Static information storage and retrieval – Interconnection arrangements
Reexamination Certificate
2002-11-29
2004-09-28
Nguyen, Viet Q. (Department: 2818)
Static information storage and retrieval
Interconnection arrangements
C365S185240, C365S185100, C365S226000
Reexamination Certificate
active
06798682
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is related to integrated circuit power consumption and more particularly to reducing static random access memory (SRAM) power consumption.
2. Background Description
Semiconductor technology and chip manufacturing advances have resulted in a steady increase of on-chip clock frequencies, the number of transistors on a single chip and the die size itself, coupled with a corresponding decrease in chip supply voltage and chip feature size. Generally, all other factors being constant, the power consumed by a given clocked unit increases linearly with the frequency of switching within it. Thus, not withstanding the decrease of chip supply voltage, chip power consumption has increased as well. Both at the chip and system levels, cooling and packaging costs have escalated as a natural result of this increase in chip power. For low end systems (e.g., handhelds, portable and mobile systems), where battery life is crucial, net power consumption reduction is important but, without degrading performance below acceptable levels.
To minimize power consumption, most integrated circuits (ICs) used in such low end systems (and elsewhere) are made in the well-known complementary insulated gate field effect transistor (FET) technology known as CMOS. A typical CMOS circuit includes paired complementary devices, i.e., an n-type FET (NFET) paired with a corresponding p-type FET (PFET), usually gated by the same signal. Since the pair of devices have operating characteristics that are, essentially, opposite each other, when one device (e.g., the NFET) is on and conducting (ideally modeled as a closed switch), the other device (the PFET) is off, not conducting (ideally modeled as an open switch) and, vice versa.
For example, a CMOS inverter is a series connected PFET and NFET pair that are connected between a power supply voltage (V
dd
) and ground (GND). Both are gated by the same input and both drive the same output, the PFET pulling the output high and the NFET pulling the output low at opposite input signal states. Ideally, when the gate of a NFET is below some positive threshold voltage (V
T
) with respect to its source, the NFET is off, i.e., the switch is open. Above V
T
, the NFET is on conducting current, i.e., the switch is closed. Similarly, a PFET is off when its gate is above its V
T
, i.e., less negative, and on below V
T
. Thus, ideally, the CMOS inverter in particular and CMOS circuits in general pass no static (DC) current. So, ideal CMOS circuits use no static or DC power and only consume transient power from charging and discharging capacitive loads.
In practice however, transient power for circuit loads accounts for only a portion of the power consumed by CMOS circuits. A typical FET is much more complex than a switch. FET drain to source current (and so, power consumed) is dependent upon circuit conditions and device voltages. FETs are known to conduct what is known as subthreshold current below threshold for NFETs and above for PFETs. Subthreshold current increases with the magnitude of the device's drain to source voltage (V
dS
) and inversely with the magnitude of the device V
T
. Among other things, V
T
is inversely proportional to gate oxide thickness and, to some extent channel length, both of which are related to feature size. In addition, gate leakage, to channel, to source or drain and gate induced drain leakage (GIDL) can also contribute to static power and are also related in particular to oxide thickness. Thus, as chip features shrink, these leakage sources become more predominant. This is especially true in what is known as partially depleted (PD) silicon on insulator (SOI) technology, where subthreshold leakage has been shown to increase dramatically, such that it may be the dominant source of leakage. When multiplied by the millions and even billions of devices on a state of the art IC, even 10 picoAmps (100 pA) of leakage in each devices, for example results in chip leakage on the order of 100 milliAmps (100 mA).
For logic chips such as general and special purpose processors, non-load related power dissipation is fairly randomly distributed throughout the logic. Transient power tends to dominate logic chip power consumption. Arrays, however, such as random access memories (RAMs) and especially static RAMs (SRAMs), have a large areas that may remain dormant during any one operation. Thus, in these dormant areas, leakage can become a substantial source of power consumption and is nearly the sole source of standby power consumption.
A typical SRAM array is an array of SRAM cells that are each essentially a pair of cross-coupled inverters selectively couple to a pair of complementary bit lines by a pair of pass gates or word line devices. Typically, the cells may be organized n word lines by m bit lines (complementary bit line pairs) by k bits. So, accessing one bit from one of the k (or more) subarrays entails selecting one of the n word lines. Of the m cells partially selected by that word line, only one (on one of the m bit lines) may actually be accessed. During a read, each of the bit line pairs rises/droops only to develop enough signal (e.g., 50 mV) for a sense amplifier. During a write, the pair for the cell being accessed may be driven at least what is termed “rail to rail,” i.e., to opposite extremes (V
dd
and GND) and for a short period of time, a portion of those write voltages are passed to the cell. Then, the word line drops isolating the cell from the bit line and, the cell completes latching what was written.
It is well known that any mismatch in the cross coupled inverters, whether inverter load, in inverter devices or layout can introduce cell sensitivities or noise that causes the cell to favor one state over the other. These cell sensitivities can impair cell performance, cell reliability and in the extreme render the RAM useless. Therefore, typically cell designers strive to design perfectly balanced cells.
Since low end systems may include several chips, it is important to reduce power in each. However, typically, memory accounts for a large number of those chips. Especially with low end systems those memory chips are SRAM. So, SRAM chip power is multiplied by the number of SRAM chips included. While high chip power may be tolerable for a single (e.g., processor) chip, when multiplied by a number SRAM chips it can account for a significant portion of system power, making the difference between acceptable and unacceptable system battery life. Thus, SRAM design and especially, SRAM cell designers must balance power consumption concerns and especially leakage with performance.
Thus, there is a need for reduced SRAM chip power consumption.
SUMMARY OF THE INVENTION
It is a purpose of the invention to reduce integrated circuit power consumption without impacting circuit performance;
It is another purpose of the invention to reduce SRAM power consumption without degrading SRAM performance;
It is yet another purpose of the invention to reduce subthreshold leakage in SRAM arrays.
The present invention is an integrated circuit that may include an array such as a static random access memory (SRAM) with gate oxide selectively thickened in the array and in selected other devices based on threshold voltage (V
T
) variations with gate oxide thickness for the particular technology, e.g., bulk insulated gate FET, bulk CMOS, PD SOI CMOS, fully depleted SOI or double gate CMOS. Some or all array devices may have thicker gate oxide. Thicker oxide may be used in non-core circuits, e.g., test circuits. Also, non-critical paths may be identified and a non-critical path margin identified. A thicker gate oxide is selected for non-critcal path FETs based on the non-critical path margin. Non-critical path delays are re-checked. FETs are formed with the selected thicker gate oxide for any non-critical paths passing the re-check and in array FETs with non-selected FETs being formed with normal gate oxide thickness.
REFERENCES:
patent: 6515521 (2003-02-01), Kono et al.
patent: 6534807 (2003-03
Chuang Ching-Te K.
Joshi Rajiv V.
Rosenfield Michael G.
International Business Machines Corp.
Law Office of Charles W. Peterson, Jr.
Nguyen Viet Q.
Percello Louis J.
LandOfFree
Reduced integrated circuit chip leakage and method of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Reduced integrated circuit chip leakage and method of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Reduced integrated circuit chip leakage and method of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3268717