Super Computing News - Page 1
NVIDIA's new Hopper GPU architecture has been unveiled and with it comes some new DGX H100 systems powered by the monster new NVIDIA H100 GPU.
The new NVIDIA DGX H100 system has 8 x H100 GPUs per system, all connected as one gigantic insane GPU through 4th-Generation NVIDIA NVLink connectivity. This enables up to 32 petaflops at new FP8 precision, a gigantic 6x performance improvement over the previous-gen Ampere-based GPUs.
NVIDIA will be using the new Hopper-based DGX H100 systems as the "building blocks" of the next-gen NVIDIA DGX POD and NVIDIA DGX SuperPOD AI infrastructure platforms. NVIDIA's new DGX SuperPOD architecture has the new NVIDIA NVLink Switch System -- capable of up to 32 nodes and a total of 256 x H100 GPUs. At this level, we're talking about 1 exaflops of FP8 AI performance, also 6x more than its predecessor.
TSMC has just announced its N4X process technology, which has been tailor-made for the demanding workloads of high-performance computing (HPC) products.
The new N4X process technology is the first of TSMC's HPC-focused technology offerings, which the company says "representing ultimate performance and maximum clock frequencies in the 5-nanometer family. The "X" designation is reserved for TSMC technologies that are developed specifically for HPC products".
Taiwan Semiconductor Manufacturing Company (TSMC) enhanced its already impressive technology with features made for HPC products to create N4X, these features include:
- Device design and structures optimized for high drive current and maximum frequency
- Back-end metal stack optimization for high-performance designs
- Super high density metal-insulator-metal capacitors for robust power delivery under extreme performance loads
The National Renewable Energy Laboratory (NREL) has chosen Hewlett Packard Enterprise (HPE) to build its third-gen, high-performance computing (HPC) system, called Kestrel.
NREL's new Kestrel supercomputer was named after a falcon that has "keen eyesight and intelligence, Kestrel's moniker is apropos for its mission-to rapidly advance the U.S. Department of Energy's (DOE's) energy research and development (R&D) efforts to deliver transformative energy solutions to the entire United States".
The new supercomputer will be installed in the fall of 2022 at NREL's Energy Systems Integration Facility (ESIF) data center and will have a mind-boggling 44 petaflops of computing power. But the most interesting thing here is that NREL's new Kestrel supercomputer is powered by future next-gen Intel Xeon Scalable processors (Sapphire Rapids) and NVIDIA A100NEXT Tensor Core GPUs "to accelerate AI", not play Crysis.
AMD engineers have been working on something from the future it seems, with a new patent called "Look Ahead Teleportation for Reliable Computation in Multi-SIMD Quantum Processor".
The patent in question is for a system that would use quantum teleportation in order to boost a quantum computer's reliability, while at the same time reducing the number of qubits required for a given calculation. This "teleportation" technology would help solve scaling issues and calculation errors that arise from system instability.
One of the main issues behind quantum development is once you start pushing the pedal to the metal, there are major issues when it comes to scalability and stability. Quantum computing is far different to the 0s and 1s of traditional technology, so AMD's new teleportation patent is quite an important step towards solving that issue.
Pawsey Supercomputing Centre down in Perth, Australia chose Hewlett Packard Enterprise (HPE) for its next-gen Setonix supercomputer, but now we have some more details on the specs inside of the new supercomputer.
Setonix will have over 200,000 AMD EPYC "Milan" CPU cores, over 750 AMD Mi-Next GPUs with 128GB of VRAM per GPU, over 548TB of system memory, near-node NVMe storage, 15PB ClusterStor Lustre filesystem with 2.7PB SSD and 90PB of Ceph storage.
The additional details on Pawsey's next-gen Setonix supercomputer were provided by Pawsey CTO Ugo Varetto.
Hewlett Packard Enterprise (HPE) have just unveiled its next-gen LUMI supercomputer, which is powered by AMD's next-gen Zen 3-based EPYC processors and Radeon Instinct GPUs.
The new LUMI supercomputer will find its new home in Kajaani, Finland in 2021 -- and will be using the HPE Cray EX architecture to spin up 550 Petaflops of peak horsepower. The new LUMI supercomputer will be a part of EuroHPC's GPU-accelerated supercomputing platform powered by next-gen AMD CPUs and GPUs.
Forrest Norrod, senior vice president and general manager, data center and embedded systems group, AMD explains:
"AMD is proud to join with HPE to power the upcoming LUMI supercomputer to advance scientific research in artificial intelligence, weather forecasting, pharmaceutical discovery, and more. Our next-generation AMD EPYC CPUs and AMD Instinct GPUs, coupled with HPE's unique supercomputing technologies, are fueling new capabilities in high performance computing, and we are excited to strengthen the European research community through our support".
D-Wave Systems Inc. has just announced that it has expanded its Leap quantum cloud service to two new markets: India and Australia.
This move is a big deal for both India and Australia as it means developers, researchers, and businesses in those countries get access to something quite incredible -- and something not available in their respective countries: access to D-Wave 2000Q quantum computers.
D-Wave Systems recently opened up their D-Wave 2000Q quantum computers over the same Leap quantum cloud service to researchers and scientists to help out in the COVID-19 pandemic. This news now unlocks India and Australia researchers access to use the quantum computers for their tasks.
ARM is in the news for all the right reasons today, where first Apple announced plans that it would be transitioning away from Intel -- where it has been a partner for its processors for many years now, to ARM. The second, is that ARM chips now power the world's fastest supercomputer -- and that supercomputer, is insanely fast.
The new RIKEN Center for Computational Science's Fugaku supercomputer packs Fujitsu's new 48-core A64FX system-on-a-chip. It has 158,976 of these 48-core processors, meaning there is a mind boggling 7,299,072 processor cores powering the Fugaku supercomputer. Yeah, 7.29 million CPU cores -- you read that right.
This means the new Fugaku supercomputer is 2.8 times more powerful than the previous #1 supercomputer; Oak Ridge National Lab's Summit supercomputer. Summit has 2,414,592 processor cores with 148 petaflops of computer performance, compared to Fugaku and its game-changing ARM chips pumping out 415 petaflops.
AMD along with its technology partner in Penguin Computing Inc. which is a division of SMART Global Holdings, Inc. have announced that they are lending AMD-powered, high-performance computing (HPC) resources from the AMD HPC Fund for COVID-19 research.
The high-performance computing systems will be used by New York University (NYU), Massachusetts Institute of Technology (MIT), and Rice University. Others will follow, but these three were the first out of the gate from the AMD HPC Fund. AMD will also contribute a cloud-based system that is powered by AMD EPYC and AMD Radeon Instinct hardware on-site at Penguin Computing.
In total, AMD is donating a rather huge 7 petaflops of compute performance to the fight against COVID-19. AMD president and CEO, Dr. Lisa Su, said: "High performance computing technology plays a critical role in modern viral research, deepening our understanding of how specific viruses work and ultimately accelerating the development of potential therapeutics and vaccines".
NVIDIA launched its next generation Ampere GPU architecture during its online GTC 2020 keynote, with company founder and CEO Jensen Huang delivering the GPU Technology Conference 2020 keynote from his house.
The first GPU on the Ampere architecture announced was the new A100, which you can read all about right here -- as well as the new DGX A100 supercomputer, which you can read about here. The new DGX A100 supercomputer is an absolutely beast, rolling out as a third-generation AI supercomputer from NVIDIA super-charged by the Ampere GPU architecture and ridiculously fast A100 GPU.
NVIDIA's new DGX A100 supercomputer is now helping the fight against COVID-19, with Rick Stevens, associate laboratory director for Computing, Environment and Life Sciences at Argonne, explaining: "We're using America's most powerful supercomputers in the fight against COVID-19, running AI models and simulations on the latest technology available, like the NVIDIA DGX A100".