A capacitorless DRAM memory cell uses an intermediary lightly-doped drain region to buffer electrical signals.
XOR operations on selection signals set internal counter codes, enabling stacked memory chips to switch active modes without complex control logic.
Multiport memory system supplies immediate status updates between ports, resolving data transfer delays inherent in traditional handshake methods.
Flying decoder lines use vertical M4 and M2 track stacking to cut parasitic resistance by 50% and lower access time.
Replacing slow epitaxy, this sputtering process grows high-quality topological insulator films at industrial speeds for spin-orbit torque memory devices.
Tracking circuit mirrors leakage currents from the memory array to compensate for temperature variations and improve write reliability.
A memory controller adjusts refresh rates based on detected error levels to maintain data integrity.
Interface circuit combines staggered data bursts from multiple memory devices into contiguous streams for higher bandwidth.
A reset selection unit selectively outputs reset signals to specific I/O switches, eliminating unnecessary power consumption from non-operational regions.
Dynamic voltage adjustment compensates for threshold voltage variations caused by shrinking feature sizes, improving stability and writability.
Dynamic voltage adjustment via a compensation circuit maintains data accuracy while preventing speed limitations in SRAM write operations.
A neuromorphic neuron apparatus with a switchable accumulation block processes temporal sequences using state variables and decay functions.
A dedicated complementary bit line driver maintains high voltage levels during logic zero writes in low voltage SRAM cells.
Segmenting memory rows into independent modules with locally coupled redundancy sections resolves signal non-uniformity caused by repair structures.
A multi-level cell sensing circuit uses switch control units to manage read bits for efficient data retrieval.
Dynamic substrate and well bias adjustment lowers SRAM supply voltage requirements while preventing transistor state flipping at reduced power levels.
Adaptive resistive memory reading applies sample pulses to determine logic states, compensating for resistance drift in crossbar arrays.
Weakly coupled magnetic layers distribute spin torque across multiple interfaces to reduce programming current requirements.
A memory module uses data buffers to regenerate locally synchronized clock signals for independent device groups.
Auxiliary circuits define operation margins that prevent parameter uncertainty and ensure correct switching between modes.
Inverting layer sequence protects exposed racetrack from etching damage while reducing stray tunneling current to improve thermal stability.
A ferroelectric tunneling field-effect transistor uses asymmetric source and drain doping to control charge carriers.
Back gate control circuits dynamically adjust voltage to prevent forward bias and reduce power loss during high-side path activation.
A column control circuit generates a unified pulse to synchronize read and mode register operations in DDR SDRAMs.
A memory controller programs a programmable delay line to adjust chip select timing using the write leveling mechanism.
Dummy memory cell logic circuit detects potential change to activate sense amplifier, reducing area overhead and optimizing timing.
Off-chip reference cells copy memory current characteristics to compensate for temperature drift without adding on-chip complexity.
A memory controller determines timing offsets for mask and data strobe signals to align write operations.
Overlapping bitline operations reduces cycle time in pseudo-dual port memory designs.
A segmented forming scheme applies specific voltage biases to word line combs and bit lines in cross-point memory arrays.
Detection logic monitors activation commands from external and internal processors to identify frequently accessed rows in DRAM memory banks.
A shift register uses a node controller to manage set node charging duration for stable scan pulse output.
Staged clock gating reduces IR drops and transients during data detector startup.
A memory circuit divides storage into independently powered sub-arrays controlled by shared peripheral logic to manage active states.
Dynamic pulse sequences resolve wide resistance distribution in RRAM, improving writability and readability without exceeding transistor current limits.
Segmented sub-mats and local ORing units reduce line loading and data skew, enabling faster read operations in static random access memory.
Isolation gate disconnects cross-coupled inverters during read cycles, preventing latch feedback disruption from transistor variability.
Page buffering circuit latches damaged bitline addresses during voltage testing, eliminating time-consuming address calculation processes.
An offset voltage mechanism reduces power consumption by suppressing leak current in non-selected cross-point memory cells.
Non-multiplexed address pins disable defective DRAM quadrants, reducing manufacturing waste from discarded chips.
A semiconductor system adjusts operation periods based on active command counts to manage timing constraints.
A self-reference sensing scheme generates a cell-specific voltage reference using two access lines to compare read voltages.
Segmenting main decoders per bank while sharing predecoders resolves the trade-off between storage capacity and decoder area.
Block refreshing applies specific voltage potentials to memory cells, reducing power consumption and space occupation while maintaining data integrity.
Time-windowed access monitoring detects row hammer risks and applies dynamic throttling rates to prevent data corruption in adjacent memory rows.
A semiconductor internal voltage generator uses a sensor to detect voltage levels and selectively enable drivers.
A data alignment circuit aligns sequential odd and even data using distinct strobe signal edges to support high-speed prefetch operations.
A series continuous time linear equalizer uses passive LC circuits to filter signals without active amplification.
Dynamic ID code allocation reassigns identifiers to functional chips in a stack package, ensuring continued system operation after a single chip failure.