The Future of IT

1
The Future of IT

• Dynamic and Seamless Personal Networks using
  voice recognition, flat screen displays, talking
  to other PNs.
• Needs
• More Powerful Computers
• Improved User Interfaces (speech recognition and
  speech synthesis)
• Faster Communications (IR-4Mbps, wireless
  radio-10-33Kbps, future 155Mbps, optical
  fibers-2Mbps, Teledesic array of satellites)
• Security (passwords, voice, fingerprints, iris,
  cryptography)
• Better Batteries lithium ion technology

2
Silicon-based Computers

• The shrinking of the transistor (1 cm..100 nm)
• Integrated circuits.
• Moores law
• Transistors per chip (1000 ..10 million ..)
• Atoms per bit (1019 .. 1)
• Clock frequency (1 MHz .. 1GHz ..)
• Energy per logical operation (107 pJ .. 10-5 pJ
  ..)
• Limits of the current semiconductor chip
  technology (transistors of 70 nm in 2010). If the
  problems related to these limits are solved, the
  size of a transistor will reach the size of an
  atom (i.e.below 1 nm) in 2020.

3
Quantum Effects

• The Uncertainty Principle (Heisenberg 1928)
• Probability Waves (deBroglie 1925, Schrodinger
  1927)
• Superposition Principle (entangled states)
• Quantum Interference (the 2 slit experiments
  1985, 1991)
• Quantum Teleportation (Bennett et al 1993)

4
Simulating a Quantum Computer

• Bits are replaced by Qubits vector states
• Logical operations (hardware) become a
  Hamiltonian (large matrix), which describes the
  evolution in time of the qubit.
• Quantum computers Benioff 1980, Feynman 1985,
  Deutsch 1985
• Complicated method to determine when to stop the
  calculation
• Error models and error correction

5
Quantum Computers and Cryptography

• The RSA system (Rivest et al 1978) uses the fact
  that the factoring a very large integers is very
  difficult. To calculate RSA-129 it took 17 years
  and a network of about 1600 classical computers.
  Shor 1994 showed how a quantum computer can do it
  very efficiently.
• Bennett et al 1989 used polarized photons to
  build the prototype of the first quantum
  cryptography machine.

6
Quantum Teleportation

• Teleportation is usually understood as a transfer
  of an object by transmitting the information
  about the construction of the object. Because of
  the Uncertainty Principle it was assumed that one
  cannot fully get full information even about a
  very simple particle.
• In 1993 Bennett et al showed how the principle of
  entangled states circumvents this problem. In a
  2-state quantum system (such as 2 electrons with
  opposite spins or 2 photons polarized
  differently) the determination of the 2 particles
  is done simultaneously. Brassard 1996 built a
  first quantum circuit to teleport a quantum state
  between two parts of a quantum computer.

7
How to Build a Quantum Computer ?

• Storing a qubit Bathe an atom in laser light of
  energy equal to the energy difference between
  ground state and an excited state.
• Reading a qubit Bathe the atom in laser light of
  energy equal to a jump to higher excited state (1
  is affected but not 0)
• Quantum computations NOT, COPY and AND

8
Quantum Computations
Initial
Final
0
1
NOT gate
1
0
A/B
A
Initial
Final
Light
1
1
1
COPY gate
0
0
0
B
A
A
B
9
Quantum Computations
A/A
Initial States
Final States
B
1
1
1
0
0
0
0
Light
0
1
1
0
0
A
B
A
B
AND gate
A
A
10
Multiparticle Quantum Computer

• A QC which can simulate any quantum system. Lloyd
  1993 showed that the more bits in QC the faster
  the calculation. A 40-bit QC could with 100 steps
  calculate a quantum system that a normal computer
  would do in tens of years.
• If a QC can have parallel architecture (ex.double
  spin resonance in crystals) gives huge speed.
  However, even regular QCs have quantum
  parallelism (Deutsch 1990).

11
Promising Quantum Computer (QC) Projects

• Note All projects consider both specialized and
  general purpose computers.
• Heteropolymer-based QC linear array of atoms
  (Teich et al 1988)
• Ion Trap-based QC linear array of ions
  (CiracZoller 1995)
• Cavity QED-based QC (Caltech 1995)
• NMR-based QC (Adleman 1994)

12
Economic Issues

• The cost of a new semiconductor plant has been
  doubling every 3 years (Hutcheson 1996). If this
  continues by 2020 the cost a new plant would be
  1 trillion (5 of the predicted US GDP) and to
  be profitable it would need sales of 10 trillion
  (50 of the US GDP) !
• Today 5 of the power generated in US is used by
  computers (Malone 1995). It is reasonable to
  assume that in 2020 computers at 40GHz with 160Gb
  RAM will dissipate the same power as today (about
  40W), but their numbers will almost certainly
  increase a lot.
• Quantum computers promise to be significantly
  more efficient. Reversible QCs with no net energy
  consumption.

13
Social and Political Pressures

• Quantum computing research requires a mix of
  computer science and quantum physics, which is
  available in North America.
• CIA and other similar organizations are extremely
  interested in the impact that QCs can have on
  cryptography.
• Governments have reduced support for pure science
  and make available significant resources for
  quantum computing.