8 results about "Modular computations" patented technology

This disclosure provides a method, apparatus, and device medium for GPU redundancy deployment based on AI server supernodes. The method includes: acquiring and adjusting the internal and external interfaces of GPUs in a modular computing unit; configuring link paths for the nine GPUs in the modular computing unit to achieve pairwise interconnection of the nine GPUs; during the operation of the modular computing unit, setting eight of the nine GPUs as regular GPUs and setting the remaining one as a redundant GPU; when any one of the eight regular GPUs is detected to have failed, adjusting the remaining seven regular GPUs to achieve pairwise interconnection with the redundant GPU. In the embodiments of this disclosure, by adjusting GPU interface allocation, introducing a configurable optical switching module, and a redundancy management mechanism, the reliability and fault tolerance of the supernode system are improved without significantly increasing hardware costs.

Modular computing systems can include a plurality of computing blocks, a frame to physically support the computing blocks, and an interconnection infrastructure. The interconnection infrastructure can provide power, cooling, and communications for the computing blocks. Each computing block can include a housing, a computing portion within the housing, a cooling portion within the housing to extract heat from the computing portion, a power infrastructure to power the computing portion, a communications infrastructure to facilitate communications between the computing portion and an external network, a cooling infrastructure to transfer heat from within the housing to outside the housing, and a service corridor within the housing to provide physical access to at least the computing portion.

A computational architecture employing self-pruning fractal branch management for achieving supercomputer-class performance on standard hardware and noisy intermediate-scale quantum (NISQ) devices. Unlike conventional parallel computing systems requiring massive hardware resources or genetic algorithms requiring extensive population evolution, this invention utilizes hierarchical fractal doubles—modular computational units organized in self-similar tree structures—with real-time adaptive pruning eliminating non-promising solution branches based on geometric performance metrics computed via √2-scaled fractal analysis. Controlled perturbations (branch shaking) inject stochastic exploration preventing premature convergence while pruning maintains computational efficiency. The system achieves quantum-competitive performance on classical hardware through fractal interference patterns mimicking quantum superposition, and enables NISQ quantum computers to operate effectively despite hardware noise by pruning decoherence-corrupted branches before they contaminate computation. Core innovation: geometric pruning criterion comparing branch trajectory fractal dimension against optimal threshold, triggering instant elimination of branches exhibiting non-productive exploration patterns. Applications include neural architecture search, protein folding simulation, quantum system modeling, combinatorial optimization, and multi-agent coordination—all achieving 10-100× speedup versus conventional approaches while consuming 60-80% less energy through aggressive branch elimination. Technical advantages: (1) no training dataset required (deterministic pruning), (2) hardware-agnostic (runs on CPU / GPU / QPU), (3) noise-tolerant (quantum error mitigation via pruning), (4) energy-efficient (eliminates wasted computation), (5) scalable (fractal recursion to arbitrary depth).

A computing device includes internal components within a modular computing housing assembly that includes an enclosure, front lid, first and second side covers, and mounting arrangement. The enclosure includes connected panels defining a front opening and internal volume. The front lid is removably coupled to the enclosure, covers the front opening, and includes front, first side, and second side portions having side openings that allow computing ports to protrude therethrough. Each side cover is removably coupled to the enclosure and includes an outer face, an inner face, and computing port plugs that plug the computing ports. Removal of a side cover exposes the computing ports to be directly plugged into corresponding computing ports on a separate computing device. The mounting arrangement is located along an enclosure outer surface and includes mounting components that move between positions to lock onto and mount the computing device onto a separate external rail.

The present invention relates to a modular computing system (1) forming at least one computer, the system comprising: a motherboard comprising a plurality of PCIe slots and at least one PCIe switch, the PCIe switch being able to route the data communication between the various PCIe slots; a plurality of expansion cards, each expansion card being provided with a PCIe connector and one or more electronic components, the plurality of expansion cards comprising at least one expansion card of a first type, the one or more electronic components of each expansion card of the first type comprising at least one central processing unit and at least one memory, the expansion cards each being plugged into a slot of the respective motherboard, the PCIe switch being configured to define at least one expansion card of the first type as a root complex, RC, to form a computer.

A modular compute platform for artificial-intelligence inference is disclosed. Each compute module delivers exaFLOPS-class throughput within a six-hundred-watt envelope and can be installed on diverse carrier boards such as PCI Express, automotive, robotic, or rackmount systems. Modules include vertically integrated logic and memory for high-bandwidth communication and local storage of neural-network parameters. Uniform mechanical and thermal interfaces allow identical modules to serve multiple markets, enabling wide deployment of exa-scale inference capability.

Modular computing system comprising at least one computer The present invention relates to a modular computing system (1) comprising at least one computer, the system comprising: a motherboard comprising a plurality of PCIe slots and at least one PCIe switch, the PCIe switch being suitable for routing data communication between the different PCIe slots;a plurality of expansion cards, each expansion card having a PCIe connector and one or more electronic components, the plurality of expansion cards comprising at least one expansion card of a first type, the electronic component(s) of each expansion card of the first type comprising at least one central processing unit and at least one memory, the expansion cards each being plugged into a slot of the respective motherboard, the PCIe switch being configured to define at least one expansion card of the first type as the root complex, RC to form a computer. Figure for the abstract: 4;

A modular computing system can include a plurality of computing blocks, a framework for physically supporting the computing blocks, and interconnecting infrastructure. The interconnecting infrastructure can provide the computing blocks with power, cooling, and communications. Each computing block can include an enclosure, a computing section within the enclosure, a cooling section within the enclosure to extract heat from the computing section, power infrastructure to supply power to the computing section, communications infrastructure to facilitate communications between the computing section and an external network, cooling infrastructure to transfer heat from within the enclosure to outside of the enclosure, and a service corridor within the enclosure to provide physical access to at least the computing section.