Ephemeral Computing and Bioinspired Optimization:

Challenges and Opportunities

Cotta, Fernández-Leiva, Fernández de Vega, Chávez, Merelo, Castillo, Camacho, Bello-Orgaz

Presented by JJ Merelo / @jjmerelo / @ephemech

... The use and exploitation of computing devices whose availability is ephemeral in order to carry out complex computational tasks.

Computers + smartphones + cloud + embedded oh my!

Asynchronous, heterogeneous, fleeting

Flexible

Resilient

Self-Adaptive

Decentralized

Issues:

Algorithms

Big data

Energy consumption

EphC secret origins:

Ubiquitous computing

Tries to hide ephemeral substrate

On the other hand, in ubiquitous computing the main goal is to leverage computation everywhere and anywhere, so that computation can occur using any kind of device, in any location, starting and ending at any time and using any format and during any amount of time. The main efforts in this area have been ori- ented to design and develop the underlying technolo- gies needed to support ubiquitous computing (Lyyti- nen and Yoo, 2002) (like advanced middleware, op- erating systems, mobile code, sensors, microproces- sors, new I/O and user interfaces, networks or mo- bile protocols). However, and in the same way it hap- pens with cloud computing, the main target in ubiqui- tous computing is to allow stable and persistent com- putation processes perform a complete execution of the programs. When this area handles the concept of ephemeral devices, services or computation, the main solution is to stop the process, or processes, and re- sume once new devices are ready (Wang et al., 2004). Previous hypothesis and assumptions can be extrapo- lated to distributed computing, where the concept of ephemeral services can be a problem that could even- tually generate a failure in the execution of the process (Sharmin et al., 2005).