摘要:
In a first aspect, a first method is provided for transferring data using an Infiniband (IB) protocol. The first method includes the steps of (1) receiving a non-IB packet having header data and payload data at a first node of a computer system; and (2) modifying data in the non-IB packet to convert the non-IB packet to an IB packet having header data and payload data. The header data of the non-IB packet is not included in the payload data of the IB packet resulting from the conversion. Numerous other aspects are provided.
摘要:
Methods and apparatus that allow lost packets on one virtual channel to be retried without requiring all subsequently issued packets, sent over other virtual channels, to be retried. In other words, packet retries may be performed on a “per virtual channel” basis. As a result, other virtual channels, not experiencing lost packets, may not suffer reductions in their bandwidth due to a lost packet occurring on another virtual channel.
摘要:
Embodiments of the present invention provide methods, a module, and a system for calculating a credit limit for an interface capable of receiving multiple packets simultaneously. Generally, the multiple packets are simultaneously received at an interface on the second device, each packet being one of a plurality of packet types, and a flow control credit limit to be transmitted to the first device is adjusted based on the combination of packet types of the simultaneously received packets.
摘要:
Methods and apparatus that may be utilized in an effort to ensure bytes of data sequentially received on multiple single-byte data paths with properly aligned when presented on a multi-byte interface are provided. A sufficient number of bytes received each channel may be stored (e.g., buffered) and examined to properly match data from each single-byte path. Once matched, the data may be output in a proper order on the multi-byte interface, for example, via some type of multiplexor arrangement.
摘要:
Computer systems with direct updating of cache (e.g., primary L1 cache) memories of a processor, such as a central processing unit (CPU) or graphics processing unit (GPU). Special addresses are reserved for high speed memory. Memory access requests involving these reserved addresses are routed directly to the high speed memory. Memory access requests not involving these reserved addresses are routed to memory external to the processor.
摘要:
An enclosure is removably securable to a support member including a first part of a releasable fastener. The enclosure includes a panel of material sized to enclose the support member. A flexible flap is affixed to the panel, and a second part of the releasable fastener is affixed to the flexible flap. The second part of the releasable fastener is positioned for engagement with the first part when the panel encloses the support member. By affixing the second part of the releasable fastener to the flexible flap, the position of the fastener is adjustable to accommodate panel shrinkage, manufacturing tolerances and the like.
摘要:
A method and apparatus are disclosed for the manufacture of a colored chip or colored swatch bearing sheet, i.e. a color chart, comprising a base sheet bearing an array of adhesively attached colored chips in rows with indicia identifying the color of each chip with a particular paint. The base sheets are fed automatically from a stack of sheets to an adhesive applying means which applies spots of adhesive to the sheet at swatch receiving areas on the sheet. Narrow webs or ribbons of swatch material are severed to form the individual swatches or color chips which then are transferred automatically to a passing sheet and applied thereto. The sheets pass through a series of swatch forming and applying stations with a row of swatches being applied to each sheet at each of the respective stations. Preferably, the sheets are fed continuously and at a relativey fast rate of speed. The arrangement and size of the swatches may be easily changed to allow the apparatus to be used for various types of color charts.
摘要:
A method and apparatus that allow packet based communication transactions between devices over an interconnect bus to be captured to measure performance. Performance metrics may be determined by capturing events at various locations as they pass through the system. Performance may be verified at run time by computing performance metrics for captured events and comparing such metrics to predefined performance ranges and/or self learned performance ranges. Furthermore, embodiments of the present invention provide for dynamic tailoring of bus traffic to generate potential failing conditions. For some embodiments, performance verification as described herein may be performed in a simulation environment.
摘要:
In a first aspect, a first method is provided. The first method includes the steps of (1) transmitting data on a bus, wherein data is presented on the bus using varying widths; (2) configuring a cyclic redundancy check (CRC) to be performed on the data based on the manner in which the data is presented on the bus; and (3) performing the CRC on the data. Numerous other aspects are provided.
摘要:
In a first aspect, a first method is provided. The first method includes the steps of (1) transmitting data on a bus, wherein data is presented on the bus using varying widths; (2) configuring a cyclic redundancy check (CRC) to be performed on the data based on the manner in which the data is presented on the bus; and (3) performing the CRC on the data. Numerous other aspects are provided.