Avnet Silica’s market segment manager for aerospace and defence, Paul Leys, shows how FPGA technology is aligning with the New Space phenomenon
The New Space phenomenon is shaking up the approach to deploying hardware into space. Access to easily available hardware and software platforms has played a big part in this, alongside migration from the need to use custom-built military-grade components towards commercial-of-the-shelf (COTS) options.
Online communities are encouraging engineers to swap ideas. Also, ride sharing is helping spread launch costs. Consequently, the sector’s barriers to entry have been substantially lowered. It is now possible for SMEs or university research groups to conduct space projects that would have been completely out of reach a decade ago.
The New Space philosophy is proving effective regarding CubeSats, PocketQubes and other forms of nanosatellites intended for low-earth orbit (LEO) applications. These items are relatively inexpensive to construct, generally weighing <10kg (some <1kg), taking up minimal rocket payload.
Regarding on-board computing resources, programmable logic has greater appeal than hardwired custom ASICs, since FPGAs better align with tight project deadlines and budget constraints. FPGA usage also enables a platform approach to space hardware design. Functionality incorporated into the FPGA for an initial nanosatellite implementation is straightforward to alter. Further functionality can be added to future generations without expending extra engineering effort or incurring the unwanted costs of starting each project from scratch.
To cope with radiation exposure, selected FPGAs must have strong total ionising dose (TID) figures. Resilience to single-event upsets (SEUs) and single-event latch-ups (SELs) is also mandated.
Different underlying technologies on which programmable logic devices are produced each have their own respective properties. Antifuse is effective at handling radiation, with strong SEU/SEL resilience. However, there are limitations on the number of gates it can support, suiting only small-scale computing requirements. When greater processing capacity is required, flash or SRAM need to be considered.
Using SoI process technology—rather than bulk CMOS—in conjunction with very thin epitaxial layers, certain SRAM-based FPGAs are now able to achieve better protection against SEU/SEL incidents than conventional SRAM equivalents.
For lower price tags, radiation-tolerant FPGAs are available. Being derived from COTS parts, they also leverage more advanced processes than the radiation-hardened devices mentioned above—leading to elevated gate counts. Radiation-tolerant FPGAs are now available on 7nm architectures. This allows dramatic increases in the number of logic cells and enables integration of new concepts such as AI Engines and Network on Chip.
The latest FPGAs have a pivotal role to play in the democratisation of space and Avnet Silica is a trusted source for this technology.
www.avnet-silica.com