Powerful AI-Augmented HPC Simulations Now Possible With Atos's BullSequana Hybrid Supercomputer

Powerful AI-Augmented HPC Simulations Now Possible With Atos's BullSequana Hybrid Supercomputer

Atos launched its most efficient supercomputer the BullSequana XH2000 at SC'18, the High-Performance Annual Conference. The BullSequana XH2000 is a hybrid supercomputer and designed to accommodate and orchestrate modern workloads between on-premises, public and private Cloud environments.

AI-augmented HPC simulation

The BullSequana XH2000 designed for and capable of running converged workloads (HPC simulation, Artificial Intelligence, Deep Learning) simultaneously on the same one machine, at the same time. According to an announcement, the simultaneous workload will help organizations to overcome the limits of traditional simulations and explore new applications, accelerating business transformation and innovation.

Atos Logo

More power to enable AI-augmented HPC simulations

As workloads expand to include AI, businesses need more computing power, and the BullSequana XH2000 provides this in a system that supports the latest CPU and GPU processor and accelerator architectures, including Marvell ThunderX2® Arm®, NVIDIA Volta V100, Intel® Xeon® processors and the latest AMD® EPYC Rome, for ultimate power, speed, and accuracy. This is the first Atos supercomputer that supports AMD processors. The next generation AMD EPYC processor, codenamed “Rome”, is expected to be available in Atos’ new BullSequana XH2000 liquid-cooled supercomputer by 2019. The BullSequana XH2000 allows a choice of system interconnect technologies, including InfiniBand HDR and BXI (Bull eXascale Interconnect) and Fast Ethernet.

Energy efficient

The BullSequana XH2000 consumes 10 times less energy than current systems. It is 100% water-cooled. It uses Atos' patented DLC (Direct Liquid Cooling) solution that minimizes global energy consumption by using warm water up to 40°C.

The Open Sequana architecture allows system upgrade and fully compatible with future computing glades and interconnect technologies, allowing to scale and making it futureproof. The BullSequana XH2000 is exascale-ready and capable of processing an exaflop (a billion billion operations per second) by 2020.

Executive Opinion

Chief Operating Officer, Big Data & Security at Atos, Pierre Barnabé, said, "With this announcement, the Atos Group confirms its strategic commitment to providing high-computing power to research centers and enterprises to carry out their research projects and to support their need for innovation."

Vice President HPC & Quantum at Atos, Agnès Boudot, said, “Atos is excited to reinforce the partnership with AMD with the planned launch of an EPYC Rome blade for our new BullSequana X2410 next year. The HPC market was waiting for a new compute engine, offering higher memory bandwidth to increase application performance efficiency. The EPYC family will greatly complement our solution’s portfolio.”

Corporate vice president and general manager, AMD datacenter products, Scott Aylor, said, “AMD welcomes the announcement by Atos of their upcoming BullSequana X supercomputing system with AMD EPYC processors. As a global leader in high-performance computing, Atos is perfectly suited to deliver on the promise of the EPYC product line and the industry’s first 7nm datacenter CPU coming to market from AMD next year.”

Availability of The BullSequana XH2000 will be from February 2019.

PC:pablo,pixabay

Note: We at TechSutram take our ethics very seriously. More information about it can be found here.
Mandar Pise Opinions expressed by techsutram contributors are their own. More details

Mandar is a seasoned software professional for more than a decade. He is Cloud, AI, IoT, Blockchain and Fintech enthusiast. He writes to benefit others from his experiences. His overall goal is to help people learn about the Cloud, AI, IoT, Blockchain and Fintech and the effects they will have economically and socially in the future.

No comments:

Post a Comment

    Your valuable comments are welcome. (Moderated)