IT/Datacenter & Super Computing

NVIDIA has been forming its quantum computing team, something that will see an alliance between American and Taiwanese companies.

In a new report from UDN spotted by Dan Nystedt on X, we're hearing that NVIDIA will be including Quanta Computing, Compal Electronics, and Supermicro for its new quantum computing plans. Quanta will be taking care of hardware testing and validation using NVIDIA's CUDA-Q, while Compal will be performing quantum optimization using CUDA-Q, leaving Supermicro to be involved with research.

NVIDIA CEO Jensen Huang said earlier this year at CES 2025 that useful quantum computers were 15-30 years away, but then just a couple of months later at GC's first-ever "Quantum Day", Jensen walked back his earlier statements saying that "quantum computing has great potential".

Underwater data centers are not a new idea - neither are 'floating cities' of compute clusters, for that matter - but the scope of a growing initiative over in China is certainly impressive.

Global Times (via TechRadar Pro) reports that a new data center capsule - measuring 18 meters long, and 3.6 meters in diameter - has been launched in Lingshui (a county in Hainan Province, China).

The new capsule is linked to an existing underwater data center, with this project in Hainan first kicked off by the data center firm Highlander back in 2023 (as the first commercial effort of its kind).

Microsoft has just unveiled its new Majorana 1 quantum computing chip, with the company creating a new "state of matter" in the process. Check it out:

The company says the design of the Majorana 1 processor is meant to fit up to 1 million qubits, which should be enough quantum computing power to see what the world of quantum can truly do: including cracking cryptographic codes, designing new drugs and materials, and so much more.

Microsoft announced the creation of the first-ever "topological qubits" in a device that stores information in an exotic state of matter, which could be the key to significant breakthroughs in quantum computing.

Fujitsu has just shown off a sample of its new Monaka processor: a new Armv9-based 144-core chip that is destined for data centers of the future.

Last week, the Japanese company announced it was working with Broadcom to use its new 3.5D eXtreme Dimension System in Package (3.5D XDSiP) platform, and now we're being introduced to Fujitsu's new Monaka processor.

The new Monaka chip is a large CoWoS system-in-package (SiP) that has 4 x 36-core compute chiplets, made on TSMC's N2 process node with 144 Armv9-based cores with enhancements that are stacked on top of SRAM tiles, in a face-to-face (F2F) layout, using hybrid copper bonding (HCB).

Google has just introduced its state-of-the-art quantum chip dubbed Willow, with Willow performing a standard benchmark computation in under 5 minutes, something that would take one of the world's fastest supercomputers 10 septillion years or 10,000,000,000,000,000,000,000,000 years... or longer than the universe has existed.

Google Quantum AI Founder and Lead, Hartmut Neven, wrote an extensive blog explaining the new Willow quantum chip and its journey beginning over 10 years ago when Neven founded Google Quantum AI back in 2012.

The vision was to build a useful, large-scale quantum computer that could use quantum mechanics -- the "operating system" of nature to the extent we know it today -- in order to benefit society by advancing scientific discovery, developing helpful applications, and solving some of the world's biggest problems.

AMD has announced its leadership in the high-performance computing (HPC) space with the new El Capitan supercomputer online, recognized as the fastest supercomputer in the world, and it's powered by AMD hardware.

The new El Capitan supercomputer is housed at the Lawrence Livermore National Laboratory (LLNL) and is powered by AMD Instinct MI300A APUs, with Hewlett Packard Enterprise (HPE) building the system. El Capitan hit a High-Performance Linpack score of 1.742 exaflops of supercomputing performance.

Forrest Norrod, executive vice president and general manager, AMD said: "We are thrilled to see El Capitan become the second AMD powered supercomputer to break the exaflop barrier and become the fastest supercomputer in the world. Showcasing the incredible performance and efficiency of the AMD Instinct MI300 APUs, this groundbreaking machine is a testament to the dedicated work between AMD, LLNL and HPE. At AMD, we are driving the future of computing with leadership performance and capabilities that will continue to define the convergence of HPC and AI for years to come".

We all thought banking and military security was impenetrable? That was, until Chinese researchers used a D-Wave quantum computer to execute what they claim as the first successful quantum attack on widely used encyption algorithms which pose a "real and substantial threat" to the banking and military sectors.

In a new report by the South China Morning Post picked up by The Quantum Insider, which I saw posted on X which you can see embedded above, the Chinese quantum computer has cracked multiple SPN algorithms, which is "way bigger than cracking the ENIGMA code, the future of encryption is at stake" we're warned.

We have thought that the digital world is secure, especially in a during and post-COVID world where in many countries across the world they pushed hard into the all-digital, all-cloud life. Well, now we have Chinese scientists led by "brainiac" Wang Chao from Shanghai University has "just pulled off a stunt that should have the NSA, the White House, and every crypto bro out there sweating bullets" says Huo on X.

Elon Musk recently flicked the 'power on' switch to his new Memphis Supercluster, which is the world's fastest AI data center, with 100,000 liquid-cooled NVIDIA H100 AI GPUs, but just how much power does this AI data center need? Lots, lots and lots of power.

In a new X post by AI and semiconductor analyst who runs SemiAnalysis, Dylan Patel, ruminated on X that it would be impossible for Elon Musk to run his Memphis Supercluster because he doesn't have enough power. He said that only 7MW is being drawn from the grid, which is enough power to run just 4000 of the H100 AI GPUs.

Patel is aware that the Tennessee Valley Authority (TVA) will deliver an additional 50MW to the Memphis Supercluster if xAI signs a pending deal, but that would be by August 1 at the very earliest. Patel also noted that the 150MW substation on the xAI site is still under construction and won't be complete by Q4 2024.

HPE was showing off a server blade from its upcoming El Capitan supercomputer at the recent ISC High Performance event in Hamburg, Germany.

The server blade had its front cover removed, showing off the internals that include AMD's new Instinct MI300A AI accelerator. The blade is called the HPE Cray Supercomputing EX255a accelerator blade, featuring a single-slot 1U blade chassis. It might be small, but it packs an incredible 8 x MI300A chips, with copper cooling blocks and copper cooling pipes linking everything together, rather beautifully I might add.

Each of the blades uses liquid cooling to keep the crazy amounts of heat generated by the 8 x AMD Instinct MI300A chips, with each of the MI300A APUs featuring a 550W TDP, and a peak power rating of 760W. This means that the cooling for the blades needs to be able to handle an incredible 4400W of power on average, and peak power of up to an even higher 6080W.

The UK government has just announced it's working on the most powerful supercomputer it's ever built, which will be known as Isambard-AI, which, when it's built, will be one of the most powerful AI supercomputers in the world.

Isambard-AI will be capable of over 200 quadrillion calculations per second once it's powered on in summer 2024, where it will be housed at the University of Bristol. Hewlett Packard Enterprise is building the new AI supercomputer, which will be powered by 5448 x NVIDIA GH200 Grace Hopper Superchips, capable of over 21 exaflops of AI performance and over 21 quintillion floating point operations per second for AI applications like large language models (LLMs).

The University of Bristol says that the new Isambard-AI supercomputer will be 10x faster than the fastest AI supercomputer that's available in the UK today. Isambard-AI will also get connected to other supercomputers that are based in the UK, which will see performance increase even more. Simon McIntosh-Smith from the Univesity of Bristol said that Isambard-AI "will be one of the most powerful AI systems for open science anywhere".

TSMC is working directly with NVIDIA and Broadcom to super-speed R&D efforts into the next step in the semiconductor industry: silicon photonics.

First, what are silicon photonics? The semiconductor industry is moving into the world of silicon photonics as its next-gen approach to traditional copper transmission cables, which sees the combination of laser and silicon technology that pumps high data transfer speeds. Silicon photonics will have conventional "electricity" delivery with light, meaning we're in for a world of seamless, quicker, secure transmission.

ChinaTimes reports that Taiwan Semiconductor Manufacturing Company -- TSMC -- pumping 200 new R&D specialists into the arms of silicon photonics use in the AI industry. This is why TSMC is teaming with NVIDIA and Broadcam, as the Taiwanese giant is currently walking into Compact Universal Photo Engine (COUPE) that will usher in an energy-efficient solution that will be the big push into its large-scale adoption. GlobalFoundries is also working on silicon photonics, but TSMC is ahead of GlobalFoundries... which shouldn't surprise anyone who knows how much further ahead TSMC is above the competition.

Microsoft is hiring for a new role at the company, Principal Program Manager of Nuclear Technology. Microsoft is going nuclear, with the company planning to implement a global Small Modular Reactor (SMR) and microreactor energy strategy.

With the rise of AI and the exponential increase in data center power requirements, the company is looking to the controversial nuclear power for its Cloud and AI business.

"The ideal candidate will have experience in the energy industry and a deep understanding of nuclear technologies and regulatory affairs," the job listing writes. It's worth noting that this will not involve Microsoft building massive nuclear power reactors all over the globe but will use specialized microreactors that are compact, transportable, and "plug-and-play."

Microsoft has released a scientific-sounding roadmap for building its first quantum supercomputer, built on the research and advances it has made with topological qubits in recent years. Even though more work is needed, Krysta Svore, Microsoft's VP of advanced quantum development, has told TechCrunch that it's on track to build a functional quantum supercomputer in less than 10 years.

"We think about our roadmap and the time to the quantum supercomputer in terms of years rather than decades," Krysta Svore said. It's an ambitious roadmap for sure, and quantum computing is an exciting prospect when it comes to high-performance computing. Quantum computers use qubits, capable of running multidimensional algorithms, and the real kicker is that their power increases exponentially as more qubits are added. Compared to the linear progression of standard computing, it's no wonder it's seen as the next big frontier.

The issue is that quantum computing can have high error rates and needs to be kept extremely cold and free from even the slightest physical interference.

NVIDIA's nifty little Jetson Nano modules have been used by two students at the Southern Methodist University in Dallas, Texas to create a mini supercomputer.

The desktop supercomputer is powered by 16 x NVIDIA Jetson Nano modules, with the students building it into a clear acrylic case, including the power supplies required, cooling fans to get rid of the heat, and a network switch to connect it all together.

Southern Methodist University received outside funding to design the NVIDIA Jetson Nano-powered desktop supercomputer, which will be used to educate students about how a computer cluster works. The project took four months of learning and labor for the design and creation, while the students were learning about networking.

The National Oceanic and Atmospheric Administration has just replaced two outdated supercomputers with new supercomputers that will bring massive upgrades to weather forecasting.

The National Oceanic and Atmospheric Administration (NOAA) announced back in 2020 that it would be replacing some of its older, now-outdated supercomputers that are used to run weather forecasts and predict patterns. NOAA has now replaced the older models, which were Cry and IBM supercomputers located in Reston, Virginia, Florida, and Orlando - with two Hewlett Packard Enterprise (HPE) Cray supercomputers.

These two supercomputers come with 2,560 AMD Epyc Rome 64-core 7742 server CPUs that collectively provide 327,680 cores that are capable of operating up to 12.1 petaflops. This new setup is approximately three times faster than what NOAA was previously using, and with the new upgraded power, the agency believes it will be able to provide more accurate and detailed weather forecasts to the public. Notably, the two new NOAA supercomputer systems are called Dogwood and Cactus, and rank as the 49th and 50th fastest supercomputers in the world.

The latest generation of Sunway supercomputer rivals the performance of Frontier, which was recently labelled the world's most powerful supercomputer.

According to the researchers behind the latest Sunway supercomputer, the Sunway has over 37 million CPU cores, quadrupling the number found in Frontier, nine petabytes of memory, and 96,000 semi-independent computer systems referred to as 'nodes,' that can exchange data at rates greater than 23 petabytes per second. The Sunway is capable of exascale computing, allegedly up to 5.3 exaFLOPS (5.3 quintillion floating-point operations per second).

The research team trained an artificial intelligence (AI) model, named bagualu (meaning alchemist's pot), with 174 trillion parameters using the Sunway. According to the South China Morning Post, this number rivals that of the number of synapses in the human brain, though some estimates of the true number of synapses go as high as 1,000 trillion.

Atos is building the new MareNostrum 5 supercomputer, with a new $160 million contract signed in part of the EuroHPC JU initiative, and once it's built it will be the fastest supercomputer in the European Union.

NVIDIA's new Grace Superchips will be powering the new Atos MareNostrum 5 supercomputer, which will be built in Spain and delivered to the Barcelona Supercomputing Centre (BSC) in 2023. Inside, the new MareNostrum 5 supercomputer will pack 314 petaflops of FP64 computing performance.

The compute and storage partitions will be operational within the year, reports HPC Wire, adding that the remainder of the MareNostrum 5 supercomputer to be "operational within the year". NVIDIA has also added that the system is "expected to enter deployment in 2023". BSC says that the new MareNostrum 5 supercomputer will be "fully powered with green energy, and will utilize heat reuse technology".

The new NREL Kestrel supercomputer details have been unveiled, with the National Renewable Energy Laboratory (NREL) division of the US Department of Energy (DOE) packing some serious horsepower into its new supercomputer.

We're looking at a Minajatwa of silicon between Intel Sapphire Rapids Xeon CPUs, AMD EPYC "Genoa" CPUs, and NVIDIA H100 GPU accelerators. We have 44 petaflops of peak compute performance, up from just 8 petaflops on the previous-gen Eagle supercomputer.

CPU upgrades are big: Intel Sapphire Rapids Xeon CPUs with 52 cores, 112 threads each -- up from the 18 cores and 36 threads from the Intel Xeon-Gold Skylake CPUs in the Eagle supercomputer.

It's a huge day for supercomputers, with AMD powering the world's fastest supercomputer -- the Oak Ridge National Laboratory (ORNL) "Frontier" supercomputer -- and now NVIDIA has announced it is powering the "Venado" supercomputer.

NVIDIA announced last week that Taiwan tech giants were preparing NVIDIA Grace CPU-powered servers, and now we have our first: the new Venado supercomputer that's being constructed by the Los Alamos National Laboratory. The new VENDAO supercomputer is capable of a huge 10 exaflops of peak AI performance.

It feels like a little bit of "me too" with NVIDIA's announcement of the Venado supercomputer, but more concrete details will be provided at the International Supercomputing Conference in Hamburg, Germany later today. The new AMD CPU + GPU-powered Frontier supercomputer was teased yesterday, and now we have the NVIDIA CPU + GPU-powered Venado supercomputer here today.

The massively powerful new Oak Ridge National Laboratory (ORNL) supercomputer dubbed "Frontier" has broken the 1.1 exaflops barrier, becoming the first machine in the world to breach the historic exascale barrier.

The Department of Energy (DOE) will operate the new Frontier supercomputer in Tennessee, USA, with the system costing up to $1.8 billion to build and is now the world's fastest supercomputer, overtaking the Fugaku supercomputer in Japan. ORNL's new supercomputer is powered by AMD 3rd Gen EPYC CPUs and AMD's newest Radeon Instinct MI250X GPUs.

Inside, we have 74 purpose-built HPE Cray EX supercomputer cabinets with 9408 AMD EPYC CPUs for a total of (a bonkers) 8,730,112 processing cores, and 37,632 AMD Instinct MI250X GPUs with a power efficiency rating of 52.23 gigaflops/watt. There's 700PB of data storage with peak write speeds of an insane 5TB/sec (5000GB/sec).