Time in Physics and Computer Science Transactions

Patents Pending | Earth Computing Inc. | Paul Borrill
Time in Physics and
Computer Science
Paul Borrill
EARTH Computing

View Slide

Time emerges from entanglement
Everything you thought you
knew about time is now obsolete

View Slide

Time is change we can count
There is no evidence whatsoever
for a smooth background of spacetime
Bell’s Inequalities have been telling us for a while now
that there’s something wrong with Minkowski’s 4D “spacetime”

View Slide

Time is not continuous
Everything we measure is discrete

View Slide

Time is not irreversible
All our theories are time reversal invariant
Classical (Newtonian)
Relativistic (Special and General)
Standard Model (Quantum Field Theory)

View Slide

Everything Moves at SOL

View Slide

Everything Moves at SOL
You can’t see
photons
coming;
you can only
see them
arrive

View Slide

So What?

View Slide

We cannot achieve exactly
once delivery unless it is paired
(intimately) with exactly once
transmission
not guaranteed by the “end to end principle”

View Slide

Switches are DReDDful
(they Drop, Reorder, Delay and
Duplicate packets)
We Need Something Else

View Slide

Compare and set atomicity
cannot be guaranteed across a
switched network, or any
network that can be partitioned

View Slide

Heartbeats don’t scale

View Slide

Timeouts create cascade failures

View Slide

Cell
Agent
A
∅
Half Link A
Half-Edge
Entangled (Hidden ENTL Packets)
ENTL
Circulating
ENTL Event
ENTL
ENTL
w x
z y
EARTH Computing Link
• Reversible
• No-Copying
• Atomic Information Transfer (AIT) 
Guaranteed in the sense that both sides can “know” if it failed or succeeded
Cell
Agent
A
Cell
Agent
B
Half
Link A
Half
Link B
Tx/Rx Media

View Slide

Mathematical Foundation
Groupoid semantics for timeless networks
and secure classical key distribution
David Reutter
Department of Computer Science, University of Oxford
[email protected]
Jamie Vicary
Department of Computer Science, University of Oxford
[email protected]
Abstract—We provide a mathematical foundation for timeless
networks, a new paradigm for distributed communication which
has recently been proposed. Our approach is based on groubits,
generalizations of classical bits arising from groupoids with spe-
cial properties. Our techniques give a clean mathematical model
for timeless networks, and allow the description and veriﬁcation
of a number of interesting protocols, including message routing
without timeouts, and information theoretically–secure classical
key distribution, under minimal security assumptions. We also
build classical-physics implementations of a number of quantum
protocols on networks of groubits, including dense coding and
teleportation.
I. INTRODUCTION
A. Overview
In this paper we study the foundations of timeless networks,
a new paradigm for distributed communication presented re-
cently by Borrill [1] and currently under commercial devel-
opment by Earth Computing1. We study a mathematically-
The read operation destroys a groubit and creates a conven-
tional bit, while the write operation destroys a conventional
bit and creates a new groubit. Pairs of groubits can also be
connected by a link, enabling the Tick operation, where A and
B label the two connected nodes, and ⊕ is addition modulo 2:
• Tick((AL,AI
),(BL,BI
) =((AL,AI
⊕BL
),(BL,BI
⊕AL
))
Intuitively, for each node in the pair, we ﬂip the internal bit
just when the other node has logical bit equal to 1. Nodes can
belong to multiple links, in general forming a graph topology.
B. Assumptions
We make some assumptions about these groubit operations.
• Atomicity. The operations Init, Swap, Read, Write and
Tick are implemented atomically.
• Security. A user can never discover any information about
the state of a node, except via Read.
We emphasize that claims we make about the functionality
arxiv.org/abs/1707.00966

View Slide

P
P:0
Q:--
R:--
Q
P:--
Q:0
R:--
R
P:--
Q:--
R:0
P
P:1
Q:2
R:1
P
P:2
Q:2
R:1
P
P:3
Q:3
R:3
Q
P:--
Q:1
R:1
Q
P:--
Q:2
R:1
Q
P:--
Q:3
R:1
Q
P:2
Q:4
R:1
Q
P:2
Q:5
R:1
R
P:--
Q:--
R:1
R
P:--
Q:3
R:2
R
P:--
Q:3
R:3
R
P:2
Q:5
R:4
R
P:2
Q:5
R:5
P
P:4
Q:5
R:5
t
Process
Causal History
Future
Effect
slope ≤
c
slope ≤
c
slope
≤ c
slope
≤ c
11 12 13 14
21 22
23
24 25
32
31 33 34 35

View Slide

Why Can’t Links be Reversible?
the snapshot
problem

View Slide

Thank You
EARTH Computing
Solid Ground Beneath The Clouds

View Slide

References
• Stanford 2016: “The Time-Less DataCenter" (November 2016)
• Video: https://www.youtube.com/watch?v=IPTlTmH-YvQ
• Slides: http://web.stanford.edu/class/ee380/Abstracts/161116-slides.pdf
• Info: http://web.stanford.edu/class/ee380/Abstracts/161116.html
• Papers We Love 2016: “Lamport’s Unﬁnished Revolution” (July 2016)
• Video: https://www.youtube.com/watch?feature=youtu.be&t=32m27s&v=CWF3QnﬁhL4
• Slides: https://speakerdeck.com/pborrill/time-clocks-and-the-reordering-of-events-pwl-san-francisco-14-jul-2016
• Info: https://www.meetup.com/papers-we-love-too/events/228341271/
• Stanford 2014: “Time in Physics, and Implications for Computer Science”:
• Video: https://www.youtube.com/watch?v=SfvouFIVCmQ
• Slides: http://web.stanford.edu/class/ee380/Abstracts/140416-Borrill-slides.pdf
• Info: http://web.stanford.edu/class/ee380/Abstracts/091111.html
• Time’s arrow emerges from observers. https://www.wired.com/2016/09/arrow-of-time/
• PBS SpaceTime https://www.youtube.com/watch?v=GguAN1_JouQ
• Groubits https://arxiv.org/abs/1707.00966

View Slide

Time in Physics and Computer Science Transactions

Time in Physics and Computer Science Transactions

Boz

More Decks by Boz

Other Decks in Technology

Featured

Transcript