International Business Machines Corporation

Osher Yifrach

Erez Barak

Giora Biran

Nicol Hofmann

Amir Turi

Cédric Lichtenau

Shlomit Koyfman

Cedric Lichtenau

Michael Klein

Eyal Naor

Ziv Nevo

Shiri Moran

Ido Rozenberg

A system for processing instructions with extended results includes a first instruction execution unit having a first result bus for execution of processor instructions. The system further includes a second instruction execution unit having a second result bus for execution of processor instructions. The first instruction execution unit is configured to selectively send a portion of results calculated by the first instruction execution unit to the second instruction execution unit during prosecution of a processor instruction if the second instruction execution unit is not used for executing the processor instruction and if the received processor instruction produces a result having a data width greater than the width of the first result bus. The second instruction execution unit is configured to receive the portion of results calculated by the first instruction execution unit and put the received results on the second results bus.

Reusing adjacent SIMD unit for fast wide result generation

The present disclosure provides reducing clock power consumption of a computer processor by simulating, in a baseline simulation of a computer processor design using a software model of the computer processor design, performance of an instruction by the computer processor design, to produce a baseline result of the instruction, and identifying a circuit of the computer processor design that receives a clock signal during performance of the instruction, and in a comparison simulation of the computer processor design using the software model of the computer processor design, simulating performance of the instruction by the computer processor design while injecting a corruption signal into the circuit, to produce a comparison result of the instruction, and designating the circuit for clock gating when processing the instruction, if the comparison result of the instruction is identical to the baseline result of the instruction.

Reducing clock power consumption of a computer processor

Examples of techniques for modifying testing tools are described herein. An example computer-implemented method includes receiving, via a processor, a netlist comprising a complex coverage event that depends on a singular independent signal. The method includes detecting, via the processor, that complex coverage event can be separated into the singular independent signal and a logic state based on a structural logic analysis. The method also includes modifying, via the processor, a testing tool to test the netlist based on the singular independent signal.

Testing netlists based on singular independent signals

Reducing clock power consumption of a computer processor by simulating, in a baseline simulation of a computer processor design using a software model of the computer processor design, performance of an instruction by the computer processor design, to produce a baseline result of the instruction, and identifying a circuit of the computer processor design that receives a clock signal during performance of the instruction, and in a comparison simulation of the computer processor design using the software model of the computer processor design, simulating performance of the instruction by the computer processor design while injecting a corruption signal into the circuit, to produce a comparison result of the instruction, and designating the circuit for clock gating when processing the instruction, if the comparison result of the instruction is identical to the baseline result of the instruction.

A computer-implemented method, computerized apparatus and computer program product for modified design debugging using differential trace back. An indication of an interface signal in a time unit in an execution resulting in a value miscompare between a design and a modification thereof is obtained. For each of the design and the modification, a data record detailing each signal value in each time unit, and a structure description detailing all components and interconnections thereamong, are obtained. A suspect root cause of the value miscompare is traced back from the interface signal in the time unit, the tracing back comprising comparing values in the data records of candidate signals selected based on the data records and the structure descriptions.

Modified design debugging using differential trace back

Checking a computer processor design for soft error handling. A baseline simulation of a computer processor design is monitored to identify a target processing cycle of the baseline simulation during which a predefined event occurs during the baseline simulation. The baseline simulation is performed in accordance with a software model of the computer processor design, and the event is associated with processing an instruction that directly involves a predefined error injection target. A test simulation of the computer processor design is performed in accordance with the software model of the computer processor design. An error is injected into the predefined error injection target during a target processing cycle of the test simulation. A determination is made as to whether the error is detected by error-checking logic of the computer processor design.

Growing community of inventors

Modiin, Israel

Osher Yifrach