A decoder for decoding a data stream into which media data is coded has a mode switch configured to activate a low-complexity mode or a high-efficiency mode depending on the data stream, an entropy decoding engine configured to retrieve each symbol of a sequence of symbols by entropy decoding using a selected one of a plurality of entropy decoding schemes, a desymbolizer configured to desymbolize the sequence of symbols to obtain a sequence of syntax elements, a reconstructor configured to reconstruct the media data based on the sequence of syntax elements, selection depending on the activated low-complexity mode or the high-efficiency mode. In another aspect, a desymbolizer is configured to perform desymbolization such that the control parameter varies in accordance with the data stream at a first rate in case of the high-efficiency mode being activated and the control parameter is constant irrespective of the data stream or changes depending on the data stream, but at a second lower rate in case of the low-complexity mode being activated.

A decoder for decoding a data stream into which media data is coded has a mode switch configured to activate a low-complexity mode or a high-efficiency mode depending on the data stream, an entropy decoding engine configured to retrieve each symbol of a sequence of symbols by entropy decoding using a selected one of a plurality of entropy decoding schemes, a desymbolizer configured to desymbolize the sequence of symbols to obtain a sequence of syntax elements, a reconstructor configured to reconstruct the media data based on the sequence of syntax elements, selection depending on the activated low-complexity mode or the high-efficiency mode. In another aspect, a desymbolizer is configured to perform desymbolization such that the control parameter varies in accordance with the data stream at a first rate in case of the high-efficiency mode being activated and the control parameter is constant irrespective of the data stream or changes depending on the data stream, but at a second lower rate in case of the low-complexity mode being activated.

Racks and rack pods to support a plurality of sleds are disclosed herein. Switches for use in the rack pods are also disclosed herein. A rack comprises a plurality of sleds and a plurality of electromagnetic waveguides. The plurality of sleds are vertically spaced from one another. The plurality of electromagnetic waveguides communicate data signals between the plurality of sleds.

Racks and rack pods to support a plurality of sleds are disclosed herein. Switches for use in the rack pods are also disclosed herein. A rack comprises a plurality of sleds and a plurality of electromagnetic waveguides. The plurality of sleds are vertically spaced from one another. The plurality of electromagnetic waveguides communicate data signals between the plurality of sleds.

The present disclosure provides a decoding method and apparatus. The decoding method mainly includes: extracting pre-indexed information, storing the pre-indexed information in another file associated with a to-be-decoded file or a tail end of the to-be-decoded file, then reading the pre-indexed information before decoding is performed, and performing parallel decoding on multiple data segments in the to-be-decoded file according to the pre-indexed information. Using the foregoing storage method for the pre-indexed information may effectively reduce an I/O operation when the pre-indexed information is read, so as to avoid, to some extent, a system frame freezing phenomenon that may be caused when decoding is performed.

The present disclosure provides a decoding method and apparatus. The decoding method mainly includes: extracting pre-indexed information, storing the pre-indexed information in another file associated with a to-be-decoded file or a tail end of the to-be-decoded file, then reading the pre-indexed information before decoding is performed, and performing parallel decoding on multiple data segments in the to-be-decoded file according to the pre-indexed information. Using the foregoing storage method for the pre-indexed information may effectively reduce an I/O operation when the pre-indexed information is read, so as to avoid, to some extent, a system frame freezing phenomenon that may be caused when decoding is performed.

Technologies for generating manifest data for a sled include a sled to generate manifest data indicative of one or more characteristics of the sled (e.g., hardware resources, firmware resources, a configuration of the sled, or a health of sled components). The sled is also to associate an identifier with the manifest data. The identifier uniquely identifies the sled from other sleds. Additionally, the sled is to send the manifest data and the associated identifier to a server. The sled may also detect a change in the hardware resources, firmware resources, the configuration, or component health of the sled. The sled may also generate an update of the manifest data based on the detected change, where the update specifies the detected change in the hardware resources, firmware resources, the configuration, or component health of the sled. The sled may also send the update of the manifest data to the server.

Technologies for generating manifest data for a sled include a sled to generate manifest data indicative of one or more characteristics of the sled (e.g., hardware resources, firmware resources, a configuration of the sled, or a health of sled components). The sled is also to associate an identifier with the manifest data. The identifier uniquely identifies the sled from other sleds. Additionally, the sled is to send the manifest data and the associated identifier to a server. The sled may also detect a change in the hardware resources, firmware resources, the configuration, or component health of the sled. The sled may also generate an update of the manifest data based on the detected change, where the update specifies the detected change in the hardware resources, firmware resources, the configuration, or component health of the sled. The sled may also send the update of the manifest data to the server.

The present disclosure relates generally to techniques for improving the implementation of certain operations on an integrated circuit. In particular, deep learning techniques, which may use a deep neural network (DNN) topology, may be implemented more efficiently using low-precision weights and activation values by efficiently performing down conversion of data to a lower precision and by preventing data overflow during suitable computations. Further, by more efficiently mapping multipliers to programmable logic on the integrated circuit device, the resources used by the DNN topology to perform, for example, inference tasks may be reduced, resulting in improved integrated circuit operating speeds.

The present disclosure relates generally to techniques for improving the implementation of certain operations on an integrated circuit. In particular, deep learning techniques, which may use a deep neural network (DNN) topology, may be implemented more efficiently using low-precision weights and activation values by efficiently performing down conversion of data to a lower precision and by preventing data overflow during suitable computations. Further, by more efficiently mapping multipliers to programmable logic on the integrated circuit device, the resources used by the DNN topology to perform, for example, inference tasks may be reduced, resulting in improved integrated circuit operating speeds.