News facts:

  • PRIMERGY GX2460 M1 expands Fujitsu’s range of servers to power artificial intelligence (AI) and other GPU-accelerated workloads
  • GPU-accelerated computing accessible to a wide range of customers thanks to an excellent price-performance ratio for the all-new dual-socket rack server
  • High-speed, power-packed performance optimized for advanced workloads such as AI, data science and deep learning, enabling organizations to harness the power of data

Fujitsu is fuelling global data-driven transformation by making NVIDIA GPU-accelerated computing for advanced AI workloads accessible to mainstream customers. The expansion of its NVIDIA GPU-driven PRIMERGY GX server portfolio delivers high-speed performance for advanced processing tasks both today and into the future.

Most data collected and stored by organizations currently goes unused. But NVIDIA GPU-accelerated computing is set to change this. According to analyst firm Gartner, the server GPU market continues to expand at more than 20% CAGR1. Enterprises are increasingly using GPUs in data centers to significantly accelerate workloads, including AI and data science, deep learning, data analytics, virtual desktop infrastructure (VDI) and graphics.

Explicitly designed for GPU-accelerated workloads, the new Fujitsu Server PRIMERGY GX2460 M1 optimizes both compute capability plus NVIDIA GPU operation and data transfer, leveraging the latest processor technology with advanced AMD processors and high-speed PCIe Gen4.0 lanes. This brings a considerable boost in server performance for data-intensive workloads. Organizations can create a competitive advantage by unlocking more insights and deriving faster value from their data, allowing them to transform customer experiences, open up new markets, and make employees and processes more productive.

Excellent price-performance for advanced enterprise workloads

Offering an excellent price-performance ratio for a wide range of advanced enterprise workloads, the dual-socket rack server PRIMERGY GX2460 M1 features second-generation AMD EPYC series CPU processors up to 32 cores, 1TB DDR4 memory and up to four NVIDIA data center GPUs. Eight NVMe/SATA drives guarantee lightning-fast storage for large amounts of data, while up to six2 high-speed PCIe Gen4.0 interfaces support GPUs, storage and networking card options for flexible expandability. With its compact form factor, the versatile server is designed for energy efficiency, ensuring lower total cost of ownership. It is easily managed alongside other data center infrastructure via the Fujitsu Software Infrastructure Manager (ISM).

Oliver Delachapelle, Head of Category Management Products Europe at Fujitsu, says: “Data is the key business enabler in today’s data-centric world, for all industries, regardless of company size. As information continues to grow exponentially, organizations need to move it faster, store more of it and process it across edge, data center and cloud. GPU-accelerated computing is the platform they need to do this intelligently. That’s why we have expanded the PRIMERGY portfolio with the introduction of the PRIMERGY GX2460 M1 – an all-new, state-of-the-art server category that gives customers an ideal solution for all their AI, data science, graphics, high-performance computing and VDI workloads. By leveraging increased computing power, businesses will be able to learn more from their data, unlock new opportunities and follow through on their most ambitious data-driven growth strategies.”

Pricing and availability

The PRIMERGY GX2460 M1 is available to order from November 2020, either directly from Fujitsu or via Fujitsu SELECT channel partners.

Notes to editors
1. Forecast Analysis: Discrete GPUs, Worldwide by Gartner, published July 13, 2020.
2. The maximum number depends on the number and type of GPUs installed.

Download PDF