Light-speed functions in non-electric computers? (with a perhaps laughable aside on 'faster-than-light' processes)

Of course, historically the mechanical computers like Zuse's Z1 (c1939) and electromechanical computers like the first differential analyzer at MIT (c1930) could nowhere near approach the calculating speeds of even primitive electronic computers like ENIAC (1945) or EDVAC.

Can non-electric computing achieve such speeds?

How do we achieve this? Light run along fiberoptic cables is the obvious way, perhaps with short-term memory on a photosensitive medium (eg. a fluorescent substance).

Murad (2000) (see below), writes, '...pseudo-fluid dynamic processes interestingly approach near steady-state conditions at light speed.' [I haven't applied for permission to quote this.] Has this got any implications for the use of light in computing -- eg. is there an analogue of solid-state systems here? Using microfluidics? [The blogger (me) clearly knows nothing about pseudo-fluid processes, and just thinks there's a similarity with microfluidics because of the name!]


A note on faster-than-light processes (that's a laugh):

There seem to be a number of observations in nature, as well as inferences in theory, suggesting the possibility of faster-than-light processes. If so, can we harness any of them in an eco-benign and equitable fashion, for use in nonE computing, and elsewhere? Some such processes may include:

• radio jets from quasars (Sams, Bruce J., Andreas Eckart, and Rashid Sunyaev. "Near-infrared jets in the Galactic microquasar GRS1915+ 105." Nature 382, 47 - 49 (04 July 1996).); 
• the motion of electromagnetic solitons (Bugay, A. N., & Sazonov, S. V. (2004). Faster-than-light propagation of electromagnetic solitons in nonequilibrium medium taking account of diffraction. Journal of Optics B: Quantum and Semiclassical Optics, 6(7), 328.);  
• superluminal (=faster-than-light) tunnelling of light pulses observed in photonic barrier experiments (Winful, H. G. (2003). Optics (communication arising): Mechanism for 'superluminal' tunnelling. Nature, 424(6949), 638-638.) 

But of course some of these are electric or EM processes, if they exist. How can we 'catch' a hypothetical faster-than-light particle at the quantum level and translate its action into >125-nm-scale work, without running into problems of the use of artificial substances built at the quantum level? Or of artificially-organized light-speed particles getting into material where we don't want them and toxifying it (a superluminal version of the 'grey goo' or 'green goo' problem of nanotechnology and synthetic biology).

Can we get superluminal tunnelling through a fiberoptic cable that functions as a logic gate, etc.?

See also Murad, P. A. (2000). Hyper-light dynamics and the effects of relativity, gravity, electricity and magnetism. Acta Astronautica, 47(2), 575-587. Various claims exist for superluminal devices -- just go to GoogleScholar or other academic search engine and search for 'superluminal devices.'