Data storage strategies for data intensive scientific research

Scientific computing teams are confronting surging data volumes from Artificial Intelligence, simulation and genomics, forcing a rethinking of storage architectures across on premises and cloud environments. An expert roundtable report profiles how leading research institutions are balancing performance, scalability, security and cost while planning for future workloads.

Scientific research infrastructures are facing a rapid escalation in data volumes generated by Artificial Intelligence and machine learning, large scale simulations and genomics, which is forcing research teams to manage petabytes of data while maintaining performance, security and cost efficiency. Storage strategy decisions are becoming more complex as organisations weigh high performance NVMe, scalable tiered storage and cloud based options, and must also consider how Artificial Intelligence workloads will reshape infrastructure requirements. Research computing leaders are increasingly focused on aligning storage capabilities with high performance computing environments so that data intensive workflows are not constrained by bottlenecks.

An expert roundtable report brings together high performance computing leaders from organisations including the University of Cambridge, Lund University and Simula Research Laboratory to describe practical approaches to modern storage design. The report details how leading research institutions design tiered storage architectures that can match different performance and capacity needs, and it highlights the key factors influencing infrastructure choices, such as workload type, budget constraints and long term retention requirements. Contributors explain how Artificial Intelligence and GPU workloads are changing storage requirements, prompting demand for faster access to active data as well as scalable systems that can accommodate rapid growth without sacrificing reliability.

The roundtable also explores why many research organisations remain cautious about cloud storage despite its flexibility, citing concerns over cost predictability, data sovereignty and performance for large scale scientific workloads. The report is positioned as a white paper for high performance computing and research computing leaders, data centre architects, research IT and infrastructure teams, Artificial Intelligence and data platform specialists, and university and laboratory technology managers who support data intensive environments. Readers are promised peer insights into how research organisations are preparing infrastructure for the next wave of data driven discovery, including perspectives on what experts believe the future of research data storage will look like.

50

Impact Score

Business anniversary meets Artificial Intelligence arrival

A reflection on nine years of building a digital local business newsroom turns into a broader assessment of how generative Artificial Intelligence is already reshaping professional communication and editorial work. The piece balances optimism about productivity gains with concern over overreliance on automated thinking.

Penguin Solutions launches cxl-based kv cache server

Penguin Solutions introduced a production-ready KV cache server built on CXL memory technology for enterprise-scale inference and agentic Artificial Intelligence workloads. The system is positioned to ease memory bottlenecks, improve GPU cluster efficiency, and reduce latency.

Micron completes Tongluo P5 site acquisition in Taiwan

Micron has completed its acquisition of PSMC’s P5 site in Tongluo, Miaoli County, Taiwan. The facility adds cleanroom capacity near Micron’s Taichung campus and is intended to support more leading-edge DRAM and HBM output for growing Artificial Intelligence-driven demand.

Qwen3.5 recipes shared for Jetson Thor

A Jetson Thor forum post shares setup recipes for running multiple Qwen3.5 models with NVIDIA’s latest vllm repository for Thor. The largest reported working model is Qwen3.5-122B-A10B, with notes on NVFP4 and INT4 tradeoffs.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.