Fifteen years ago, Jen-Hsun Huang, who set his sights on medical care, exclaimed that computer-aided drug discovery was “truly a miracle.” He once said that, like computer-aided chip design, in the field of drug discovery, one can move from computer-aided drug discovery to computer-aided drug design. "If we scale it up a billion times, we can simulate biology."
In Jen-Hsun Huang's view, life science is a "new continent" that has not yet been fully developed by human scientific and technological achievements, and new computer systems may be able to help biologists further understand and process extremely complex life systems.
"In the future, life sciences will be as highly engineered as traditional industries. When data science, artificial intelligence and automation are combined, biology will improve exponentially, becoming the next golden track."
This scene is happening. Since this year, the technical support of AI tools such as ChatGPT has enabled AI medical care to attract money around the world. In the first half of 2024, there were 69 global AI pharmaceutical financings, with an investment amount of US$3.336 billion, close to the level of the same period in 2022, and far exceeding the same period in 2023.
This week, Jen-Hsun Huang appeared in Vietnam and announced: NVIDIA has acquired VinBrain, a Vietnamese healthcare startup.
This transaction seems unexpected, but it has traces. Since this year, Jen-Hsun Huang has strongly expressed his enthusiasm for AI medical care in multiple public occasions. This time, he bought a star AI medical model company VinBrain from Vietnam, becoming another footnote in betting on AI medical care. Now when you open the official website of VinBrain, the notice “VinBrain is now a part of NVIDIA” has appeared on the homepage.
VinBrain focuses on applying AI, Internet of Things and other technologies to the medical field, using computer vision and machine learning to design products and create customized solutions for the medical industry, enabling AI diagnosis to view vital signs, Blood tests, medical imaging, etc. For example, applying AI to imaging can help radiologists quickly identify serious diseases such as lung cancer and brain tumors, improving the possibility of early diagnosis.
The acquisition of NVIDIA is not a whim: According to local media reports, since its establishment, VinBrain has long used NVIDIA DGX SuperPOD for AI training, and used NVIDIA GPUs to improve operating efficiency and deployment. It has developed more than 300 tools for processing voice, AI models for text, video, and imaging including X-ray, CT, and MRI data. The company also streamlines inference on more than hundreds of AI models with NVIDIA Triton Inference Server and NVIDIA TensorRT to improve model efficiency.
Last year, the company developed DrAid, an artificial intelligence driven medical assistant using NVIDIA GPU drivers ™ It has been used by over 100 hospitals in Vietnam and the United States, helping it establish a reputation internationally. When you open the official website of VinBrain, NVIDIA is placed at the top of the list of "Research and Business Partners." The company has also appeared in related events several times as a chief member of the NVIDIA Startup Acceleration Program.
In the past two years, "AI + medical care" has become the focus of participation in venture capital in Nvidia - its investment arm Nventures was born in early 2023. So far, it has invested in at least 13 AI medical companies, and its investment speed is even faster than that of many professional biomedical VC.
The investment portfolio includes companies familiar to the industry such as Recursion, a large AI biological model unicorn that has just completed mergers and acquisitions, Genesis Therapeutics, an AI protein star company, and lambicTherapeutics, an AI small molecule drug platform. “For NVIDIA, the primary criterion for investing in start-ups is relevance.” The person in charge of NVentures once explained that the relevant companies it invests in all use the chip products from NVIDIA and build businesses based on this technology.
