVIDIA, NVIDIA DGX Cloud Lepton, NVIDIA’s world’s EU developers will connect a cloud provider with tens of thousands of GPUs.
“DGX Cloud Lepton, connecting the power to strengthen Cloud AI and GPU, helps to solve the critical problem in ensuring the secure, high-effective GPU resources, the NVIDIA computing ecosystem.” press release.
In his speech, Huang announced plans for a new office for a new office in Taiwan, where the world’s largest electronics manufacturer Hon Hai Technology Group will establish a Taiwan Foxconn and the EU Supercomputer project.
“We are happy to establish the infrastructure of the infrastructure of the Taiwan’s EI infrastructure and Foxconn and Taiwan and Taiwan to help develop the AI and other leading companies in the age of TSMC and other leading companies,” he said.
[ad_2]
Source link