
Noisebased logic, computing and brain
circuitry
Presentation (two talks combined) at Hot Topics of Physical Informatics (HOTPI2013), November 13, 2013, China
New
application: Physical Uncloneable Function
(PUF) noncounterfeitable
hardware
keys.
Presentation: Noisebased logic: Why
Noise?
Presented as invited talk at ICCAD 2012, at the
Special Session: Computing in the Random Noise:
The Bad, the Good, and the Amazing Grace;
November 5, 2012, San Jose, CA, USA.
Artistic animation: Randomtelegraphwave (RTW)
hyperspace operations in noisebasedlogic
Some
of the features:
Original motive:
Noisebased logic is a new approach inspired by
the fact that the
neural spike trains in the
brain are usually stochastic.
New
deterministic, multivalued
logic scheme, and
neural circuitry
(logic gates) proposed to understand the
brain.
Able to perform "intelligently":
quick decisions, based on a small amount
of information, with
reasonably low error rate.
Exponential gain in
computing performance at proper
specialpurpose applications.
Potential
advantages, among others, are: noisebits
(similar to qubits)
with hyperspace vector products offering
2^N orthogonal vectors
and 2^(2^N) logic values in
a single wire;
reduced power dissipation; analog circuits
(with natural saturation
thresholds); switching errors do not
propagate/accumulate; robustness
against noises and other usual types of
interference (ground plane EMI,
transients, etc).
Papers/manuscripts:
13. New application:
PUF noncounterfeitable hardware keys.
L.B. Kish, C. Kwan,
"Physical Uncloneable Function Hardware Keys
Utilizing KirchhoffLawJohnsonNoise Secure Key
Exchange and NoiseBased Logic", submitted for
publication. click
here
12.
Complex noisebit are
introduced as carriers of the logic information.
Ghost noisebits are also introduced for fast
measurements and
manipulations in large products and
superpositions. H. Wen, L.B. Kish,
A. Klappenecker, "Complex
NoiseBits and
LargeScale Instantaneous Parallel Operations
with Low Complexity", Fluctuation and Noise
Letters accepted for publication. click here
for the preprint
11. Why
noise, orthogonality of
base
signals is not enough? The answer is: noise can
be exponentially
better: H. Wen, L.B. Kish, "Noise based logic:
why noise? A comparative
study of the necessity of randomness out of
orthogonality", Fluctuation and Noise
Letters 11 (2012) 1250021, click
here to get the manuscript
10.
Exponential
speedup at evaluating
product strings (identifying arbitrary
hyperspace vectors) in
instantaneus noisebased logic when the
timeshifting method described
in [9, see below] cannot be used: L.L. Stacho,
"Fast measurement of
hyperspace vectors in noisebased logic", Fluctuation
and Noise Letters, in press (2012), click
here to get the manuscript
9.
Exponential speedup at evaluating
product strings (identifying arbitrary
hyperspace vectors) in
instantaneus noisebased logic: H. Wen, L.B.
Kish, A. Klappenecker, F.
Pepper, "New noisebased logic representations
to avoid some problems
with time complexity", Fluctuation
and Noise Letters, in press (2012), click
here to get the manuscript
8. Fast,
binary, noisebased logic without time
averaging, where both the Low
and
the High values are represented by independent
noises. Transforming
binary noisebased logic values up to the
multidimensional logic
hyperspace
can also be performed easily in this way:
F. Peper and L.B. Kish,
"Instantaneous, nonsqueezed, noisebased
logic", Fluctuation
and Noise Letters, 10 (2011)
231–237 (Open Access). click
here to get the manuscript
7. Performing
in an intelligent way: quick
decisions, based on a small
amount of information, with reasonably
low error rate. Computation
using
Noisebased Logic: Efficient String Verification
over
a Slow Communication Channel, by L.B. Kish, S.
Khatri, T. Horvath,
accepted for publication in European
Journal of Physics B 79 (2011)
8590, click here to
get
the paper
6.
Noisebased
logics and their Boolean gates without time
averaging; random telegraph
wave and Boolean brain logic; L.B. Kish, S.
Khatri, F. Pepper, "Instantaneous noisebased
logic", Fluctuation
and Noise Letters 9 (2010)
323–330, click
here
to get
the paper
5. Some
practical/conceptual aspects of spike based
hardware; Z. Gingl, S.
Khatri, L.B. Kish, "Towards braininspired
computing", Fluctuation and
Noise Letters 9 (2010)
403–412, click here to get
the paper
4.
Deterministic logic based on noisy spikes. First
noisebased logic
without time averaging. S.
Bezrukov, L.B. Kish, "Deterministic multivalued
logic scheme for
information processing and routing in the
brain", Physics
Letters A 373
(2009)
23382342, click
here to get
the paper
3. Utilization
of the noise hyperspace to make 2^N orthogonal
noise
vectors. A string search algorithm outperforming
Grover's quantum
search algorithm with real data at the same
hardware complexity class;
L.B. Kish, S. Khatri, S. Sethuraman,
"Noisebased logic hyperspace with the
superposition of 2^N states in a
single wire", Physics
Letters A 373 (2009) 1928–1934, click here to
get
the paper
2.
First concrete solutions for noisebased
logic:
"Noisebased
logic: binary,
multivalued, or fuzzy, with optional
superposition of logic
states", Physics Letters A 373 (2009)
911–918, click
here to get
the paper
1. First
(now outdated) vision with an estimation
of the
energyneed but
without concrete logic solution: "Thermal noise
driven computing"; Applied Physics
Letters 89, 144104 (2006), click
here to get
the paper
New, related conference series: Hot
Topics of Physical
Informatics, HotPI2013
click here
In the media. Features
of noisebased logic:
1.
"Breaking
the noise barrier",
by Justin Mullins, New Scientist,
September 29, 2010
,
click
here to read the
extract
2. "Cover story:
What's
all this noise about?", by David Boothroyd, New Electronics
magazine (UK),
February 22, 2011,
the part
featuring noisebased logic is in the 3rd
part, see also the two figures there. click
here to read the story
