Powering the Next Generation of Scientific Discovery

We have reached extreme scale in our measurements of computation and our data sets, now marked by an increasingly familiar metric of “exa-” short for exascale (noted as 10^18 for a billion billion). For data, this means we have gone from counting in terabytes and petabytes – to exabytes. To put this into perspective, an exabyte of data would be equivalent to enough HD video for every one of the 8 billion people on Earth to stream Netflix continuously – every second of every day – for over 5 years. While that would be a truly impossible case of “binge watching”, it provides a frame of reference for the scale we have now reached.

In fact, new scientific instruments will easily generate this staggering volume of information annually - with facilities like these becoming more and more common. The next generation of research facilities like CERN's High-Luminosity Large Hadron Collider (HL-LHC) and the Square Kilometer Array Observatory (SKAO) will each produce an exabyte of data in a single year. This unprecedented scale of scientific data isn't just a technical milestone; it represents humanity's most ambitious attempt yet to unlock the fundamental mysteries of our universe, including the nature of dark matter.

To that end, today I'm proud to announce NextSilicon's selection to participate in ODISSEE (Online Data Intensive Solutions for Science in the Exabytes Era), a groundbreaking European Union project that brings together 14 leading partners across industry, research, and academia to tackle one of science's most pressing challenges: processing and analyzing unprecedented volumes of scientific data. This significant milestone positions NextSilicon as a key partner to Europe's leading research facilities, such as CERN and SKAO, delivering cutting-edge technology that powers Europe's most ambitious mega-projects in the field of science.

ODISSEE: A New Chapter in Scientific Computing

Our participation in ODISSEE represents more than just a technical collaboration—it's a commitment to advancing the frontiers of scientific discovery. The project's focus on developing innovative methodologies for on-the-fly AI data processing aligns perfectly with NextSilicon’s Maverick-2 intelligent compute accelerator (ICA) capabilities and our vision for the future of scientific computing.

Working alongside partners like CERN, SKAO, and leading European research institutions, we'll be contributing to the development of: Energy-efficient processing solutions for massive scientific datasets Flexible, open computing architectures that can adapt to emerging research needs Advanced capabilities to support the next generation of Exascale supercomputers

Among ODISSEE's most exciting aspects is its pilot program focusing on dark matter research. This fundamental mystery of our universe requires combining data from both SKAO and HL-LHC, setting new standards for data handling and processing. It's precisely the kind of challenge that Maverick-2 was built to address: complex calculations that demand both high- speed and high precision.

The Challenge: Next-Generation Science Demands Next-Generation Computing

The scale of data generated by modern scientific instruments is staggering. Traditional computing approaches simply cannot keep pace with these massive data streams, creating a real risk that potentially groundbreaking discoveries could go unnoticed. This isn't just a matter of storage—it's about our ability to process, analyze, and extract meaningful insights from this data in real-time.

The computational requirements push beyond conventional limits. Scientists must perform complex calculations with unwavering precision while processing data volumes that transcend the scale of today's largest content and video streaming platforms. This requires a new approach to computing—one that combines raw processing power with intelligence, adaptability, and efficiency.

Maverick-2: Purpose-Built for High-Performance Computing

The Maverick-2 Intelligent Compute Accelerator (ICA) was designed from the ground up to handle the high-precision, complex calculations that define modern scientific and commercial research. With its novel software-defined hardware architecture that eliminates the need for Domain Specific Languages (DSLs), Maverick-2 breaks from limiting software requirements of traditional accelerators; freeing researchers from vendor lock-in and allowing them to leverage existing application codebases without extensive rewrites.

The Maverick-2 ICA features an intelligent, self-optimizing architecture. Its advanced telemetry and patented algorithms dynamically adapt to workload demands in real-time, balancing power consumption, silicon utilization, and performance to achieve peak efficiency at extreme scale. In an era where computational demands and energy concerns are reaching unprecedented levels, Maverick-2 empowers commercial and scientific research institutions to achieve breakthroughs and insights faster while maximizing computational efficiency. This software-defined approach ensures computing infrastructure can both meet today's challenges and evolve to support tomorrow's discoveries.

Looking Ahead

The selection of NextSilicon for ODISSEE is truly an honor. It further validates our approach to develop the industry’s first intelligent compute accelerator that is focused on powering the most complex applications of scientific computing. Collaborating on these types of grand research projects inspires me and our team here at NextSilicon. It reinforces our commitment to pushing the boundaries of what's possible in high-performance computing and extreme scale data processing. As we work alongside Europe's leading research institutions and technology providers, we're not just hoping to solve today's computational challenges—we're helping to shape the future of scientific discovery.

The era of exascale science is in full swing, bringing with it opportunities and challenges that were unimaginable just a few years ago. Through ODISSEE and the capabilities of Maverick-2, we're ensuring that computational power keeps pace with scientific ambition, enabling researchers to extract meaningful insights from the ocean of new data being generated by modern scientific instruments.

Together with our ODISSEE project partners, we're building the foundation for the next generation of scientific breakthroughs. The journey toward understanding dark matter and other fundamental mysteries of our universe requires not just the right questions, but the right tools to find the answers. At NextSilicon, we're proud to be providing those tools to researchers and contributing to this historic endeavor.

Stay Connected

Stay tuned as we share more details in future blogs and announcements via our Nextletter.

About the Author:

Elad Raz is the founder and CEO of NextSilicon, a company pioneering a radically new approach to HPC architecture that drives the industry forward by solving its biggest, most fundamental problems.