ASUS Announces ESC N4A-E11 and ESC8000A-E11 Servers at NVIDIA GTC

ASUS has been refining AI workloads for several years by delivering exclusive performance-tuning technologies, hardware security Root-of-Trust and BMC remote-management solutions such as the ASMB10-iKVM, resulting in a best-in-class server portfolio in AI applications. Today, ASUS is launching the ESC N4A-E11 , its first NVIDIA HGX-based server, which fully leverages the NVIDIA Ampere architecture to provide up to twice the compute capabilities of previous FP32 implementations while harmoniously balancing this increased performance with energy efficiency. Armed with AMD EPYC™ 7003 processors, the ESC N4A-E11 is targeted to meet the stringent requirements of full-scale research facilities and data centers. ASUS is also launching the ESC8000A-E11, which can support up to eight PCIe® GPUs and features PCIe 4.0 storage and networking technology, in response to the ever-growing demand for ultra-high speed data transmissions.
ASUS ESC4000A-E10 is a high-demanding NVIDIA-Certified server powered by AMD EPYC processors, featuring NVIDIA A100, A40, A30, A6000 and future planned support for the recently announced A2 GPUs. It enables enterprises to meet requirements for a variety of demanding workloads like deep-learning inference, while running the NVIDIA AI Enterprise software suite for AI in virtualized environments. The NVIDIA AI Enterprise software suite includes AI workflow tools such as NVIDIA TensorRT for optimization and NVIDIA Triton™ Inference Server for model deployment. ASUS also offers NVIDIA-Certified servers ESC4000-E10 and ESC4000-E10S, powered by an Intel® platform with support for NVIDIA A100 80GB GPUs for AI training and inferencing. ASUS plans to certify these systems with NVIDIA A2 and A16 GPUs in the future.

At GTC, ASUS will conduct two sessions about its recent healthcare and medical solutions. In the first talk, titled “Accelerating Your AI Journey, ASUS Cloud AI & Cloud Solution at NTUH Smart ED Project (A31742),” Peter Wu of ASUS Cloud and Chien-Hau Huang of National Taiwan University Hospital (NTUH) will discuss a complete AI solution for NTUH to improve patient flow in emergency services. ASUS integrated the NVIDIA Clara architecture with its servers to greatly improve overall efficiency in the emergency room. ASUS will also have Chien-Hau Huang, the chief clinical professor at NTUH, share his hands-on experience of this three-way collaborative model.
The second talk, titled “Clinical Development of Knowledge-Driven, Perspective-Based, AI-assisted Tumor Diagnosis” (A31737) and presented by InwinSTACK data scientist Mingta Tu, will cover topics around delivering AI applications for assisted medical treatment. ASUS will cover its collaboration with InwinSTACK to engineer carefully trained modeling of AI-driven visualizations that improved diagnostic accuracy for brain tumors. ASUS servers were validated by inwinSTACK’s ari solution and will be positioned to deliver more comprehensive medical solutions in the future.

ASUS servers are the top data center solution choices for compute performance and energy efficiency. ASUS 2P and 1P Intel and AMD servers continue to achieve top-ranked speed, with over 900 world records based on SPEC CPU® 2017 benchmarks on SPEC.org. ASUS also offers servers that are focused on more efficient energy consumption, and these have achieved the No. 1 ranking on the SPECpower® green computing benchmark on both Windows and Linux, thanks to ASUS-exclusive energy-saving technology.