Information Technology Reference
In-Depth Information
Repeat steps (1)-(4) roughly 2 n/2 times. When the iterations are completed,
sampling may take place to provide with good probability a solution to the problem
of what satisfies the given function. The solution must be checked by substitution,
and if it fails, the method is repeated.
This method is a vast improvement over having to evaluate a function for up to 2 n
times (once for each possible input combination), trying to hit upon what satisfies it.
Increasing the Capacity of Long-Term Memory
Simulated qubit combinations might be useful for portions of long-term memory.
For instance, using two simulated qubits, binary-like information may be created
with combinations of [1 0] 0 ,[01] 0 , and
[1 1] 0 , employed two at a time. Using [1 0] 0 ,
[0 1] 0 the following probability vectors may be created: [1 0] 0 ,[10] 0 ;[10] 0 ,[01] 0 ;
[0 1] 0 ,[10] 0 ;[01] 0 ,[01] 0 , these being ordinary binary coding 00 , 01 , 10 , 11 which
can be sampled and read with certainty. In addition, using [1 0] 0 ,
η
[1 1] 0
η
the following additional codes may be created: [1 0] 0 ,
[1 1] 0 ;
[1 1] 0 [1 0] 0 ;
η
η
using [0 1] 0 ,
[1 1] 0 the following additional codes may be created: [0 1] 0 ,
[1 1] 0 ;
η
η
[1 1] 0 ,[01] 0 ; and finally there is the code
[1 1] 0 ,
[1 1] 0 . Components containing
η
η
η
[1 1] 0 have to be read more than once to determine their presence. Assuming this
is done, two simulated qubits can store 9 codes, which is more than 2 2
η
¼
4 codes.
The improvement increases exponentially with an increase in n.
Generally the number of additional data items grows exponentially with the
number n, the number of independent multivibrators. By considering independent
elements as being [1 0] 0 ,[01] 0 , and
[1 1] 0 there could be 2 n + n{2 n1 +2 n2 ,
η
,2 1 } + 1 codes. This, the author's calculation, is the binary count 2 n of the basic
variables 0 , 1 ; plus the binary count with
[1 1] 0 in place of one variable; plus the
η
binary count with
η
[1 1] 0 in place of two variables; and so on to
η
[1 1] 0 in place of
all n variables.
The net count using n simulated qubits is far more than the measly 2 n codes using
n bits. However, since some of the codes are probabilistic, waveforms would have
to be sampled several times in order to see faithfully the original data. But this is
easily accomplished in a classical system by permitting several sampling pulses and
letting the results accumulate in a special register.
Conclusions
This chapter introduces multivibrator functions on simulated qubits. Multivibrator
functions are those which modify the position of a state vector with the end result
that one or more probability combinations are tagged with negative signs. Such tags
are invisible to a sampling system because sampling only indicates a positive
probability for a given output combination. Sampling cannot determine phase.
Search WWH ::




Custom Search