-
公开(公告)号:US20230418579A1
公开(公告)日:2023-12-28
申请号:US18464044
申请日:2023-09-08
发明人: Hanwoong JUNG
摘要: A compile method for a neural network, the compile method includes receiving data related to the neural network, generating a grouped layer by grouping layers comprised in the neural network based on the data, generating a set of passes executable in parallel based on a dependency between a plurality of passes to process the neural network, generating a set of threads performing a plurality of optimization functions based on whether optimization operations performed by the optimization functions is performed independently for the layers, respectively, or sequentially based on a dependency between the layers, and performing compilation in parallel based on the grouped layer, the set of passes, and the set of threads.
-
公开(公告)号:US11797291B2
公开(公告)日:2023-10-24
申请号:US17429784
申请日:2020-02-04
CPC分类号: G06F8/65 , G06F8/451 , G06F9/505 , G06F3/123 , H04W52/0219
摘要: A software update management apparatus includes a storage unit adapted to divide a network into one or more blocks and store block management information indicating whether each of network devices belonging to each of the resulting blocks is an active device or a standby device; an update instruction receiving unit adapted to receive software update instructions; a software update information generating unit adapted to generate software update information; a software updating unit adapted to perform software update processes after transferring traffic to standby devices in same blocks as respective active devices when it is determined that the network devices are active devices according to the software update information and thereby perform the software update processes for active devices or standby devices in different blocks in parallel.
-
3.
公开(公告)号:US20230336811A1
公开(公告)日:2023-10-19
申请号:US18335276
申请日:2023-06-15
申请人: DISH Network L.L.C.
发明人: Kan Man Wong
IPC分类号: H04N21/433 , H04N21/238 , G10L15/22 , H04N21/45 , H04N21/262 , H04N21/61 , H04N21/81 , H04N21/4363 , H04N21/462 , H04B7/185 , H04N21/2665 , G06F9/451 , G06F9/44 , H04N21/475 , H04N21/478 , H04L61/5014 , H04L41/0893 , H04L65/611 , G06F8/41 , G06F8/70
CPC分类号: H04N21/433 , H04N21/238 , G10L15/22 , H04N21/4518 , H04N21/4516 , H04N21/26291 , H04N21/6118 , H04N21/6125 , H04N21/6143 , H04N21/818 , H04N21/43637 , H04N21/4622 , H04B7/18523 , H04N21/2665 , G06F9/451 , G06F9/44 , H04N21/475 , H04N21/478 , H04L61/5014 , H04N21/6106 , H04L41/0893 , H04L65/611 , G06F8/45 , G06F8/451 , G06F8/453 , G06F8/70 , G10L2015/223 , H04L2101/668
摘要: Various arrangements for facilitating smart television content receivers are provided. A primary television receiver (PTR) having a first operating system may be configured to receive digital content from a remote content provider and distribute the content to one or more devices in response to a request for the content. A secondary television receiver (STR) configured to be in communication with a PTR and having a second operating system may be configured to receive the digital content from the PTR and provide the content to a display for presentation. The STR may include a first software stack including a first inter-process communication (IPC) mechanism, and a second software stack including the first IPC mechanism and a second IPC mechanism. The first IPC may support communication within the first stack as well as between the first stack and the second stack. The second IPC may support communication within the second stack.
-
4.
公开(公告)号:US20230315415A1
公开(公告)日:2023-10-05
申请号:US17705099
申请日:2022-03-25
发明人: Skyler Arron Windh , Allan Kennedy Porterfield , Douglas John Vanesko , Randall Paul Meyer , Patrick Alan Estep , Bashar Romanous
IPC分类号: G06F8/41
摘要: An assembly language program for a coarse grained reconfiguration array (CGRA), having dispatch interface information indicating operations to be performed via a dispatch interface of the CGRA to receive an input, memory interface information indicating operations to be performed via one or more memory interfaces of the CGRA, tile memory information indicating memory variables referring to memory locations to be implemented in tile memories of the CGRA, a flow description specifying one or more synchronous data flows, through the memory locations referenced via the memory variables in the tile memory information, to produce a result from the input using the CGRA.
-
公开(公告)号:US20230267024A1
公开(公告)日:2023-08-24
申请号:US18012924
申请日:2020-10-27
发明人: Fang CAO , Zhenhua GUO , Li WANG , Kai GAO
摘要: Disclosed are a method for realizing an nGraph framework supporting an FPGA backend device, and a related apparatus. The method includes: integrating an OpenCL standard API library into an nGraph framework; creating, in the nGraph framework, an FPGA backend device creation module for registering an FPGA rear-end device, initializing an OpenCL environment and acquiring the FPGA backend device; creating, in the nGraph framework, an FPGA buffer space processing module for opening up an FPGA buffer space and for reading and writing an FPGA cache; creating, in the nGraph framework, an OP kernel implementation module for creating an OP kernel and compiling the OP kernel; and creating, in the nGraph framework, an FPGA compiling execution module for registering, scheduling and executing the OP kernel.
-
公开(公告)号:US09934010B1
公开(公告)日:2018-04-03
申请号:US14675935
申请日:2015-04-01
CPC分类号: G06F8/445 , G06F8/30 , G06F8/314 , G06F8/41 , G06F8/427 , G06F8/437 , G06F8/447 , G06F8/451 , G06F8/453 , G06F9/48 , G06F9/546 , G06F2209/548
摘要: Programming in a multiprocessor environment includes accepting a program specification that defines a plurality of processing modules and one or more channels for sending data between ports of the modules, mapping each of the processing modules to run on a set of one or more processing engines of a network of interconnected processing engines, and for at least some of the channels, assigning one or more elements of one or more processing engines in the network to the channel for sending data between respective processing modules.
-
公开(公告)号:US09864582B2
公开(公告)日:2018-01-09
申请号:US14675674
申请日:2015-03-31
申请人: CAVIUM, INC.
CPC分类号: G06F8/33 , G06F8/314 , G06F8/41 , G06F8/427 , G06F8/443 , G06F8/4434 , G06F8/445 , G06F8/447 , G06F8/451 , G06F8/70 , G06F9/30145 , G06F15/76 , G06F15/7825
摘要: A processing network including a plurality of lookup and decision engines (LDEs) each having one or more configuration registers and a plurality of on-chip routers forming a matrix for routing the data between the LDEs, wherein each of the on-chip routers is communicatively coupled with one or more of the LDEs. The processing network further including an LDE compiler stored on a memory and communicatively coupled with each of the LDEs, wherein the LDE compiler is configured to generate values based on input source code that when programmed into the configuration registers of the LDEs cause the LDEs to implement the functionality defined by the input source code.
-
8.
公开(公告)号:US09836283B2
公开(公告)日:2017-12-05
申请号:US14675710
申请日:2015-03-31
申请人: CAVIUM, INC.
CPC分类号: G06F8/33 , G06F8/314 , G06F8/41 , G06F8/427 , G06F8/443 , G06F8/4434 , G06F8/445 , G06F8/447 , G06F8/451 , G06F8/70 , G06F9/30145 , G06F15/76 , G06F15/7825
摘要: A processing network including a plurality of lookup and decision engines (LDEs) each having one or more configuration registers and a plurality of on-chip routers forming a matrix for routing the data between the LDEs, wherein each of the on-chip routers is communicatively coupled with one or more of the LDEs. The processing network further including an LDE compiler stored on a memory and communicatively coupled with each of the LDEs, wherein the LDE compiler is configured to generate values based on input source code that when programmed into the configuration registers of the LDEs cause the LDEs to implement the functionality defined by the input source code.
-
公开(公告)号:US09785423B2
公开(公告)日:2017-10-10
申请号:US14694856
申请日:2015-04-23
申请人: Google Inc.
发明人: Albert Meixner
CPC分类号: G06F8/451 , G06F8/441 , G06F8/4441 , G06F8/445 , G06F9/30003 , G06F9/3001 , G06F9/30032 , G06F9/3004 , G06F9/30043 , G06F9/30134 , G06F9/30145 , G06F9/345 , G06F9/3887
摘要: A method is described that includes translating higher level program code including higher level instructions having an instruction format that identifies pixels to be accessed from a memory with first and second coordinates from an orthogonal coordinate system into lower level instructions that target a hardware architecture having an array of execution lanes and a shift register array structure that is able to shift data along two different axis. The translating includes replacing the higher level instructions having the instruction format with lower level shift instructions that shift data within the shift register array structure.
-
公开(公告)号:US20170102967A1
公开(公告)日:2017-04-13
申请号:US15387312
申请日:2016-12-21
发明人: Chen Tian , Handong Ye , Ziang Hu
CPC分类号: G06F9/4881 , G06F8/41 , G06F8/423 , G06F8/451 , G06F8/48 , G06F8/51 , G06F9/4812 , G06F13/24
摘要: An embodiment includes a method includes designating a portion of a plurality of processing cores as an input/output (I/O) core and compiling a program source code to produce compiled program source code, including identifying an I/O operation region of the program source code, determining a number of I/O operations for the I/O operation region, and determining a number of system resources and system resource types for the I/O operation region. The method also includes executing the program source code using the plurality of processing cores, including scheduling the I/O operation region of the program source code on the I/O core of the plurality of processing cores.
-
-
-
-
-
-
-
-
-