Leakage power is consumed when the power is on, regardless of whether the circuit is operating or not. For example, when a remote control unit is used to turn off the power to a liquid crystal display (LCD) television the circuit stops operating, but the power may still be input to the IC. This "fast start" mode is often needed when the TV is to be used together with other equipment. If power to the IC is cut off, it can take several seconds for it to become ready, so power is left on constantly. For an LCD TV and associated equipment, this can translate into leakage power of several W to about a dozen W.
Worse, leakage power tends to rise as process technology rules shrink. For a chip with a 10% on ratio, for example, leakage accounts for no more than 10% of total dissipation in a 130nm-generation IC, but for "over half" in a 65nm chip, says a source at Rohm.
One effective method
of reducing leakage power already in use (Fig 3) is "power gating," a
method of precisely shutting off power to individual circuits that are
not needed. Leakage can be also be reduced by raising the transistor
threshold voltage, but that simultaneously lowers transistor speed.
Power gating does not suffer from this drawback.
The power supply can be cut off to the entire chip, or in circuit block units, for example.
Normally when the power is cut, flip-flop data is stored to another location to preserve it, and in most cases this means transferring it to a retention flip-flop before shutting off the power. Only the retention flip-flop is powered, maintaining the data. However, it is always possible that power supply on/off noise can corrupt data in the retention flip-flop, which makes it essential to precisely control power-off timing (Note 1).
Note 1: This approach also means that the retention flip-flop must be constantly supplied with power, so IC leakage does not drop to zero.
When this type of precise timing control is needed, explains a source at
Rohm, "It is difficult to turn very many circuits on and off." This is
because power-off timing can get very complex when many circuits are
involved. At present, Rohmadds, "We can control about a dozen circuit
blocks, but it gets very tough very quickly beyond that."
Nonvolatile flip-flops (Fig 4) resolve this problem. Flip-flop data is preserved in the nonvolatile device even when the power is cut off, eliminating possible effect from power supply noise. As a result, explains Rohm, "We can control about ten times as many circuits."
Choice: FeRAM? MRAM?At present there are two proposed methods of making nonvolatile flip-flops: the Rohm approach, based on ferroelectric random-access memory (FeRAM), and the NEC approach, using magnetic RAM (MRAM) devices (Fig 5). Both designs add nonvolatile devices to conventional flip-flops. The nonvolatile devices are disconnected during normal operation, so that the flip-flop operates just like a standard design, but when the power is to be cut off data is stored to the nonvolatile memory.
Nikkei Electronics Asia magazine is available each month free of charge to engineers, managers and other qualified readers.
Nikkei Electronics |
Nikkei Monozukuri |
Nikkei Automotive Technology |
|---|---|---|
 |
 |
 |
| The magazine for electronics designers and managers. | Design You Won't Regret - How to Reduce the Risk of Accidents | The comprehensive automotive technology magazine. |
Copyright © 1995-2010 Nikkei Business Publications, Inc. All rights reserved.
All editorial content and graphics on this Web site may not be reproduced in whole or in part without the express permission of the copyright owner.
February 2010
January 2010
December 2009