読むのに 3 分
プレスリリース
SAN JOSE, Calif., July 30th, 2019 – Dr. Radoslav “Rado” Danilak, CEO of Tachyum Inc., will speak about the need for new approaches, and capabilities of data center architectures, interconnects, storage and memory types, to enable the industry to unleash the full power of AI in the multi-billion dollar hyperscale market, at the Flash Memory Summit in the Santa Clara Convention Center in Santa Clara, Calif. on Wednesday, August 7 at 4:35 p.m. PDT.
Danilak’s presentation, “Flash Solutions for the Skyrocketing AI/ML Applications Market,” addresses how storage designers must address an entirely new series of challenges and problems to overcome the massive and growing demand for AI/ML applications across all industries. AI/ML programs require huge data sets and present increasingly challenging bandwidth and latency needs. Danilak’s company Tachyum is leading the charge in developing a new class of high performance / ultralow power universal processors capable of handling normal data center workloads as well as the needs of emerging AI/ML applications with a single unified hardware architecture. The new bottleneck in Data Center architectures is now becoming the data processing. Prodigy will solve this bottleneck by enabling data center flash storage to be utilized more efficiently across a broad spectrum of AI workloads. The impact of Danilak’s new universal processor architecture on Flash memory markets is expected to be substantial.
Tachyum’s Prodigy Universal Processor Chip is the smallest and fastest general purpose, 64-core processor developed to date, requiring 10x less processor power and reducing processor cost by 3x. Prodigy will directly enable a 32-Tensor Exaflop supercomputer allow to build machines more powerful than the human brain by 2021, years ahead of industry expectations. Prodigy reduces data center TCO (annual total cost of ownership) by 4x, through its disruptive processor architecture and a smart compiler that has made many parts of the hardware found in typical processors redundant. Fewer transistors, fewer and shorter wires, due to a smaller, simpler core, translates into much greater speed and power efficiency for the Prodigy processor.
“The world is at an inflection point where the needs of next-generation applications must be met with next-generation flash solutions capable of delivering the performance needs of data center, AI and HPC workloads,” said Danilak. “I look forward to sharing with Flash Memory Summit attendees how these needs can best be accomplished through an approach that will replace existing chips provisioned in hyperscale data centers to significantly improve performance, energy consumption, server utilization and space requirements.”
FMS is the world’s premier flash memory conference and exposition. Now in its 14th year, FMS features the latest technology trends, the most innovative products, and the broadest coverage of this rapidly expanding market with demonstrations, speaking engagements and panels covering the latest storage technologies, such as NVMe, NVMe-oF, persistent memory, computational storage, 3D flash, 3D XPoint, neuromorphic computing, and new non-volatile memory. Additional information about the event is available at https://www.flashmemorysummit.com/