Find me at (links open in a new window or tab):

PostDoc, MSc and PhD opportunities

I currently do not have a funded project; watch this space for future opportunities. In the meantime, check

the university web site for postgrad and postdoc options
as well as the National Research Foundation (NRF)

Start looking at these early in the year, as some have application deadlines long ahead of the usual February start of academic year. Should you wish to apply for funding to work with me, let me know, and I will work through with you whether this is a good idea and, if so, advise on how to go about it. Contact me at p.machanick AT-nospam ru.ac.za. Not that research degrees can start any time of year; timing is only an issue in terms of securing funding.

My Research Interests

Watch this space for more detail

My research is based on the observation that Moore’s Law can be interpreted in two ways:

hot – if you have a fixed budget and wait a while, what hotter, faster technology can you buy?
cool – if want do something that’s currently too expensive, how long do you have to wait until it’s affordable?

In the cool space, the Raspberry Pi represents a new class of computer that is about as powerful as a PC of 10 years ago, but a fraction of the price. What can we do with Raspberry Pi? Aside from running a reasonably standard Linux distribution, it has a range of hardware interfaces to which we can add devices. Possible projects include hardware add-ons, investigating applicability of the original design goal of introducing children as young as eight to programming, and using Raspberry Pi in rural applications where cost and power consumption are major concerns.

In the hot space, I am investigating making GPUs more general purpose. General-purpose use of GPUs (GPGPU) starts from GPU designs primarily targeting graphics but with extensions to support general-purpose coding. I propose starting from the other end: a multicore general-purpose design with extensions to support graphics. The idea is to end up with a design with more accessible parallelism for general code than that of GPUs.

A hot new idea on which I plan to build is RISC-V, which aims to take the RISC movement back to its roots: really simple designs that are easy to implement in hardware. RISC-V provides a platform on which to build extensions of a simple, pure RISC instruction set that should be easy to scale up to high speed.

Also, bioinformatics is a burgeoning field in which we can tackle increasingly interesting tasks as costs decline. My current project in investigating population regulomics, a new field made possible by the growing number of human genomes. The general idea: investigate regulatory differences within and across populations.

Source WikiPedia
The hot view of Moore’s Law	The cool view of Moore’s Law

source NHGRI
Gene sequencing progress by comparison.

For more see my publications.