摘要:
Techniques involving the transfer of signals across interconnection media are disclosed. For instance, an apparatus may include a configuration module and a driver module. The configuration module establishes an operational mode setting from a first operational mode and a second operational mode. The driver module may operate according to this setting. For instance, according to the first operational mode, the driver module continually drives an interconnection medium. However, for the second operational mode, the driver module drives the interconnection medium when it receives an input signal and otherwise refrains from driving the interconnection medium. According to the first operational mode, a termination module may couple a pull-up resistance to the interconnection medium. However, for the second operational mode, the termination module does not couple such a resistance to the interconnection medium.
摘要:
A non-volatile random access memory (NVRAM) is used in a computer system to perform multiple roles in a platform storage hierarchy, specifically, to replace traditional mass storage that is accessible by an I/O. The computer system includes a processor to execute software and a memory coupled to the processor. At least a portion of the memory comprises a non-volatile random access memory (NVRAM) that is byte-rewritable and byte-erasable by the processor. The system further comprises a memory controller coupled to the NVRAM to perform a memory access operation to access the NVRAM in response to a request from the software for access to a mass storage.
摘要:
In one embodiment, the present invention includes a method for receiving a request in a router from a first endpoint coupled to the router, where the request is for an aggregated completion. In turn, the router can forward the request to multiple target agents, receive a response from each of the target agents, and consolidate the responses into an aggregated completion. Then, the router can send the aggregated completion to the first endpoint. Other embodiments are described and claimed.
摘要:
In one embodiment, a method includes determining whether producer-consumer ordering rules have been met for a first transaction to be sent from a source agent to a target agent via a fabric, and if so a first request for the first transaction is sent from the source agent to the fabric in a first clock cycle. Then a second request can be sent from the source agent to the fabric for a second transaction in a pipelined manner. Other embodiments are described and claimed.
摘要:
In one embodiment, the present invention includes a method for receiving a request in a router from a first endpoint coupled to the router, where the request is for an aggregated completion. In turn, the router can forward the request to multiple target agents, receive a response from each of the target agents, and consolidate the responses into an aggregated completion. Then, the router can send the aggregated completion to the first endpoint. Other embodiments are described and claimed.
摘要:
In one embodiment, the present invention includes a method for receiving a request in a router from a first endpoint coupled to the router, where the request is for an aggregated completion. In turn, the router can forward the request to multiple target agents, receive a response from each of the target agents, and consolidate the responses into an aggregated completion. Then, the router can send the aggregated completion to the first endpoint. Other embodiments are described and claimed.
摘要:
An enhanced general input/output communication architecture, protocol and related methods are presented. In one embodiment, a method is described comprising receiving a datagram at general input/output (GIO) interface from a remote GIO interface coupled through a GIO link, validating content of one or more packets embedded within the received datagram, and issuing an acknowledgment to the remote GIO interface that the datagram was successfully received on positive validation of the datagram before promoting the embedded packets to a transaction layer of the GIO interface. Other embodiments are also described.
摘要:
According to the invention, systems, apparatus and methods are disclosed for automatically replacing semiconductor based memory in a system while that system is operating. Replacement is accomplished by blocking access to the memory devices to be swapped, copying the contents of the faulty device to the replacement device, swapping the IDs of the two devices, and re-enabling access to the replaced device. This replacement is triggered by interrupts from the error detection logic. After replacement, the system automatically checks the faulty device to determine its suitability for use as a spare in the future. The determination is made by repeatedly writing and reading a pseudo-random pattern into the device and logging any errors.
摘要:
A computer graphics processor capable of reading from, and writing to, virtual memory. The invention provides a graphics processing unit which includes, among other things, an graphic processor in the form of an address generator which retrieves data from memory locations, and writes data to memory locations. The address generator retrieves data from memory locations memory access request directly to a memory control unit, which retrieves the contents of the memory location. Prior to issuing the request, the address generator sends the address to a virtual translation unit, which translates the virtual address to a physical address. The virtual translation/FIFO control unit also contains three translation buffers, in which are stored the most recently accessed virtual addresses, which, in many situations, enables the virtual translation/FIFO control unit to translate the virtual address using less memory accesses.