There is an increasing need for research at the intersection of computer science and electronics. One aspect is industrial pull, where partners increasingly want to explore the spaces between traditional hardware versus software-oriented products. Further, blue-sky academic research relies increasingly on the ability to work across the complete spectrum from computational hardware to high-level software. Electronics and Computer Science (ECS) is committed to attracting researchers who can operate within the wide field of computer engineering. We see this as strategic in terms of research portfolio, teaching, and in-house expertise for external collaboration. We are looking to recruit applicants who have a vision of how computational hardware and software will co-evolve. Our vision is to move beyond the traditional constraints of hardware and software, as vital enabling technologies offer versatile, cheap resource bases to build completely novel compute systems.
A residual common theme of these compute systems, aside from the sheer size is the importance of the network interconnecting the processor. The cost of moving data exceeds that of processing it, and it is not reasonably possible to separate the core from the communications: the “computer” is the network. This has profound implications for the future of both research and teaching in this area. You can no longer develop software, however elegantly crafted and rigorously designed, and subsequently port it to your choice of architecture. Software design at the most intimate level is now a function of the (class of) target hardware architectures.
The big wins of the next decade will come from the intimate synergy of new architectural models (networks and distributed systems) and specific language types, for example GPGPUs and CUDA, TPUs and TensorFlow, and event-based architectures. Southampton will be in the vanguard of this activity