Abstract: Hardware accelerator designs for neural networks are improved with various approaches to reduce circuit area, improve power consumption, and reduce starvation. Convolutional layers of a neural network may multiply a set of weights with a set of inputs. One example defers two's complement arithmetic from the parallelized multiplication circuits and completes the two's complement arithmetic when the results are accumulated. In another example, a multiplication circuit initially multiplies an input by an initial value of the maximum (or minimum) multiplication range before applying the magnitude of a multiplication encoded relative to the multiplication range. In another example, after dimensional reduction earlier in the network hardware, circuitry for a convolutional layer uses a reduced number of convolutional block circuits that are reused across a plurality of clock cycles to apply different subsets of weight channels.

Abstract: The present invention relates generally to a population of stem cells (e.g., iPSCs or HSCs) that comprise nucleic acids encoding a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, for example two distinct tumour antigenic determinants. The present invention is also directed to a population of T cells that co-express a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, such as two distinct tumour antigenic determinants. The cells of the present invention can be derived from chosen donors whose HLA type is compatible with significant sectors of the populations, and are useful in a wide variety of applications, in particular in the context of the therapeutic treatment of neoplastic conditions.

Abstract: This disclosure relates to treatment of T cell diseases, particularly T cell lymphomas (TCLs) including, in particular, cutaneous T cell lymphomas (CTCLs) such as Sezary Syndrome (SS) and mycosis fungoides (MF), by targeting tumor-associated glycoprotein-72 (TAG-72).

Abstract: A cutting assembly for a rock excavation machine having a frame includes a boom and a cutting device supported on the boom. The boom includes a first portion and a second portion, the first portion supported for pivotable movement relative to the frame. In some embodiments, the first portion includes a first structure extending along a longitudinal base axis and a second structure moveable relative to the first portion in a direction parallel to the longitudinal base axis, and at least one bearing supports the second portion for movement relative to the first portion. In some embodiments, the second portion is pivotably coupled to the first portion by a universal joint, and a suspension system including a plurality of biasing members may be coupled between the first portion and the second portion.

Abstract: A cutting assembly for a rock excavation machine having a frame includes a boom and a cutting device supported on the boom. The boom includes a first portion and a second portion, the first portion supported for pivotable movement relative to the frame. In some embodiments, the first portion includes a first structure extending along a longitudinal base axis and a second structure moveable relative to the first portion in a direction parallel to the longitudinal base axis, and at least one bearing supports the second portion for movement relative to the first portion. In some embodiments, the second portion is pivotably coupled to the first portion by a universal joint, and a suspension system including a plurality of biasing members may be coupled between the first portion and the second portion.

Abstract: Methods, systems, apparatus, and articles of manufacture to reduce memory latency when fetching pixel kernels are disclosed. An example apparatus includes first interface circuitry to receive a first request from a hardware accelerator at a first time including first coordinates of a first pixel disposed in a first image block, second interface circuitry to receive a second request including second coordinates from the hardware accelerator at a second time after the first time, and kernel retriever circuitry to, in response to the second request, determine whether the first image block is in cache storage based on a mapping of the second coordinates to a block tag, and, in response to determining that the first image block is in the cache storage, access, in parallel, two or more memory devices associated with the cache storage to transfer a plurality of image blocks including the first image block to the hardware accelerator.

Abstract: A cutting assembly for a rock excavation machine having a frame includes a boom and a cutting device supported on the boom. The boom includes a first portion and a second portion, the first portion supported for pivotable movement relative to the frame. In some embodiments, the first portion includes a first structure extending along a longitudinal base axis and a second structure moveable relative to the first portion in a direction parallel to the longitudinal base axis, and at least one bearing supports the second portion for movement relative to the first portion. In some embodiments, the second portion is pivotably coupled to the first portion by a universal joint, and a suspension system including a plurality of biasing members may be coupled between the first portion and the second portion.

Abstract: This disclosure relates to methods for producing immune cells with enhanced function. More specifically, disclosed herein is a method for enhancing the function of an immune cell comprising modifying an immune cell to inhibit the function of at least one gene selected from the group consisting of RC3H1, RC3H2, A2AR, FAS, TGFBR1, and TGFBR2. Also disclosed herein is a method comprising modifying a stem or progenitor cell capable of differentiating into an immune cell to inhibit the function of at least one gene selected from the group consisting of RC3H1, RC3H2, A2AR, FAS, TGFBR1, and TGFBR2. Also disclosed herein are immune cells or stem cells made by the present methods, as well as the use of immune cells in therapeutic treatment.

Abstract: An DNN accelerator may perform 1×N kernel decomposition to decompose a convolutional kernel into kernel vectors, each of which includes multiple weights. Through the kernel decomposition, a weight operand may be generated from a filter. The DNN accelerator converts an input tensor into input operands. An input operand includes activations and has the same size as the weight operand. The DNN accelerator may read a first activation in the input operand from memory to an internal memory of a first PE and read a second activation in the input operand from the memory to an internal memory of a second PE. The first PE may receive the second activation from the second PE through activation broadcasting between the two PEs and perform MAC operations on the input operand and weight operand. The second PE may perform MAC operations on another input operand in the input tensor and the weight operand.

Abstract: The present invention relates generally to a population of stem cells (e.g., iPSCs or HSCs) that comprise nucleic acids encoding a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, for example two distinct tumour antigenic determinants. The present invention is also directed to a population of T cells that co-express a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, such as two distinct tumour antigenic determinants. The cells of the present invention can be derived from chosen donors whose HLA type is compatible with significant sectors of the populations, and are useful in a wide variety of applications, in particular in the context of the therapeutic treatment of neoplastic conditions.

Abstract: This disclosure relates to immune cells (such as T cells or NK cells) modified in their cell surface receptors to recognize one or more target antigens, in particular tumor-associated antigens. This disclosure also relates to a simple method for editing cell receptors, in particular cell surface receptors naturally expressed by immune cells such as T cells or NK cells, to create modified immune cells (e.g., cytotoxic cells) targeted against one or more target antigens, in particular tumor-associated antigens. Further, this disclosure relates to stem cells modified in one or more endogenous genes encoding one or more cell surface receptors and capable of differentiating into immune cells expressing modified cell surface receptors that recognize one or more target antigens. In addition, this disclosure relates to methods of making such modified stem cells.

Abstract: The present invention relates generally to a population of stem cells (e.g., iPSCs or HSCs) that comprise nucleic acids encoding a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, for example two distinct tumour antigenic determinants. The present invention is also directed to a population of T cells that co-express a T cell receptor and a chimeric antigen receptor directed to multiple distinct antigenic determinants, such as two distinct tumour antigenic determinants. The cells of the present invention can be derived from chosen donors whose HLA type is compatible with significant sectors of the populations, and are useful in a wide variety of applications, in particular in the context of the therapeutic treatment of neoplastic conditions.

Abstract: A cutting assembly for a rock excavation machine including a frame. The cutting assembly includes a boom supported on the frame and a cutting device. In some aspects, the boom includes a first portion and a second portion, and the first portion includes a first structure and a second structure slidable relative to the first structure. The second portion includes a first member pivotably coupled to the second structure, and a second member pivotably coupled to the first member. The cutting device is supported on the second member. In some aspects, a material handling device is supported independently of the boom and movable between a retracted position and an extended position independent of the boom.

Abstract: Methods, systems, apparatus, and articles of manufacture to reduce memory latency when fetching pixel kernels are disclosed. An example apparatus includes first interface circuitry to receive a first request from a hardware accelerator at a first time including first coordinates of a first pixel disposed in a first image block, second interface circuitry to receive a second request including second coordinates from the hardware accelerator at a second time after the first time, and kernel retriever circuitry to, in response to the second request, determine whether the first image block is in cache storage based on a mapping of the second coordinates to a block tag, and, in response to determining that the first image block is in the cache storage, access, in parallel, two or more memory devices associated with the cache storage to transfer a plurality of image blocks including the first image block to the hardware accelerator.

Abstract: A cutting assembly is provided for a rock excavation machine having a frame. The cutting assembly includes a boom, a cutting device, and a plurality of fluid actuators. The boom includes a base portion and a movable portion. The base portion is configured to be supported by the frame, and the movable portion is supported for sliding movement relative to the base portion in a direction parallel to a longitudinal axis of the base portion. The boom includes a wrist portion pivotably coupled to the movable portion at a pivot joint. The cutting device is supported on a distal end of the wrist portion. The fluid actuators are coupled between the base portion and the wrist portion. The fluid actuators are operable to move the movable portion and the wrist portion parallel to the longitudinal axis, and are also operable to bias the wrist portion against cutting loads exerted on the cutting device.

Abstract: Hardware accelerator designs for neural networks are improved with various approaches to reduce circuit area, improve power consumption, and reduce starvation. Convolutional layers of a neural network may multiply a set of weights with a set of inputs. One example defers two's complement arithmetic from the parallelized multiplication circuits and completes the two's complement arithmetic when the results are accumulated. In another example, a multiplication circuit initially multiplies an input by an initial value of the maximum (or minimum) multiplication range before applying the magnitude of a multiplication encoded relative to the multiplication range. In another example, after dimensional reduction earlier in the network hardware, circuitry for a convolutional layer uses a reduced number of convolutional block circuits that are reused across a plurality of clock cycles to apply different subsets of weight channels.

Abstract: A cutting assembly is provided for a rock excavation machine having a frame. The cutting assembly includes a boom, a cutting device, and a plurality of fluid actuators. The boom includes a base portion and a movable portion. The base portion is configured to be supported by the frame, and the movable portion is supported for sliding movement relative to the base portion in a direction parallel to a longitudinal axis of the base portion. The boom includes a wrist portion pivotably coupled to the movable portion at a pivot joint. The cutting device is supported on a distal end of the wrist portion. The fluid actuators are coupled between the base portion and the wrist portion. The fluid actuators are operable to move the movable portion and the wrist portion parallel to the longitudinal axis, and are also operable to bias the wrist portion against cutting loads exerted on the cutting device.

Abstract: Methods, systems, apparatus, and articles of manufacture to reduce memory latency when fetching pixel kernels are disclosed. An example apparatus includes a prefetch kernel retriever to generate a block tag based on a first request from a hardware accelerator, the first request including first coordinates of a first pixel disposed in a first image block, a memory interface engine to store the first image block including a plurality of pixels including the pixel in a cache storage based on the block tag, and a kernel retriever to access two or more memory devices included in the cache storage in parallel to transfer a plurality of image blocks including the first image block when a second request is received including second coordinates of a second pixel disposed in the first image block.

Abstract: A cutting assembly for a rock excavation machine including a frame. The cutting assembly includes a boom supported on the frame and a cutting device. In some aspects, the boom includes a first portion and a second portion, and the first portion includes a first structure and a second structure slidable relative to the first structure. The second portion includes a first member pivotably coupled to the second structure, and a second member pivotably coupled to the first member. The cutting device is supported on the second member. In some aspects, a material handling device is supported independently of the boom and movable between a retracted position and an extended position independent of the boom.

Abstract: This disclosure relates to treatment of T cell diseases, particularly T cell lymphomas (TCLs) including, in particular, cutaneous T cell lymphomas (CTCLs) such as Sezary Syndrome (SS) and mycosis fungoides (MF), by targeting tumor-associated glycoprotein-72 (TAG-72).