TMTPost -- Foxconn, the world's largest electronics manufacturer, is set to achieve a milestone with the largest production facility of Nvidia’s GB200 superchips, aiming to meet the rising demand from Nvidia’s next-generation Blackwell computing platform amid the artificial intelligence (AI) boom.

打开网易新闻 查看精彩图片

Credit:Nvidia

"We're building the largest GB200 production facility on the planet," said Benjamin Ting, Foxconn senior vice president for the cloud enterprise solutions business group, at the company's annual "Hon Hai Tech Day" in Taipei Tuesday. ”Everyone is asking for Nvidia's Blackwell chips, the demand is awfully huge.”

Foxconn Chairman Young Liu said at the same event that Foxconn would be "the first to ship these superchips". He then told reporters the new plant was being built in Mexico and its capacity would be “very, very enormous.” Reuters quoted a Mexican government source that the plant would be built in the city of Guadalajara,a city in western Mexico.

Once put into operation in 2025, the facility is expected to become the largest in the world for assembling motherboards and servers for Nvidia Blackwell-based servers. The plant's planned capacity is 20,000 servers, presumably per month, which means 240,000 machines per year. Blackwell is also known under the brand name GB200. The GB200 chip, a crucial component of the Blackwell platform, is driving the development of advanced AI applications across various sectors.

Foxconn is right now on the tack to ship servers based on Nvidia’s next-generation Blackwell GPU (graphics processing unit), an advanced AI superchip, in the fourth quarter of 2024, Liu said in an interview with CNBC. Asked about the order book for Blackwell servers, Liu said that demand for the chip is “much better than we thought.” Liu told Bloomberg Television that demand for GB200-based NVL72 servers with 72 B200 GPUs for AI and HPC workloads had exceeded initial expectations and was “crazy.”

Foxcoon also announced at Hon Hai Tech Day it is partner with Nvidia to build Taiwan’s largest supercomputer under a project called Hon Hai Kaohsiung Super Computing Center. The project will be built around Nvidia’s Blackwell architecture and feature the GB200 NVL72 platform, which includes a total of 64 racks and 4,608 Tensor Core GPUs. With an expected performance of over 90 exaflops of AI performance, the machine would easily be considered the fastest in Taiwan.

Foxconn plans to use the supercomputer, once operational, to power breakthroughs in cancer research, large language model development and smart city innovations, positioning Taiwan as a global leader in AI-driven industries.

Foxconn’s “three-platform strategy” focuses on smart manufacturing, smart cities and electric vehicles. The new supercomputer will play a pivotal role in supporting Foxconn’s ongoing efforts in digital twins, robotic automation and smart urban infrastructure, bringing AI-assisted services to urban areas like Kaohsiung.

Construction has started on the new supercomputer housed in Kaohsiung, Taiwan. The first phase is expected to be operational by mid-2025. Full deployment is targeted for 2026.

The GB200 NVL72 is a state-of-the-art data center platform optimized for AI and accelerated computing. Each rack features 36 NVIDIA Grace CPUs and 72 Nvidia Blackwell GPUs connected via Nvidia’s NVLink technology, delivering 130TB/s of bandwidth.

Foxcoon last Saturday reported highest-ever quarterly revenue for the three-month period ended September, highlighting strong demand for AI servers. Revenue soared 20.2% year-over-year (YoY) to 1.85 trillion Taiwan dollars, or T$1.85 trillion (US$57.3 billion), beating analysts’ expectation of T$1.79 trillion. “The result exceeded the company’s original expectations of significant growth,” the iPhone assembler said.