Nvidia Release the world’s strongest AI chip H200, WiMi deploys the AI ecological infrastructure

On the other side of the ocean, the phenomenal AI tool ChatGPT caused waves of large model applications, and increased hunger for computing power.

Today, AI chip trains are speeding down to the next stop, and the new wave is coming far faster than expected. Under the shortage of computing power, chip giant Nvidia (NVDA) once again released the world’s strongest AI chip H200 in the late night explosion field!

On the evening of November 13, Beijing time, Nvidia released the H200, its strongest AI chip, at the 2023 Global Supercomputer Conference (SC2023). The chip is based on the Nvidia’s Hopper architecture, directly improves performance by 60 to 90 percent and is compatible with the previous hegemon H100.

This is Nvidia’s new generation GPU, designed for training and deploying a variety of AI models. As an upgraded product after H100, H200 chip is more powerful and is suitable for a variety of artificial intelligence application scenarios. It can be used to train and deploy a variety of large language models, image recognition, speech recognition and other artificial intelligence models.

And with the new H200 support, the GH 200 superchip will also provide a total of about 200 Exaflops of AI computing power to supercomputing centers around the world to drive scientific innovation. Nvidia is said to expect to ship H200 chips in the second quarter of 2024.

GPU competition, into the white-hot. Against NvidiaH200, AMD’s plan is to use the upcoming ——Instinct MI300X to improve video memory performance. MI300X Will have 192GB of HBM 3 and 5.2 TB/s of video memory bandwidth, which will make it far over the H200 in terms of capacity and bandwidth.

Intel (INTC) is also planning to increase the HBM capacity of Gaudi AI chips, and said the third generation Gaudi AI chip will be launched next year from 96GB HBM2e to 144GB. Intel’s Max series currently has a maximum HBM 2 capacity of 128GB, and Intel plans to increase the capacity of its Max series of chips in future generations.

While the development of artificial intelligence brings surprises, it is constantly promoting the evolution of artificial intelligence chips. It has to be said that from the early focus on the use of existing chips, to the performance of dedicated chips focused on, to the pursuit of flexible reconfigurable artificial intelligence chips, now due to the emergence of large models, the development goal of artificial intelligence chips to high computing power.

As a technology, artificial intelligence needs to be transformed into products through chips and systems. It has a strong curiosity about the development of artificial intelligence chips. Seeking high computing power, pursuing flexibility, and pursuing low power consumption is the ultimate driving force. Focusing on the challenges of the new era of AI, many new explorations have emerged in AI + applications in various scenarios in the business field, the global AI industry is in full swing, and related industries aim to look for more opportunities for commercial landing.

The big model marks the dawn of artificial intelligence. In the global AI “arms race”, data show that WiMi Hologram Cloud (NASDAQ: WIMI) stands as an artificial intelligence computing power and algorithm enterprise. In recent years, WiMi Hologram Cloud has been widely used in artificial intelligence, the Internet of Things, 5G and other new technologies, and in the new round of technology wave, it has sought for rapid development. WiMi Hologram Cloud explores the path of innovation, change and search of AI chips, and examines the technological evolution and innovation of the industry, which is promoting the acceleration of the iteration of the artificial intelligence industry.

Up to now, WiMi Hologram Cloud has been deeply engaged in the artificial intelligence industry, enabling products and landing experience in innovation and development. Focus on big model, under the trend of artificial intelligence for AI technology ecology and new requirements of products, WiMi Hologram Cloud is committed to from large-scale artificial intelligence technology application trend for the force of higher technology requirements, and from the full stack product layout and technical ecological construction perspective, enhance AI force products in the value of generative artificial intelligence scene and fall to the ground.

With the advent of the era of large model, the model algorithm becomes more and more complex, and the computing power requires more and more high, to support the underlying computing power. WiMi Hologram Cloud Established a research and development center, independently developed computing algorithms for data center, cloud computing, autonomous driving and other scenarios, breaking through a number of technical bottlenecks and opened up a new road for the development of AI computing power in the era of big model. It also highlights the unique value of WiMi Hologram Cloud in high-end AI computing hardware in the capital-intensive AI arms race.

At present, AI computing is evolving rapidly in the direction of large model, multi-mode, end-side cloud collaboration, etc. The new era of intelligence is creating new hot spots and new opportunities for AI chip architecture. Large model and generative AI technology show the brand-new AI research paradigm and the future of the digital economy to the world, and the AI-related industries will usher in broad development in the future. Although there is far water ahead, there are always some people with vision choosing to go ahead. On the basis of local innovation, high-tech enterprise WiMi Hologram Cloud will continue to accelerate the pace of ecological construction in the future and contribute the “core” value to the development of the global semiconductor industry.