Methods, apparatus, systems, and articles of manufacture are disclosed to debug a hardware accelerator such as a neural network accelerator for executing Artificial Intelligence computational workloads. An example apparatus includes a core with a core input and a core output to execute executable code based on a machine-learning model to generate a data output based on a data input, and debug circuitry coupled to the core. The debug circuitry is configured to detect a breakpoint associated with the machine-learning model, compile executable code based on at least one of the machine-learning model or the breakpoint. In response to the triggering of the breakpoint, the debug circuitry is to stop the execution of the executable code and output data such as the data input, data output and the breakpoint for debugging the hardware accelerator.

A processor includes a central processing unit (CPU) and diagnostic monitoring circuitry. The diagnostic monitoring circuitry is coupled to the CPU. The diagnostic monitoring circuitry includes a monitoring and cyclic redundancy check (CRC) computation unit. The monitoring and CRC computation unit is configured to detect execution of a diagnostic program by the CPU, and to compute a plurality of CRC values. Each of CRC values corresponds to processor values retrieved from a given register of the CPU or from a bus coupling the CPU to a memory and peripheral subsystem while the CPU executes the diagnostic program.

A processor includes a central processing unit (CPU) and diagnostic monitoring circuitry. The diagnostic monitoring circuitry is coupled to the CPU. The diagnostic monitoring circuitry includes a monitoring and cyclic redundancy check (CRC) computation unit. The monitoring and CRC computation unit is configured to detect execution of a diagnostic program by the CPU, and to compute a plurality of CRC values. Each of CRC values corresponds to processor values retrieved from a given register of the CPU or from a bus coupling the CPU to a memory and peripheral subsystem while the CPU executes the diagnostic program.

Examples of electronic devices are described herein. In some examples, an electronic device may include a communication interface to receive information from a peripheral device. In some examples, the electronic device may include logic circuitry to determine a target latency tolerance reporting (LTR) value based on the information via a machine learning model.

Examples of electronic devices are described herein. In some examples, an electronic device may include a communication interface to receive information from a peripheral device. In some examples, the electronic device may include logic circuitry to determine a target latency tolerance reporting (LTR) value based on the information via a machine learning model.

Methods, apparatus, systems, and articles of manufacture are disclosed to debug a hardware accelerator such as a neural network accelerator for executing Artificial Intelligence computational workloads. An example apparatus includes a core with a core input and a core output to execute executable code based on a machine-learning model to generate a data output based on a data input, and debug circuitry coupled to the core. The debug circuitry is configured to detect a breakpoint associated with the machine-learning model, compile executable code based on at least one of the machine-learning model or the breakpoint. In response to the triggering of the breakpoint, the debug circuitry is to stop the execution of the executable code and output data such as the data input, data output and the breakpoint for debugging the hardware accelerator.

Methods, apparatus, systems, and articles of manufacture are disclosed to debug a hardware accelerator such as a neural network accelerator for executing Artificial Intelligence computational workloads. An example apparatus includes a core with a core input and a core output to execute executable code based on a machine-learning model to generate a data output based on a data input, and debug circuitry coupled to the core. The debug circuitry is configured to detect a breakpoint associated with the machine-learning model, compile executable code based on at least one of the machine-learning model or the breakpoint. In response to the triggering of the breakpoint, the debug circuitry is to stop the execution of the executable code and output data such as the data input, data output and the breakpoint for debugging the hardware accelerator.

Embodiments of information handling systems (IHSs) and computer-implemented methods are provided herein for testing system memory (or another volatile memory component) of an IHS. In the disclosed embodiments, memory testing is performed automatically: (a) during the pre-boot phase each time a new page of memory is allocated for the first time after a system boot, and (b) during OS runtime each time a read command is received and/or an event is detected. By proactively testing each page of memory, as the page is allocated but before information is stored therein, the systems and methods disclosed herein prevent “bad” memory pages from being used.

Embodiments of information handling systems (IHSs) and computer-implemented methods are provided herein for testing system memory (or another volatile memory component) of an IHS. In the disclosed embodiments, memory testing is performed automatically: (a) during the pre-boot phase each time a new page of memory is allocated for the first time after a system boot, and (b) during OS runtime each time a read command is received and/or an event is detected. By proactively testing each page of memory, as the page is allocated but before information is stored therein, the systems and methods disclosed herein prevent “bad” memory pages from being used.

An electronic device, which can enter a debug mode, comprising: a plurality of buttons, wherein a layout of the buttons correspond to one of a first button layout and a second button layout; a processing circuit, configured to control the electronic device to enter a debug mode when at least two of the buttons are pressed to meet a predetermined button combination. The processing circuit controls the electronic device to perform a first test corresponding to the first button layout or to perform a second test corresponding to the second button layout to detect which one of the first button layout and the second button layout does the electronic device correspond to.