IEEE Spectrum June, 2014 - 16

Still, researchers must conware," says Austin Fowler, a
staff scientist at UCSB who also
tinue to improve accuracy
worked on the quantum error- rates before they can achieve
correction device.
highly reliable quantum comThe USCB team's success with
putations, which may still
using surface code on a linear require 1000 or even 10 000
array of qubits could lead to a
physical qubits to encode a
full 2-D grid of qubits capable of
single logical qubit. The idea
making important calculations. of connecting thousands of
The qubits would be arranged
qubits without causing interin a checkerboard pattern
ference among neighboring
where "white squares" would
qubits also presents a huge enhold data qubits for performing
gineering challenge, though
operations and "black squares" one that doesn't run afoul of
would contain measurement
any physics.
qubits that detect and correct
"The physics of coupling and
errors in the neighboring data
control is not going to change,"
qubits. In this setup, the sur- says Rami Barends, a postdocface code can indirectly mea- toral fellow in physics at UCSB.
sure possible errors in the data "But what you'll have to come
qubits without disturbing their up with is the wiring and condelicate quantum states.
trol done in a 2-D system withIBM researchers have also
out hampering the fidelity."
done pioneering work in makThe next step for the UCSB
team is to run simple erroring surface-code error correction
work with superconducting qu- correction experiments-a huge
first for the quantum-computing
bits. In research posted online to
the arXiv repository last Novem- field. Previously, researchber, one IBM group demonstrated
ers showed how to correct
a smaller three-qubit system ca- big errors that were deliberpable of running surface code, ately injected into quantumalthough that system had a lower computing arrays. But the UCSB
accuracy-94 percent.
researchers want to show how
"Both our result and the one
to correct for natural errors
from UCSB are showing the
that arise in the course of real
promise for superconducting
quantum- computing operaqubits, and that architectural
tions. A combination of imand engineering challenges
proved accuracy and rigorous
lie ahead of us and should be- error correction could eventugin to be addressed to get to- ally realize the dream of practical quantum computers.
ward a fault-tolerant quantum
"We have all the requirecomputer," says Jerry Chow, a
research staff member at IBM's
ments for starting to tackle
Thomas J. Watson Research Cen- error correction for the first
time," says Julian Kelly, a gradter, in Yorktown Heights, N.Y.
Improved accuracy for super- uate student in physics at UCSB.
conducting qubits makes the "People have gone through the
technology a serious rival of
motions before, but there has
other quantum-computer sys- never been a practical way of
reducing errors in the system."
tems, such as the use of trapped
-jer em y hsu
ions as qubits.

nEwS

JUn 2014

nORTh aMERICan

SPECTRUM.IEEE.ORG

BreaChing
the BlooD-Brain
Barrier
redesigned ultrasound machines
could help important drugs reach
brain cells
There's a barrier in your brain.
Composed of very densely packed cells
in the capillary walls, it restricts the passage of substances of the wrong size or
chemistry from the bloodstream. Like a locked fence
around your home, the blood-brain barrier prevents
intruders-such as infective bacteria-from entering.
But a locked fence can also keep out rescuers in an
emergency, and the blood-brain barrier keeps out potentially helpful drugs that might be able to ease the suffering of the tens of millions of people with Alzheimer's
disease, Parkinson's disease, multiple sclerosis, amyotrophic lateral sclerosis, and other diseases of the central nervous system. Less than 5 percent of the roughly
7000 available drugs can get through. Basically, none
of the large-molecule drugs can, severely limiting the
options for new therapies.
But there's hope. Blasts of ultrasound can temporarily open the barrier in tightly focused spots of the brain
that are just millimeters in diameter. And engineers at
Chang Gung University, in Taiwan, have recently come
up with a much improved way of delivering that energy.
The prototype device they developed is a 256-channel
ultrasound phased array. According to electrical engineering professor Hao-Li Liu, his team has developed
a unique circuit design involving multiple microcontrollers and power-sensing feedback circuits that
enable the system to deliver two frequencies at once
instead of the single frequency that biomedical researchers have been working with. By altering the phase of
individual channels, the array produces millimeterswide spots of ultrasound energy that can be electronically steered to any point in the brain.
It's been known for a while that ultrasound reversibly opens the blood-brain barrier, even if the exact
workings haven't quite been nailed down. The process relies on the acoustic cavitation effect, which is
the growth and collapse of microbubbles in a liquid

http://SPECTRUM.IEEE.ORG

Table of Contents for the Digital Edition of IEEE Spectrum June, 2014