Reiner_Ha

here already in 1984 a vN software to #Xputer migration of a Mead-&Conway-based microchip design rule checker obtained #speed-ups_by_FPGAs by a factor of 15000.

reaching astronomic dimensions this is the usual software code size unit (Million Lines Of Code): e. g. for Windows XP (2001) with 40 MLOC, for Mac OS X 10.4 (2006) with 86 MLOC, and for SAP NetWeaver (2007) even with 238 MLOC.This is commented by #Nathans_Law (“Software is a gas, which expands to fill all its containers.”) This ecplains the #memory_wall causing inefficiency by additional orders of magnitude requiring large power-hungry and very slow extra memory capacity.

(“Software is a gas, which expands to fill all its containers.”) meets the #memory_wall causing inefficiency by additional orders of magnitude since software packages often reach astronomic dimensions and running them requires large power-hungry and very slow extra memory capacity. The usual software code size unit is “MLOC” (Million Lines Of Code): for Windows XP (2001) with 40 MLOC, for Mac OS X 10.4 (2006) with 86 MLOC, and for SAP NetWeaver (2007) even with 238 MLOC.

also called #Energy_Wall (the electricity consumption of computers) cannot be ignored any longer. The energy consumption of server farms, supercomputers and all other ICT infrastructures is heading toward astronomic dimensions and may become unaffordable towards the end of next century or even earlier.

also called #Power_Wall (the electricity consumption of computers) cannot be ignored any longer. The energy consumption of server farms, supercomputers and all other ICT infrastructures is heading toward astronomic dimensions and may become unaffordable towards the end of next century or even earlier.

is the counterpart of the von Neumann paradigm by using data counters instead of a program counter. Xputers being programmed by #configware instead of #software provide speed-ups and energy-efficiency improvements by up to several orders of magnitude. Also see #Energy_wall and #time_to_space_mapping

Although migrations from vN computing over to FPGAs provide speed-ups by up to more than 4 orders of magnitude, the market success of FPGAs is by orders of magnitude worse than that of other semiconductor products. This paradox os caused by the #Education_wall

means that the market success of FPGAs is by orders of magnitude worse than that of other semiconductor products. This paradox os caused by the #Education_wall

is causing the #FPGA_marketing_paradox causing by orders of magnitude worse marketing success than that of dramatically less efficient other semiconductor products. This paradox is caused by the #Education wall.#Memory_wall

The migration of an implementation from Software over to Reconfigurable Computing platforms like #FPGA requires time to space mapping by turning a program into a pipe network. For instance The decision box of a flowchart from the time domain is directly implemented by a demultiplexer, whereas the 3 operation boxes are implemented as flip-flops providing the evoke bit for triggering the executions. See "The alternative Machine Paradigm for Energy-Efficient Computing" by Reiner Hartenstein (submitted).

inevitable machine paradigm to cope with the #von_Neumann_syndrome. An Xputer runs by data counters instead of a #program_counter.

the reconfigurable processor part of an #Xputer to be programmed by #Configware - the counterpart of the hardwired CPU of #von_Neumann-processors.

Source to program the #Morphware part of an #Xputer - Configware is the counterpart of #Software to program #von_Neumann computers.