Abstract: A system of high bandwidth memory (HBM) chiplets and compute chiplets includes embedded logic bridges that extend communication distances from the HBM chiplets to other chiplets beyond the ˜6 mm limit imposed by the JEDEC standard. The embedded logic bridges include high-speed (e.g., greater than 1 Gbps) communication circuits that drive communication signals longer distances without fading below the detection threshold of the receiver. The longer high-speed communication distances enable more HBM chiplets to connect to a compute chiplet (and other chiplets, such as I/O or other compute chiplets) to support computational workloads in high-performance computing and machine learning/artificial intelligence, which depend on access to large amounts of memory for efficient operations.

Abstract: Aspects of the disclosure are directed to methods, systems, and apparatuses using an instruction prefetch pipeline architecture that provides good performance without the complexity of a full cache coherent solution deployed in conventional CPUs. The architecture can include components which can be used to construct an instruction prefetch pipeline, including instruction memory (TiMem), instruction buffer (iBuf), a prefetch unit, and an instruction router.

Abstract: Aspects of the disclosure provide for an accelerator capable of accelerating data dependent, irregular, and/or memory-bound operations. An accelerator as described herein includes a programmable engine for efficiently executing computations on-chip that are dynamic, irregular, and/or memory-bound, in conjunction with a co-processor configured to accelerate operations that are predictable in computational load and behavior on the co-processor during design and fabrication.

Abstract: Aspects of the disclosure provide for an accelerator capable of accelerating data dependent, irregular, and/or memory-bound operations. An accelerator as described herein includes a programmable engine for efficiently executing computations on-chip that are dynamic, irregular, and/or memory-bound, in conjunction with a co-processor configured to accelerate operations that are predictable in computational load and behavior on the co-processor during design and fabrication.

Abstract: Aspects of the disclosure are directed to methods, systems, and apparatuses using an instruction prefetch pipeline architecture that provides good performance without the complexity of a full cache coherent solution deployed in conventional CPUs. The architecture can include components which can be used to construct an instruction prefetch pipeline, including instruction memory (TiMem), instruction buffer (iBuf), a prefetch unit, and an instruction router.

Abstract: Aspects of the disclosure are directed to methods, systems, and apparatuses using an instruction prefetch pipeline architecture that provides good performance without the complexity of a full cache coherent solution deployed in conventional CPUs. The architecture can include components which can be used to construct an instruction prefetch pipeline, including instruction memory (TiMem), instruction buffer (iBuf), a prefetch unit, and an instruction router.

Abstract: Aspects of the disclosure provide for an accelerator capable of accelerating data dependent, irregular, and/or memory-bound operations. An accelerator as described herein includes a programmable engine for efficiently executing computations on-chip that are dynamic, irregular, and/or memory-bound, in conjunction with a co-processor configured to accelerate operations that are predictable in computational load and behavior on the co-processor during design and fabrication.

Abstract: A network device, capable of understanding communications between an end user and the core network on a RAN network is disclosed. In some embodiments, the device is able to decode the control plane and the user plane. As such, it is able to determine when the end user has requested multimedia content. Once this is known, the device can optimize the delivery of that content in several ways. In one embodiment, the device requests the content from the content server (located in the core network) and transmits this content in a just-in-time manner to the end user. In another embodiment, the device automatically changes the encoding and resolution of the content, based on overall monitored network traffic. In another embodiment, the device automatically selects or modifies the format and resolution options based on overall bandwidth limitations, independent of the end user.

Abstract: A packet scheduling method and apparatus with the knowledge of application behavior, anticipated usage/behavior based on the type of content, and underlying transport conditions during the time of delivery, is disclosed. This type of scheduling is applicable to a content server or a transit network device in wireless (e.g., 3G, WIMAX, LTE, WIFI) or wire-line networks. Methods for identifying or estimating rendering times of multi-media objects, segmenting a large media content, and automatically pausing or delaying delivery are disclosed. The scheduling reduces transit network bandwidth wastage, and facilitates optimal sharing of network resources such as in a wireless network.

Abstract: An application mobility-management entity (“AME”) in a radio-access network detects movement of a mobile device and routes application data for already-active application streams associated with the mobile device to the device's new location by establishing a connection to a second AME in the radio-access network. The second AME merges forwarded application data from/to the first AME for previously active application streams with the new application streams (e.g., new TCP connections) locally and forwards to/from the mobile device through the radio access network elements.

Abstract: An application mobility-management entity (“AME”) in a radio-access network detects movement of a mobile device and routes application data for already-active application streams associated with the mobile device to the device's new location by establishing a connection to a second AME in the radio-access network. The second AME merges forwarded application data from/to the first AME for previously active application streams with the new application streams (e.g., new TCP connections) locally and forwards to/from the mobile device through the radio access network elements.

Abstract: The current invention defines a device and a method of logically inserting the device between two other network devices, for example, in a 3GPP Radio Access Network. The device transparently monitors and interacts with one or more control protocol layers in the two neighboring devices. The invention defines methods by which the intercepting node selectively passes through or proxies (selectively modifying portions of the protocols content) in such as way that the neighbor nodes are un-aware of the intercepting device. The proxy operation implies that the intercepting node is capable of terminating some protocol elements, injecting some protocol elements, or modifying protocol elements before forwarding them in such a way that the operation is transparent to neighboring nodes.

Abstract: A network device, capable of understanding communications between an end user and the core network on a RAN network is disclosed. In some embodiments, the device is able to decode the control plane and the user plane. As such, it is able to determine when the end user has requested multimedia content. Once this is known, the device can optimize the delivery of that content in several ways. In one embodiment, the device requests the content from the content server (located in the core network) and transmits this content in a just-in-time manner to the end user. In another embodiment, the device automatically changes the encoding and resolution of the content, based on overall monitored network traffic. In another embodiment, the device automatically selects or modifies the format and resolution options based on overall bandwidth limitations, independent of the end user.

Abstract: An application mobility-management entity (“AME”) in a radio-access network detects movement of a mobile device and routes application data for already-active application streams associated with the mobile device to the device's new location by establishing a connection to a second AME in the radio-access network. The second AME merges forwarded application data from/to the first AME for previously active application streams with the new application streams (e.g., new TCP connections) locally and forwards to/from the mobile device through the radio access network elements.

Abstract: A packet scheduling method and apparatus with the knowledge of application behavior, anticipated usage/behavior based on the type of content, and underlying transport conditions during the time of delivery, is disclosed. This type of scheduling is applicable to a content server or a transit network device in wireless (e.g., 3G, WIMAX, LTE, WIFI) or wire-line networks. Methods for identifying or estimating rendering times of multi-media objects, segmenting a large media content, and automatically pausing or delaying delivery are disclosed. The scheduling reduces transit network bandwidth wastage, and facilitates optimal sharing of network resources such as in a wireless network.

Abstract: The current invention defines a device and a method of logically inserting the device between two other network devices, for example, in a 3GPP Radio Access Network. The device transparently monitors and interacts with one or more control protocol layers in the two neighboring devices. The invention defines methods by which the intercepting node selectively passes through or proxies (selectively modifying portions of the protocols content) in such as way that the neighbor nodes are un-aware of the intercepting device. The proxy operation implies that the intercepting node is capable of terminating some protocol elements, injecting some protocol elements, or modifying protocol elements before forwarding them in such a way that the operation is transparent to neighboring nodes.

Abstract: The Bit Buddy is a handy wristband that holds your drill bits while you work. The Bit Buddy allows you to continue working with the ease of having accessibility to your drill bits on your wrist. This allows you the ability to continue to use your drill and quickly change out drill bits without needing to stop and search for the correct drill bits needed.