Methods, systems, and apparatus for combined or separate implementation of coarse-to-fine neural architecture search (NAS), two-phase block NAS, variable hardware prediction, and differential hardware design are provided and described. A variable predictor is trained, as described herein. Then, a controller or policy may be used to iteratively modify a neural network architecture along dimensions formed by neural network architecture parameters. The modification is applied to blocks (e.g., subnetworks) within the neural network architecture. In each iteration, the remainder of the neural network architecture parameters are modified and learned with a differential NAS method. The training process is performed with two-phase block NAS and incorporates a variable hardware predictor to predict power, performance, and area (PPA) parameters. The hardware parameters may be learned as well using the variable hardware predictor.

In general, embodiments of the present disclosure provide methods, apparatus, systems, computer program products, computing devices, computing entities, and/or the like for altering a design of a hardware intellectual property (IP). In accordance with various embodiments, a representation of the design of the hardware IP is converted to generate a control and data flow graph (CDFG) for the design. An entropy analysis of the CDFG is conducted to identify one or more control paths and/or data paths for removal. Responsive to identifying control path(s) for removal, control logic for the control path(s) is removed from the design and replaced with first reconfigurable logic. Responsive to identifying data path(s) for removal, datapath logic for the data path(s) is removed from the design and replaced with second reconfigurable logic. Logic synthesis is then performed on the design, along with verification to check functional correctness of the design of the hardware.

Reduced-power dynamic data circuits with wide-band energy recovery are described herein. In one embodiment, a circuit system comprises at least one sub-circuit in which at least one of the sub-circuits includes a capacitive output node that is driven between low and high states in a random manner for a time period and an inductive circuit path coupled to the capacitive output node. The inductive circuit path includes a transistor switch and an inductor connected in series to discharge and recharge the output node to a bias supply. A pulse generator circuit generates a pulse width that corresponds to a timing for driving the output node.

An integrated circuit includes a plurality of logic cells arranged in a first row extending in a first direction and including different types of active areas extending in the first direction, a filler cell arranged in a second row adjacent to the first row in a second direction orthogonal to the first direction and extending in the first direction, and a first routing wiring line arranged in the second row and connecting a first logic cell and a second logic cell apart from each other by a first distance among the plurality of logic cells. A height of the first row is different from a height of the second row.

Methods, systems and media for simulating or analyzing voltage drops in a power distribution network can use an incremental approach to define a portion of a design around a victim to capture a sufficient collection of aggressors that cause appreciable voltage drop on the victim, and then an incremental simulation of just the portion can be performed rather than computing simulated voltage drops across the entire design. This approach can be both computationally efficient and can limit the size of the data used in simulating dynamic voltage drops in the power distribution network. Multiple different portions can be simulated separately in separate processing cores or elements. In one embodiment, a system can provide options of user selected constraints for the simulation to provide better accuracy or use less memory. Better accuracy will normally use a larger set of aggressors for each victim at the expense of using more memory.

The invention relates to an optimal allocation method for stored energy coordinating electric vehicles (EVs) to participate in auxiliary service market (ASM), including the following steps: 1. Predict the reported capacity of daily 96 points for EVs to participate in the ASM by least square support vector machine (LSSVM). 2. Fit the daily total load distribution of EVs. 3. Determine the error distribution between the reported capacity and the actual response capacity, and simulate the total daily load capacity of EVs in the future with Monte Carlo method. 4. Calculate the energy storage capacity required by EVs daily participating in ASM. 5. Build the objective function to minimize the scheduling risk of auxiliary service. 6. Solve the energy storage model in step 5 with particle swarm optimization (PSO), and output the configuration results of optimal energy storage capacity and energy storage power. The invention can improve the adjustable capacity of EVs participating in ASM.

Systems and methods for context aware circuit design are described herein. A method includes: identifying at least one cell to be designed into a circuit; identifying at least one context parameter having an impact to layout dependent effect of the circuit; generating, for each cell and for each context parameter, a plurality of abutment environments associated with the cell; estimating, for each cell and each context parameter, a sensitivity of at least one electrical property of the cell to the context parameter by generating a plurality of electrical property values of the cell under the plurality of abutment environments; and determining whether each context parameter is a key context parameter for a static analysis of the circuit, based on the sensitivity of the at least one electrical property of each cell and based on at least one predetermined threshold.

A method of generating a circuit model used to simulate an integrated circuit may include generating first feature element data and second feature element data by classifying feature data of a target semiconductor device according to measurement conditions, generating first target data and second target data by preprocessing the first feature element data and the second feature element data, respectively, generating a first machine learning model using the first target data and extracting a second machine learning model using the second target data, and generating the circuit model used to simulate the integrated circuit using the first machine learning model and the second machine learning model.

A method includes generating an integrated circuit (IC) layout design and manufacturing an IC based on the IC layout design. Generating the IC layout design includes generating a pattern of a first shallow trench isolation (STI) region and a pattern of a through substrate via (TSV) region within the first STI region; a pattern of a second STI region surrounding the first STI region, the second STI region includes a first and second layout region, the second layout region being separated from the first STI region by the first layout region, first active regions of a group of dummy devices being defined within the first layout region, and second active regions of a group of active devices being defined within the second layout region; and patterns of first gates of the group of dummy devices in the first layout region, each of the first active regions having substantially identical dimension in a first direction.

An integrated circuit layout is provided. The integrated circuit layout includes one or more first cell rows partially extending across a space arranged for an integrated circuit layout along a first direction. Each of the one or more first cell rows has a first height along a second direction perpendicular to the first direction. The integrated circuit layout includes one or more third cell rows partially extending across the space along the first direction. Each of the one or more third cell rows has a second height along the second direction, the second height different from the first height.