Abstract: A system including a stack of two or more layers of volatile memory, such as layers of a 3D stacked DRAM memory, places data in the stack based on a temperature or a refresh rate. When a threshold is exceeded, data are moved from a first region to a second region in the stack, the second region having one or both of a second temperature lower than a first temperature of the first region or a second refresh rate lower than a first refresh rate of the first region.

Abstract: A compute unit configured to execute multiple threads in parallel is presented. The compute unit includes one or more single instruction multiple data (SIMD) units and a fetch and decode logic. The SIMD units have differing numbers of arithmetic logic units (ALUs), such that each SIMD unit can execute a different number of threads. The fetch and decode logic is in communication with each of the SIMD units, and is configured to assign the threads to the SIMD units for execution based on such differing numbers of ALUs.

Abstract: A plurality of thermal electric cooler (TEC) elements are formed in a TEC grid structure. Control logic dynamically varies a supply current supplied to each TEC element (or group of TEC elements) in the TEC grid based on changes in power density respectively associated with areas cooled by each of the TEC elements or group of TEC elements.

Abstract: A processing system dynamically scales at least one of voltage and frequency at a subset of a plurality of compute units of a graphics processing unit (GPU) based on characteristics of a kernel or workload to be executed at the subset. A system management unit for the processing system receives a compute unit mask, designating the subset of a plurality of compute units of a GPU to execute the kernel or workload, and workload characteristics indicating the compute-boundedness or memory bandwidth-boundedness of the kernel or workload from a central processing unit of the processing system. The system management unit determines a dynamic voltage and frequency scaling policy for the subset of the plurality of compute units of the GPU based on the compute unit mask and the workload characteristics.

Abstract: A system including a stack of two or more layers of volatile memory, such as layers of a 3D stacked DRAM memory, places data in the stack based on a temperature or a refresh rate. When a threshold is exceeded, data are moved from a first region to a second region in the stack, the second region having one or both of a second temperature lower than a first temperature of the first region or a second refresh rate lower than a first refresh rate of the first region.

Abstract: An operating system (OS) of a processing system having a plurality of processor cores determines a cost associated with different mechanisms for performing a translation lookaside buffer (TLB) shootdown in response to, for example, a virtual address being remapped to a new physical address, and selects a TLB shootdown mechanism to purge outdated or invalid address translations from the TLB based on the determined cost. In some embodiments, the OS selects an inter-processor interrupt (IPI) as the TLB shootdown mechanism if the cost associated with sending an IPI is less than a threshold cost. In some embodiments, the OS compares the cost of using an IPI as the TLB shootdown mechanism versus the cost of sending a hardware broadcast to all processor cores of the processing system as the shootdown mechanism and selects the shootdown mechanism having the lower cost.

Abstract: A method includes storing a first portion of a sparse triangular matrix in a local memory and launching a kernel for executing a set of workgroups. The first portion includes a plurality of row blocks, and each workgroup in the set of workgroups is associated with one of the plurality of row blocks. The method also includes, for each workgroup in the set of workgroups, solving the row block. The row block is solved by, for each row segment of a first subset of row segments in the row block, calculating a partial sum for the row segment based on one or more matrix elements in the row segment, and writing the partial sum to a remote memory of a first remote processing unit prior to terminating the kernel.

Abstract: An operating system (OS) of a processing system having a plurality of processor cores determines a cost associated with different mechanisms for performing a translation lookaside buffer (TLB) shootdown in response to, for example, a virtual address being remapped to a new physical address, and selects a TLB shootdown mechanism to purge outdated or invalid address translations from the TLB based on the determined cost. In some embodiments, the OS selects an inter-processor interrupt (IPI) as the TLB shootdown mechanism if the cost associated with sending an IPI is less than a threshold cost. In some embodiments, the OS compares the cost of using an IPI as the TLB shootdown mechanism versus the cost of sending a hardware broadcast to all processor cores of the processing system as the shootdown mechanism and selects the shootdown mechanism having the lower cost.

Abstract: A processor utilizes instruction based sampling to generate sampling data sampled on a per instruction basis during execution of an instruction. The sampling data indicates what processor hardware was used due to the execution of the instruction. Software receives the sampling data and generates an estimate of energy used by the instruction based on the sampling data. The sampling data may include microarchitectural events and the energy estimate utilizes a base energy amount corresponding to the instruction executed along with energy amounts corresponding to the microarchitectural events in the sampling data. The sampling data may include switching events associated with hardware blocks that switched due to execution of the instruction and the energy estimate for the instruction is based on the switching events and capacitance estimates associated with the hardware blocks.

Abstract: A processing apparatus is provided that includes a plurality of memory regions each corresponding to a memory address and configured to store data associated with the corresponding memory address. The processing apparatus also includes an accelerated processing device in communication with the memory regions and configured to determine a request to allocate an initial memory buffer comprising a number of contiguous memory regions, create a new memory buffer comprising one or more additional memory regions adjacent to the contiguous memory regions of the initial memory buffer, assign one or more values to the one or more additional memory regions and detect a change to the one or more values at the one or more additional memory regions.

Abstract: A processing device includes a first memory that includes a context buffer. The processing device also includes a processor core to execute threads based on context information stored in registers of the processor core and a memory controller to selectively move a subset of the context information between the context buffer and the registers based on one or more latencies of the threads.

Abstract: Methods, devices, and systems for capturing an accuracy of an instruction executing on a processor. An instruction may be executed on the processor, and the accuracy of the instruction may be captured using a hardware counter circuit. The accuracy of the instruction may be captured by analyzing bits of at least one value of the instruction to determine a minimum or maximum precision datatype for representing the field, and determining whether to adjust a value of the hardware counter circuit accordingly. The representation may be output to a debugger or logfile for use by a developer, or may be output to a runtime or virtual machine to automatically adjust instruction precision or gating of portions of the processor datapath.

Abstract: A processing apparatus is provided that includes a plurality of memory regions each corresponding to a memory address and configured to store data associated with the corresponding memory address. The processing apparatus also includes an accelerated processing device in communication with the memory regions and configured to determine a request to allocate an initial memory buffer comprising a number of contiguous memory regions, create a new memory buffer comprising one or more additional memory regions adjacent to the contiguous memory regions of the initial memory buffer, assign one or more values to the one or more additional memory regions and detect a change to the one or more values at the one or more additional memory regions.

Abstract: Techniques described herein improve processor performance in situations where a large number of system service requests are being received from other devices. More specifically, upon detecting that certain operating conditions that indicate a processor slowdown are present, the processor performs one or more system service adjustment techniques. These techniques include throttling (reducing the rate of handling) of such requests, coalescing (grouping multiple requests into a single group) the requests, disabling microarchitctural structures (such as caches or branch prediction units) or updates to those structures, and prefetching data for or pre-performing these requests. Each of these adjustment techniques helps to reduce the number of and/or workload associated with servicing requests for system services.

Abstract: A method of multiplication of a sparse matrix and a vector to obtain a new vector and a system for implementing the method are claimed. Embodiments of the method are intended to optimize the performance of sparse matrix-vector multiplication in highly parallel processors, such as GPUs. The sparse matrix is stored in compressed sparse row (CSR) format.

Abstract: Methods, devices, and systems for capturing an accuracy of an instruction executing on a processor. An instruction may be executed on the processor, and the accuracy of the instruction may be captured using a hardware counter circuit. The accuracy of the instruction may be captured by analyzing bits of at least one value of the instruction to determine a minimum or maximum precision datatype for representing the field, and determining whether to adjust a value of the hardware counter circuit accordingly. The representation may be output to a debugger or logfile for use by a developer, or may be output to a runtime or virtual machine to automatically adjust instruction precision or gating of portions of the processor datapath.

Abstract: A processing device includes a first memory that includes a context buffer. The processing device also includes a processor core to execute threads based on context information stored in registers of the processor core and a memory controller to selectively move a subset of the context information between the context buffer and the registers based on one or more latencies of the threads.

Abstract: A system and method for efficiently performing program instrumentation. A processor processes instructions stored in a memory. The processor allocates a memory region for the purpose of creating “random branches” in the computer code utilizing existing memory access instructions. When the processor processes a given instruction, the processor both accesses a first location in the memory region and may determine a condition is satisfied. In response, the processor generates an interrupt. The corresponding interrupt handler may transfer control flow from the computer program to instrumentation code. The condition may include a pointer storing an address pointing to locations within the memory region equals a given address after the point is updated. Alternatively, the condition may include an updated data value stored in a location pointed to by the given address equals a threshold value.

Abstract: A system and method for efficiently performing program instrumentation. A processor processes instructions stored in a memory. When the processor processes a given instruction of a given instruction type, the processor updates a corresponding performance counter. When the performance counter reaches a threshold, the processor generates an interrupt and compares a location of the given instruction with stored locations in a given list. If a match is not found, then the processor processes an instruction following the given instruction in the computer program without processing intermediate instrumentation code. If a match is found, then the processor processes instrumentation code. Regardless of whether or not the instrumentation code is processed, when control flow returns to the computer program, the corresponding performance counter is initialized with a random value.

Abstract: A method of multiplication of a sparse matrix and a vector to obtain a new vector and a system for implementing the method are claimed. Embodiments of the method are intended to optimize the performance of sparse matrix-vector multiplication in highly parallel processors, such as GPUs. The sparse matrix is stored in compressed sparse row (CSR) format.