Parallel Computing with biological mass

Many a problem in modelling or factroisation , graphs ect fail as a result of its non linear and non deternimistic nature,whereby linear computationality fails and is completely squeezed out .The future as always has got a solution .In vouge being quantum computation.But to achieve the desired speed up the number of states to be maintained varies as (2 pow n) with n bits ,which in itself runs exponentially.So why cant we instead go in for a biological approach replicating our own neural network .It could be sumthing like a neural farm where neural mass is maintained in a nutrient solution and as it multplies at an exponential rate it might achieve the speed up we require .in quantum computationality thuogh better results could be achieved here we do not build the entire mechanism but we rather tinker the existing biological neural systems to provide us the computationality required.

ps I am not sure of all the data given above and would be highly thankful for any factual errors being apprised.Though this sounds similar to dna computing i would be thnkful to all to post a full account of it.