Data movement between the processing units and the memory in traditional von Neumann architecture is creating the “memory wall” problem. To bridge the gap, two approaches, the memory-rich processor (more on-chip memory) and the compute-capable memory (processing-in-memory) have been studied. However, the first one has strong computing capability but limited memory capacity/bandwidth, whereas the second one is the exact the opposite.To address the challenge, we propose DRISA, a DRAM-based Reconfigurable In-Situ Accelerator architecture, to provide both powerful computing capability and large memory capacity/bandwidth. DRISA is primarily composed of DRAM memory arrays, in which every memory bitline can perform bitwise Boolean logic operations (such as NOR). DRISA can be reconfigured to compute various functions with the combination of the functionally complete Boolean logic operations and the proposed hierarchical internal data movement designs. We further optimize DRISA to achieve high performance by simultaneously activating multiple rows and subarrays to provide massive parallelism, unblocking the internal data movement bottlenecks, and optimizing activation latency and energy. We explore four design options and present a comprehensive case study to demonstrate significant acceleration of convolutional neural networks. The experimental results show that DRISA can achieve 8.8× speedup and 1.2× better energy efficiency compared with ASICs, and 7.7× speedup and 15× better energy efficiency over GPUs with integer operations.CCS CONCEPTS• Hardware → Dynamic memory; • Computer systems organization → reconfigurable computing; Neural networks;

DRISA: a DRAM-based Reconfigurable In-Situ Accelerator

A semiconductor memory device comprises memory cells, a bitline connected to the memory cells, a read circuit including a precharge circuit, and a first transistor connected between the bitline and the read circuit, wherein a first voltage is applied to a gate of the first transistor when the precharge circuit precharges the bitline, and a second voltage which is different from the first voltage is applied to the gate of the first transistor when the read circuit senses a change in a voltage of the bitline.

Semiconductor memory device.

A disclosed example apparatus includes a row address register (412) to store a row address corresponding to a row (608) in a memory array (602). The example apparatus also includes a row decoder (604) coupled to the row address register to assert a signal on a wordline (704) of the row after the memory receives a column address. In addition, the example apparatus includes a column decoder (606) to selectively activate a portion of the row based on the column address and the signal asserted on the wordline.

Memory access methods and apparatus

The unprecedented growth in deep neural networks (DNN) size has led to massive amounts of data movement from off-chip memory to on-chip processing cores in modern machine learning (ML) accelerators. Compute-in-memory (CIM) designs performing analog DNN computations within a memory array, along with peripheral mixed-signal circuits, are being explored to mitigate this memory-wall bottleneck: consisting of memory latency and energy overhead. Embedded-dynamic random-access memory (eDRAM) [1], [2], which integrates the 1T1C (T=Transistor, C=Capacitor) DRAM bitcell monolithically along with high-performance logic transistors and interconnects, can enable custom CIM designs. It offers the densest embedded bitcell, a low pJ/bit access energy, a low soft error rate, high-endurance, high-performance, and high-bandwidth: all desired attributes for ML accelerators. In addition, the intrinsic charge sharing operation during a dynamic memory access can be used effectively to perform analog CIM computations: by reconfiguring existing eDRAM columns as charge domain circuits, thus, greatly minimizing peripheral circuit area and power overhead. Configuring a part of eDRAM as a CIM engine (for data conversion, DNN computations, and weight storage) and retaining the remaining part as a regular memory (for inputs, gradients during training, and non-CIM workload data) can help to meet the layer/kernel dependent variable storage needs during a DNN inference/training step. Thus, the high cost/bit of eDRAM can be amortized by repurposing part of existing large capacity, level-4 eDRAM caches [7] in high-end microprocessors, into large-scale CIM engines.

16.2 eDRAM-CIM: Compute-In-Memory Design with Reconfigurable Embedded-Dynamic-Memory Array Realizing Adaptive Data Converters and Charge-Domain Computing

A semiconductor structure and a method for fabricating the semiconductor structure include or provide a field effect device that includes a spacer shaped contact via. The spacer shaped contact via preferably comprises a spacer shaped annular contact via that is located surrounding and separated from an annular spacer shaped gate electrode at the center of which may be located a non-annular and non-spacer shaped second contact via. The annular gate electrode as well as the annular contact via and the non-annular contact via may be formed sequentially in a self-aligned fashion while using a single sacrificial mandrel layer.

Field effect device structure including self-aligned spacer shaped contact

A decoder circuit includes a pulse powered stage having a plurality of fan-in inputs thereto, a dynamic stage fed by the pulse powered stage, and a replica node selectively coupled to an output node of the pulse powered stage by a pass device. The pass device and the dynamic stage are controlled by a clock signal so as to enable a self-timed evaluation of the pulse-powered stage with a clocked enablement of the dynamic stage.

Fast pulse powered NOR decode apparatus for semiconductor devices

IBM introduced trench capacitor eDRAM into its high performance microprocessors beginning with 45nm and Power 7 [1] to provide a higher density cache without chip crossings. Whereas the 45 and 32nm designs employ a micro sense amplifier [2] and three-level bitline hierarchy, the design implemented for 22nm utilizes a higher gain sense amplifier and two-level bitline architecture that together provide significant reductions in area, latency, and power. This 22nm design style has been migrated into a 14nm FinFET [3] learning vehicle, complete with an ABIST engine, wordline charge pumps (VPP and VWL), and padcage interface circuitry.

17.4 A 14nm 1.1Mb embedded DRAM macro with 1ns access

A clock control method and apparatus are provided employing a clock control circuit which generates an array clock for a memory array from a system clock and a reset control signal. The reset control signal is one of a plurality of input control signals to the clock control circuit. When the system clock is below a predefined frequency threshold, the reset control signal is an array tracking reset signal, wherein the active pulse width of the array clock is system clock frequency independent, and when the system clock is above the predefined frequency threshold, the reset control signal is a mid-cycle reset signal, meaning that the active pulse width of the array clock is system clock frequency dependent. A bypass signal is provided as a third input control signal, which when active causes the clock control circuit to output an array clock which mirrors the system clock.

Clock control method and apparatus for a memory array

A method and structure is provided for an improved body contact layout for semiconductor-on-insulator (SOI) devices. In one embodiment, an insulated gate field effect transistor and method for fabrication of such a transistor is provided. The insulated gate field effect transistor includes a source, a drain, and a channel formed in a layer of a single-crystal semiconductor. The layer is disposed over and insulated from a bulk semiconductor layer of a substrate by a buried insulator layer. A gate conductor is disposed in an annular pattern overlying the channel, such that the gate conductor surrounds one of the source and drain disposed to the inside of the annular pattern, the other of the source and drain being disposed to the outside of the annular pattern. A second conductive pattern is connected to the annular pattern of the gate conductor. A conductive body contact is also disposed in the vicinity of the second conductive pattern.

Body contact layout for semiconductor-on-insulator devices

A decoder circuit includes a pulse powered stage having a plurality of fan-in inputs thereto, a dynamic stage fed by the pulse powered stage, and a replica node selectively coupled to an output node of the pulse powered stage by a pass device. The pass device and the dynamic stage are controlled by a clock signal so as to enable a self-timed evaluation of the pulse-powered stage with a clocked enablement of the dynamic stage. A pull up device restores the dynamic stage to a precharged condition, the pull up device controlled by a second clock signal independent of the first clock signal.

Kenneth J. Reyer

Papers

Fast pulse powered NOR decode apparatus for semiconductor devices

17.4 A 14nm 1.1Mb embedded DRAM macro with 1ns access

Clock control method and apparatus for a memory array

Body contact layout for semiconductor-on-insulator devices

Fast pulse powered NOR decode apparatus with pulse stretching and redundancy steering