HIPER Global at STFC Rutherford
In mid-June, we visited Rutherford Appleton Laboratory, joining the team at Sarsen Technology for our first Science and Technology Exhibition of the year. RAL is home to pioneering work across particle physics, scientific computing, laser technology and space research – the kind of place where some of the country’s finest minds tackle problems at the very edge of what’s measurable.
We sent some of our finest minds to connect with many of the 7500 scientists and engineers on site, representing more than 250 organisations that work within the £3bn Harwell Campus. Over a focused few hours, that curiosity did exactly what we hoped: it brought a steady stream of researchers and engineers to the table, many of them working at the nearby particle accelerator facilities, and most of them keen to talk in real detail about the hardware behind their experiments.
Where deterministic compute meets big science
One theme ran through almost every conversation: how to capture, process and act on data the instant it arrives, especially with regards to FPGA technologies. Radio Frequency (RF) control systems and beamline timing both demand deterministic, high-speed processing with nanosecond-level resolution – performance that general-purpose computing simply can’t guarantee. It’s a problem space where FPGAs have quietly become indispensable, and it set the tone for much of the day.
FPGA: precision at the beamline
Sarsen Technology has long been a specialist in FPGA solutions and while the SOM modules featuring the AMD Zynq from DAVE Embedded were able to satisfy many queries, the high–end products from Annapolis Micro Systems featuring the Altera Agilix 9 with high speed ADC converters were definitely catching the eye also. Their high-end platforms were a genuine draw for anyone working with high-bandwidth signals in real time. For RF, timing and data-acquisition applications, that combination of raw throughput and precision is hard to beat.
Edge AI and high-performance compute
For those wishing to use GPU acceleration for AI as well as simple number crunching, the desktop GPU was a highlight. As was the centre piece on our table, the VersaLogic Sabertooth AI. This deceptively compact single-board computer pairs an Intel Xeon-E CPU with an NVIDIA RTX 2000 GPU and dual 10GbE – a remarkable amount of capability in a footprint that opens up deployment options well outside the data centre. It was a reliable conversation-starter, and a great illustration of just how much compute can now be packed into a single board.
From research to real-world deployment
Days like this one are a reminder of how much common ground there is between cutting-edge science and the compute platforms that support it. Whether the challenge is deterministic timing, real-time signal processing or deployable AI at the edge, the goal is the same: getting the right processing power to exactly where it’s needed.
At HIPER Global and Sarsen Technology, that’s what we do – from board-level technologies through to full rack-scale infrastructure, backed by deep engineering expertise and a robust supply chain. If you’d like to pick up where we left off at RAL, we’d be glad to help.
📞 0118 929 4990
