ASUS shared success stories about comprehensive artificial intelligence solutions for emergency service and tumor diagnosis applications, as well as new server models. With the high demand for AI applications, ASUS provided solid case studies and proofs of concept from solution and hardware-oriented perspectives.
ASUS ESC N4A-E11 and ESC8000A-E11 for AI applications
ASUS, which has been trying to improve artificial intelligence workloads for many years with special technologies for performance tuning, root of trust for hardware security, BMC remote management solutions such as ASMB10-iKVM, has the best server portfolio for artificial intelligence applications. ASUS ESC N4A-E11, the first server based on NVIDIA HGX, takes full advantage of NVIDIA Ampere architecture. Offering up to twice the processing capacity of previous FP32 applications, this server provides this increased performance by balancing energy efficiency. Equipped with AMD EPYC™ 7003 processors, the ESC N4A-E11 meets the demanding demands of full-scale research facilities and data centers. ASUS also introduced the ESC8000A-E11, which supports up to eight PCIe® GPUs and has PCIe 4.0 storage technology. This model was developed to meet the ultra-high data transfer demands.
Powered by AMD EPYC processors, ASUS ESC4000A-E10 is an NVIDIA Certified server that can meet demanding demands. This model, which can be configured with the NVIDIA A100, A40, A30 and A6000 models, will also support the newly announced A2 GPUs. This model, which enables institutions to access the power required for demanding workloads such as deep learning inferences, can also be used for artificial intelligence applications in virtual environments with the NVIDIA AI Enterprise software package. The NVIDIA AI Enterprise software package includes AI workflow tools such as NVIDIA TensorRT for optimization and NVIDIA Triton™ Inference Server for model distribution. ASUS also offers NVIDIA Certified ESC4000-E10 and ESC4000-E10S models. These models, which have the Intel® platform, are supported by NVIDIA A100 80GB GPUs for artificial intelligence training and inferences. ASUS plans to make these systems compatible with NVIDIA A2 and A16 GPUs in the future.
ASUS AI solutions for healthcare workloads
ASUS also announced its current solutions for the healthcare industry, and explained it in a separate session. ASUS has integrated the NVIDIA Clara architecture into its servers to improve the overall efficiency of the emergency room. ASUS will work with InwinSTACK to develop AI-powered and carefully trained visualization models that will improve accuracy in brain tumor diagnosis. ASUS servers certified with InwinSTACK solutions will provide more comprehensive solutions in the healthcare industry in the future.
ASUS servers continue to deliver top-notch performance
ASUS servers are the center’s number one choice for computing performance and energy efficiency. ASUS 2P and 1P Intel and AMD servers SPEC at the top-class speed they reach. org holds more than 900 world records in the SPEC CPU 2017 tesdollar series. ASUS is also launching servers focused on more efficient energy consumption. These models also complete the SPECpower® green computing tesdollar series in 1st place on both Windows and Linux platforms, thanks to ASUS-exclusive energy-saving technology.