摘要:
Examples are disclosed for probabilistic dynamic random access memory (DRAM) row repair. In some examples, using a row hammer limit for DRAM and a maximum activation rate for the DRAM a probabilistic row hammer detection value may be determined. The probabilistic row hammer detection value may then be used such that a probability is acceptably low that a given activation to an aggressor row of the DRAM causes the row hammer limit to be exceeded before a scheduled row refresh is performed on one or more victim rows associated with the aggressor row. Other examples are described and claimed.
摘要:
In one embodiment, the present invention includes a method of assigning a location within a shared variable for each of multiple threads and writing a value to a corresponding location to indicate that the corresponding thread has reached a barrier. In such manner, when all the threads have reached the barrier, synchronization is established. In some embodiments, the shared variable may be stored in a cache accessible by the multiple threads. Other embodiments are described and claimed.
摘要:
The apparatus and method described herein are for handling shared memory accesses between multiple processors utilizing lock-free synchronization through transactional-execution. A transaction demarcated in software is speculatively executed. During execution invalidating remote accesses/requests to addresses loaded from and to be written to shared memory are tracked by a transaction buffer. If an invalidating access is encountered, the transaction is re-executed. After a pre-determined number of times re-executing the transaction, the transaction may be re-executed non-speculatively with locks/semaphores.
摘要:
A method and apparatus for protecting a TLB's VPN from soft errors is described. On a TLB lookup, the incoming virtual address is used to CAM the TLB VPN. In parallel with this CAM operation, parity is computed on the incoming virtual address for the possible page sizes supported by the processor. If a matching VPN is found in the TLB, its payload is read out. The encoded page size is used to select which of the set of pre-computed virtual address parity to compare with the stored parity bit in the TLB entry. This has the advantage of removing the computation of parity on the TLB VPN from the critical path of the TLB lookup. Instead it is now in the TLB fill path.
摘要:
Microprocessor architecture for an address translation unit which provides two levels of memory management is described. Segmentation registers and an associated segmentation table in main memory provide a first level of memory management which includes attribute bits used for protection, priority, etc. A page cache memory and an associated page directory and page table in main memory provide a second level of management with independent protection on a page level.
摘要:
An improved method of transferring burst data in a microprocessor is described. The improvement lies in the burst ordering of the data items to be referenced. The original address is selected as the data item that the user initially wants to access. Subsequent addresses in the burst are generated according to a mathematical algorithm. The algorithm generates the remaining addresses as a function of the internal bus width, the external memory/bus line size and the original address. Using the burst sequence of the present invention, memories/buses of different widths can be smoothly coupled to a microprocessor having a fixed CPU bus size (e.g., 32 bits).
摘要:
A method and system for managing the operation of a computing system are described herein. The method includes determining a number of workloads on the computing system. The method also includes determining a number of performance-power states for each workload and a corresponding performance range and power consumption range for each performance-power state. The method further includes managing performance and power consumption of the computing system based on the performance-power states.
摘要:
Apparatus, systems, and methods to manage memory refresh operations are described. In one embodiment, an electronic device comprises a processor and memory controller logic to determine a memory refresh frequency for a memory system and transmit refresh commands to a refresh control logic in at least one memory bank coupled to the memory controller according to the memory refresh frequency. Other embodiments are also disclosed and claimed.
摘要:
The apparatus and method described herein are for handling shared memory accesses between multiple processors utilizing lock-free synchronization through transactional-execution. A transaction demarcated in software is speculatively executed. During execution invalidating remote accesses/requests to addresses loaded from and to be written to shared memory are tracked by a transaction buffer. If an invalidating access is encountered, the transaction is re-executed. After a pre-determined number of times re-executing the transaction, the transaction may be re-executed non-speculatively with locks/semaphores.
摘要:
The apparatus and method described herein are for handling shared memory accesses between multiple processors utilizing lock-free synchronization through transactional-execution. A transaction demarcated in software is speculatively executed. During execution invalidating remote accesses/requests to addresses loaded from and to be written to shared memory are tracked by a transaction buffer. If an invalidating access is encountered, the transaction is re-executed. After a pre-determined number of times re-executing the transaction, the transaction may be re-executed non-speculatively with locks/semaphores.