Source: Broker China
Author: Shi Qian

According to the news on March 22, local time,$Hitachi (6501.JP)$Already with$NVIDIA (NVDA.US)$The partnership combines Hitachi's solutions in operational technology and leading positions in key industries such as energy, mobile and connected systems with Nvidia's expertise in generative AI to accelerate social innovation and digital transformation.

Through this partnership, generative artificial intelligence will now be extended to the industrial sector by capturing large amounts of data generated in the field of operational technology (from various sensors, devices, and industrial machines), and applied to simplify performance, gain insights, and allow organizations to automate operations. This wasn't possible before. Additionally, Hwang In-hoon recently released the 5th edition of Nvidia's artificial intelligence enterprise platform. The platform uses new technology, and Nvidia executives described it as Nvidia's Inference Microservice (NIM).

According to foreign media reports, in the era of artificial intelligence (AI) being integrated into the industry structure, a new competitor has emerged and dominated the right to speak: ChatGPT Enterprise (enterprise subscription service). This plan is the brainchild of OpenAI. It is not only a product, but also a revolution aimed at leading companies into a new era of cognitive ability. OpenAI has equipped ChatGPT Enterprise with features beyond conventional ones, changing the way businesses interact with AI.

Notably, the Enterprise Connect (EC) conference will be held in Gaylord Palms, Orlando, Florida from March 25 to 28, 2024. Microsoft will launch two new Copilot-enabled devices in April: Surface Pro 10 and Surface Laptop 6. Meanwhile, on Friday US time, Astera Labs, a company involved in connecting cloud service providers to enterprise data, surged more than 9%.

So will big enterprise-level models also usher in the limelight?

Giants work together

The big enterprise-level model is probably where the money comes in the fastest.

Nvidia is starting to work at the enterprise level. According to news on March 22, local time, Hitachi has cooperated with Nvidia to combine Hitachi's solutions in the field of operational technology and leading positions in key industries such as energy, mobile, and connected systems with Nvidia's expertise in generative AI to accelerate social innovation and digital transformation. Through this partnership, generative artificial intelligence will now be extended to the industrial sector by capturing large amounts of data generated in the field of operational technology, and applied to streamline performance, gain insights, and allow organizations to automate operations. This wasn't possible before.

Recently, Nvidia has released the 5th edition of its artificial intelligence enterprise platform. The platform uses new technology, which Nvidia executives described as Nvidia's Inference Microservices (NIM). Built for portability and control, NIM supports model deployment across a variety of infrastructures, from local workstations to the cloud to local data centers. NIM is part of NVIDIA AI Enterprise and uses enterprise-grade basic containers to provide a solid foundation for enterprise AI software through functional branching, strict verification, enterprise support for service-level agreements, and regular security updates. Comprehensive support structures and optimization capabilities highlight NIM's role as a key tool for deploying efficient, scalable, and customized AI applications in production.

By the close of the US stock market on the 22nd local time, Nvidia had risen 3.12%, an increase of more than 90% during the year, with a total market capitalization of 2.36 trillion US dollars (approximately RMB 17 trillion).

In addition, according to foreign media reports, ChatGPT Enterprise is beginning to dominate the right to speak. According to reports, OpenAI has begun customizing pricing services for this, and the price will be adjusted according to the size and purpose of the company. According to the report, this is an intellectual auction: to start bidding, companies need to provide OpenAI with details from the size of the company to the industrial sector. In return for “customization,” OpenAI creates a “set of packages” tailored for each enterprise organization.

Analysts expect many businesses to deploy small language models in-house so they can fine-tune them based on company data without moving sensitive information to the public cloud. Additionally, running a model in a data center is sometimes cheaper than running in the cloud. Robin Bordoli, chief marketing officer at AI model training platform maker Weights and Biases, said having the tools to automate model-related processes meant that traditional software engineers could do the job rather than hard-to-find AI experts.

Weights and Biases has integrated its software with Nvidia's inference engine, so developers can train and reason on a platform that supports 30 basic models. Today, Weights and Biases has 1,000 customers, many of which are government agencies and life sciences organizations, Bordoli said.

It's worth noting that Microsoft is also making big moves. According to the latest news, Microsoft will launch two new Copilot-enabled devices in April: Surface Pro 10 and Surface Laptop 6. The practical application of generative artificial intelligence changed from an abstract future concept to a concrete reality in just a few months. Businesses and organizations large and small are busy figuring out if and how artificial intelligence can help their employees be more productive and efficient. For organizations using Microsoft software, the application of artificial intelligence in business environments is being led by the Microsoft Copilot platform. The trend of artificial intelligence changing the business format seems unstoppable.

The night before the multimodal outbreak

Various indications suggest that after the rise of big models, multimodality is already on the eve of an explosion.

The essence of multimodality is to use richer sensory channels beyond language such as sight, hearing, touch, and taste to simulate human ability to understand and express information. The ideal multi-modal large model has the ability to generalize understanding and generation across modes, which is more in line with the way humans perceive the world.

CICC believes that multi-modality may further open up the upper limit of AI capabilities. The industry is also actively exploring possible technical paths for multi-modal large models to “replicate” the success of big language models in the multi-modal field. There are still endless possibilities for future industrial development.

On the evening of March 20, Artificial Intelligence Company Fourth Paradigm released its first annual results since listing. Financial reports show that in 2023, the fourth model achieved revenue of 4.2 billion yuan, a year-on-year increase of 36.4%; gross profit of 1.98 billion yuan, a year-on-year increase of 33.2%, gross profit margin of 47.1%; adjusted net loss of 415 million yuan, a decrease of 88.88 million yuan compared to 2022, a decrease of 17.6% year-on-year. The company targets the B-side market, focuses on the big model of the industry, and focuses on differentiation.

At the performance briefing on the same day, Dai Wenyuan, chairman of the board of directors and CEO of the company, said that the underlying logic of the fourth paradigm is to use artificial intelligence technology to help all walks of life discover more and more rules, form larger and larger models — big industry models, and improve the production and operation efficiency of enterprises. Due to the company's positive performance, many market participants believe that the spring of big models is really coming.

Meanwhile, at a recent media conference, IBM also announced that the 2024 Greater China Strategy will continue to promote hybrid cloud and AI, focus on enterprise-level AI applications, and build an open ecosystem partnership system.

Zhou Hongli, member of the National Committee of the Chinese People's Political Consultative Conference and founder of 360 Group, stated in the “Proposal on Deepening Multi-Scenario Applications of Artificial Intelligence to Support the Development of Large Models in the Direction of Verticalization and Industrialization” that in terms of big model applications, 2024 is the first year of big model application scenarios, and China can fully embark on a big model development path with Chinese characteristics. An important direction for China's development model should be to use the advantages of industries and scenarios to combine big models with business processes and product functions to seek multi-scenario applications, verticalization, and industrialization to help accelerate the formation of new quality productivity.

Editor/Somer