摘要:
Methods and apparatus are provided, as an aspect of a combined CPU/APD architecture system, for discovering and reporting properties of devices and system topology that are relevant to efficiently scheduling and distributing computational tasks to the various computational resources of a combined CPU/APD architecture system. The combined CPU/APD architecture unifies CPUs and APDs in a flexible computing environment. In some embodiments, the combined CPU/APD architecture capabilities are implemented in a single integrated circuit, elements of which can include one or more CPU cores and one or more APD cores. The combined CPU/APD architecture creates a foundation upon which existing and new programming frameworks, languages, and tools can be constructed.
摘要:
A computer based rendering system is provided. The computer based rendering system includes an abstraction mechanism to provide configuring instructions to two or more processors, the configuring instructions being operative to facilitate scene rendering. The configuring provides processor setup instructions to at least one driver. Each of the two or more processors renders a respective portion of the scene independent of the other of the processors.
摘要:
In a CPU, the CPU having multiple CPU cores, each core having a first machine specific register, a second machine specific register, and microcode which when executed causes a write notification to be issued to the physical address contained in the second machine specific register; receiving in the first machine specific register of a CPU core, a physical page table/page directory base address, receiving in the second machine specific register of the CPU core, a physical address pointing to a location controlled by the IOMMUv2, determining that a control register of the CPU core has been updated, and responsive to the determination that the control register has been updated, executing microcode in the CPU core that causes a write notification to be issued to the physical address contained in the second machine specific register, wherein the physical address is able to receive writes that affect IOMMUv2 page table invalidations.
摘要:
Methods and apparatus are provided, as an aspect of a combined CPU/APD architecture system, for discovering and reporting properties of devices and system topology that are relevant to efficiently scheduling and distributing computational tasks to the various computational resources of a combined CPU/APD architecture system. The combined CPU/APD architecture unifies CPUs and APDs in a flexible computing environment. In some embodiments, the combined CPU/APD architecture capabilities are implemented in a single integrated circuit, elements of which can include one or more CPU cores and one or more APD cores. The combined CPU/APD architecture creates a foundation upon which existing and new programming frameworks, languages, and tools can be constructed.
摘要:
In a CPU of the combined CPU/APD architecture system, the CPU having multiple CPU cores, each core having a first machine specific register for receiving a physical page table/page directory base address, a second machine specific register for receiving a physical address pointing to a location controlled by an IOMMUv2 that is communicatively coupled to an APD, and microcode which when executed causes a write notification to be issued to the physical address contained in the second machine specific register; receiving in the first machine specific register of a CPU core, a physical page table/page directory base address, receiving in the second machine specific register of the CPU core, a physical address pointing to a location controlled by the IOMMUv2, determining that a control register of the CPU core has been updated, and responsive to the determination that the control register has been updated, executing microcode in the CPU core that causes a write notification to be issued to the physical address contained in the second machine specific register, wherein the physical address is able to receive writes that affect IOMMUv2 page table invalidations.
摘要:
A method for rendering a scene across N number of processors is provided. The method includes evaluating performance statistics for each of the processors and establishing load rendering boundaries for each of the processors, the boundaries defining a respective portion of the scene. The method also includes dynamically adjusting the boundaries based upon the establishing and the evaluating.
摘要:
Initializing and configuring computer hardware with firmware stored in compressed form in nonvolatile semiconductor memory. Upon startup of the computer hardware, decompression software decompress the firmware, which is then stored in another memory. The computer hardware may be an adapter board (e.g., a graphics board connected to PCI bus), the nonvolatile semiconductor memory may be physically located on the adapter board, and the firmware may be firmware for initializing and configuring the adapter board. The decompression software may be stored in the same nonvolatile semiconductor memory as the firmware, and may be written in a machine-independent language (e.g., a Forth-based language). The compression technique used may include both run-length encoding and pattern compression, and may operate at the bit level.
摘要:
A method for providing two or more processors access to a single command buffer is provided. The method includes receiving instructions in the command buffer from a central processor, at least one of the instructions being designated for a particular one of the two or more processors. The method also includes sending the at least one instruction to only the particular processor.