The Physics Horserace

Physics is the horse to bet on. That is, in any matter under dispute, the player with the best hand from physics has the winning hand. Included in this argument from physics there exist details regarding all assumption - that is - basic statements about the facts of the matter. Time scales of 10^10 down to 10^-10; twenty orders of magnitude accuracy and precision. Our measurements of 'this physical specious present' have salient facts about all intervals of time (from the variations in duration over that interval) with sufficient precision to catch neutrinos traveling faster than the speed of light with respect to stationary human observers. The conclusion of this dispute will be judged by the world wide networked resources of physics. No idea regarding the meaning of the data will be discarded in this man hunt for falsification.

The falsification of theories being the primary purpose of science; the biggest prize in the falsification of the General Theory of Relativity is going to a physicist. No prayers said over this matter will be of any use. The other point being that the entire theory will no be overturned. My bet is on quantum and not relativistic causes. I am betting for one of the many physicists at the black jack table of science.

One Planet - One Architecture

So here is my bet. The ARM ISA is the salient architectural ecosystem for greater performance with lower energy. The big ideas were all tried out on X86 and POWER for system level and application level computing. They threw power at it. The 25 to 100 Watt envelope fills with all the performance that can be squeezed in for bragging rights in Server War I, the ground war over available floor space and the power grid. The metcalfe score for the ensemble of dispositions; Oracle indexing data bases, Google-verse indexing the network, Amazon selling the cloud and Apple being Apple in the cloud reached a set point. This burst of increased imagination in 7 billion pre-frontal corteses crossed the threshold of 'energy per application' that demanded Moore's law re leaf from ARM CORTEX. The intellectual property DNA of the ARM-VERSE suddenly has the system architecture that cuts out all the fat.

The ARM ecosystem grew by self-replication with variations. Each generation only kept a small sliver more of the system energy. That is, on system architecture long cycle of innovation and incremental Moore's law on a short cycle, The various design teams replicated the original with multiple variations. Every avenue in system architecture and silicon feature size was exploited ten different ways and the losers all copied the winner from each contest where possible. This is an ecosystem that could never run inefficient tasks because of severe (from the perspective of big iron) thermal constraints.

The memory management hardware in any ARM is better than the most efficient X86 server configuration. More of the important work is completed with the least possible work. The result is less bad. Less bad is the golden egg for ARM.

Physics is the horse to bet on. We can fill the world with maximum silicon cognition per Joule until quantum computing grows up. And we then improve on the orders of magnitude over which we observe our universe.

Re: Boltzmann Brain Hypothesis, a random fluctuation - an evolved CPU system architecture - emerged at precisely the right time (actually over an interval of time) to push the next emergent random fluctuation at a certain point in time. All because that is what a PDP=11 wants.