Become a SponsormatchmAIQer category
Compute is the constraint behind many AI roadmaps. This category covers accelerated compute, edge devices and the systems that manage them, whether in the cloud, in a colocation facility or on premises.
Industrial computers, edge AI systems and IoT platforms
Advantech makes industrial PCs, embedded boards and edge AI systems, many built on NVIDIA Jetson and Intel platforms, plus WISE-IoT software for device management.
Simulation, HPC workload management and data analytics software
Altair provides simulation, high-performance computing and data analytics software used by engineering and IT teams. Its HPC portfolio centers on workload schedulers and cloud-bursting tools that manage CPU and GPU clusters, and it also offers AI and data science tooling. Siemens completed its acquisition of Altair in 2025.
AI vision SoCs for cameras, vehicles and edge infrastructure
Ambarella designs CVflow-based system-on-chips for video and AI, serving security cameras, automotive and robotics, including newer parts for generative AI at the edge.
CPUs, Instinct GPUs and adaptive chips for AI from data center to edge
AMD designs processors and accelerators for servers, PCs and embedded systems. Its Instinct GPUs target AI training and inference, supported by the open ROCm software stack, and it also sells EPYC server CPUs, Ryzen AI client chips and FPGA-based adaptive SoCs from the Xilinx acquisition.
Data center networking for AI clusters with EOS software
Arista supplies high-speed Ethernet switches and EOS software for cloud and AI data centers, along with CloudVision for network management.
Processor architecture and IP licensed for data center to edge AI
Arm licenses CPU and NPU designs, including Neoverse server cores, Cortex cores and Ethos NPUs, used by chipmakers and cloud providers. It does not historically manufacture chips, and is majority-owned by SoftBank.
Metis AI processing units for edge inference
Axelera AI designs the Metis AIPU, which accelerates vision and generative inference in edge devices, offered as cards, modules and systems with the Voyager SDK.
Graph streaming processors and software for edge and hybrid AI
Blaize builds Graph Streaming Processor hardware and AI Studio software for edge and hybrid AI deployments in vision, defense and smart infrastructure. It became publicly listed in 2025.
Smart sensors, IoT software and automotive AI from a global industrial group
Bosch is a German technology group whose businesses include mobility, industrial technology and consumer goods. For AI hardware it supplies MEMS sensors with on-device intelligence through Bosch Sensortec, IoT platforms and automotive driver-assistance systems.
Custom AI accelerators and Ethernet networking silicon
Broadcom designs custom AI accelerators, known as XPUs, for hyperscale customers, along with Tomahawk and Jericho Ethernet switch chips and optical components used in AI clusters. It also owns enterprise software including VMware.
IT solutions provider supplying AI-ready hardware, software and services
CDW resells and integrates technology from vendors including NVIDIA, Dell, HPE and Cisco, serving business, government, education and healthcare customers. It offers AI readiness assessments, infrastructure design, deployment and lifecycle support for AI workloads.
Wafer-scale AI processors, systems and fast inference cloud
Cerebras builds the Wafer-Scale Engine, a very large AI processor, and CS-series systems for training and inference. It also operates a cloud inference service and sells supercomputer-scale clusters. The company went public on Nasdaq in 2026.
Direct liquid cooling for servers and data centers
CoolIT makes cold plates and coolant distribution units for liquid-cooled servers. Ecolab announced an agreement in 2026 to acquire it.
Specialized GPU cloud for AI training and inference
CoreWeave operates GPU-focused data centers and a cloud platform with Kubernetes-based services for AI workloads. It offers NVIDIA GPU capacity and managed tooling to AI labs and enterprises.
AI cloud and data center infrastructure built around energy efficiency
Crusoe builds AI data centers and operates Crusoe Cloud, offering GPU compute with managed inference and orchestration.
Power management and cooling for data centers
Eaton supplies power distribution, UPS and, following an acquisition, liquid cooling for data centers, along with Brightlayer software.
IT services group developing AI platforms, processors and computing hardware
Fujitsu is a Japanese IT company that provides services, servers, and AI technology. It builds Arm-based processors such as FUJITSU-MONAKA, supercomputing systems derived from the Fugaku program, and the Kozuchi AI platform, and offers digital annealing for optimization problems.
Servers and GPU systems with management software
GIGABYTE makes servers, motherboards and workstations; its enterprise arm Giga Computing sells GPU servers and management software.
Satellite operator offering IoT trackers and connectivity
Globalstar operates a low-earth-orbit satellite network and sells satellite voice, data and IoT products, including asset trackers and messengers. It supplies connectivity for remote monitoring devices; it does not make AI accelerators.
Intelligence Processing Units for AI, owned by SoftBank
Graphcore designs Intelligence Processing Units and software for AI workloads. SoftBank Group acquired it in 2024.
LPU inference chips and the GroqCloud inference service
Groq designs the Language Processing Unit, an inference-focused chip, and runs GroqCloud for hosted model inference and GroqRack for on-premises deployment. In late 2025 Nvidia struck a licensing deal for its technology and hired its leadership, while the company continued operating GroqCloud.
Edge AI processors and accelerator modules for vision workloads
Hailo designs AI accelerators for edge devices, offered as M.2 and PCIe modules and as vision processors for cameras. Its Hailo-8 and newer parts target low-power inference in cameras, vehicles and industrial gear.
Servers, supercomputing, networking and hybrid cloud for AI workloads
HPE sells servers, storage, networking and the GreenLake hybrid cloud platform. For AI it offers Cray supercomputing systems, ProLiant GPU servers and a private cloud AI stack developed with NVIDIA, plus fleet management tools.
PCs, workstations and printers with on-device AI and fleet management
HP Inc. sells personal computers, workstations and printing products. Its AI-focused lines include PCs with neural processing units and Z workstations for data science, along with software for managing device fleets.
CPUs, Gaudi AI accelerators and AI PC processors
Intel makes processors and accelerators for data centers, PCs and edge systems. Its AI products include Xeon CPUs with built-in AI instructions, Gaudi accelerators, Core Ultra chips with NPUs, and the OpenVINO toolkit for deploying models.
Server appliance, OEM integration and hardware support services
Iron Systems, part of IRON Global, builds custom server appliances and embedded systems and provides OEM integration, supply chain and global repair services. It lists NVIDIA GPU, AMD and Intel-based platforms and edge infrastructure services.
GPU cloud, clusters and servers for AI training and inference
Lambda offers on-demand and reserved NVIDIA GPU cloud capacity, private clusters, and GPU workstations and servers. It began as a deep learning workstation maker and now focuses on AI cloud infrastructure for training and inference.
Low-power FPGAs with sensor and edge AI solution stacks
Lattice makes small, low-power FPGAs used in sensor fusion, vision and security, with software stacks for edge inference.
PCs, servers and liquid-cooled infrastructure with hybrid AI solutions
Lenovo sells PCs, workstations, data center servers and services. Its infrastructure group offers ThinkSystem GPU servers, Neptune liquid cooling and validated hybrid AI stacks, alongside AI PCs and edge systems.
Composable infrastructure software for pooling GPUs and memory
Liqid provides software and fabric hardware that lets operators compose GPU, memory and storage resources across servers.
Custom AI silicon, optics and data center connectivity chips
Marvell supplies custom compute silicon, electro-optics, Ethernet and storage controllers for cloud and AI infrastructure.
High-bandwidth memory and storage for AI systems
Micron makes DRAM, high-bandwidth memory and SSDs used with AI accelerators and servers.
AI cloud with GPU infrastructure and inference services
Nebius is an AI infrastructure company, operating GPU clusters and a managed cloud platform with training and inference services. It emerged from the restructuring of Yandex's international business.
GPUs, systems and software platform for AI training and inference
NVIDIA designs GPUs, networking and full-stack systems for AI and accelerated computing. Its products span Blackwell data center GPUs, DGX systems, NVLink and Spectrum networking, Jetson edge modules, and the CUDA and NVIDIA AI Enterprise software stack.
Processors and microcontrollers with embedded neural acceleration
NXP Semiconductors makes processors, microcontrollers and connectivity chips for automotive, industrial and IoT markets. Its i.MX and MCX families include neural processing units, supported by the eIQ machine learning software toolkit, and it announced an acquisition of Kinara, an edge AI accelerator company.
AI cluster infrastructure, cluster management software and memory
Penguin Solutions designs and operates AI and HPC clusters, with OriginAI infrastructure designs, ClusterWare management software, and memory products.
Mobile, edge and data center AI processors and developer tools
Qualcomm makes Snapdragon processors with on-device NPUs, Dragonwing industrial and IoT chips, and Cloud AI accelerators for data center inference. It also offers Qualcomm AI Hub and acquired Edge Impulse for edge ML development.
Data center servers and racks for cloud and AI
QCT, part of Quanta, designs servers, storage and rack systems for cloud and enterprise customers, including GPU platforms.
Microcontrollers and processors with embedded AI acceleration
Renesas supplies microcontrollers and microprocessors for automotive and industrial markets, including RZ/V parts with DRP-AI acceleration and Reality AI software.
Reconfigurable dataflow chips and full-stack AI inference platform
SambaNova builds the SN-series Reconfigurable Dataflow Unit and offers SambaNova Cloud and on-premises systems for running large models. Talks with Intel on an acquisition ended and the companies signed an inference partnership.
Power, cooling and infrastructure management software for data centers
Schneider Electric supplies power, cooling and the EcoStruxure IT software used to monitor data center equipment, and publishes reference designs for AI racks.
Industrial automation, edge computing and AI for manufacturing
Siemens is a technology group serving industry, infrastructure and healthcare. In AI hardware it supplies Industrial Edge devices, SIMATIC industrial PCs and software for deploying AI models on factory floors, through the Xcelerator portfolio.
Machine learning system-on-chip platforms for physical AI at the edge
SiMa.ai builds MLSoC chips and software for running vision and multimodal models at the edge in embedded systems.
Microcontrollers and sensors with on-device neural acceleration
ST makes STM32 microcontrollers, some with a Neural-ART accelerator, and MEMS sensors with built-in machine learning, supported by the ST Edge AI Suite.
Immersion cooling and data center design for AI
Submer provides immersion and liquid cooling and data center design services for high-density compute.
GPU servers, rack-scale systems and liquid cooling for AI
Supermicro designs and builds servers, storage and rack-scale systems using building-block architecture, with strong focus on NVIDIA and AMD GPU platforms and direct liquid cooling. It also offers management software and integration services.
Ultra-low-power neural decision processors for always-on devices
Syntiant makes NDP chips for always-on speech, sensor and vision AI in earbuds, wearables and IoT devices, along with models and development tools.
RISC-V based AI processors, accelerator cards and IP licensing
Tenstorrent designs AI accelerators and RISC-V CPU IP, sells Wormhole and Blackhole cards and servers, and licenses its designs. It publishes open-source software stacks and is led by chip architect Jim Keller.
Microcontrollers and vision processors with integrated edge AI
Texas Instruments makes analog chips, microcontrollers and processors, including MCUs with an on-chip NPU and AM6xA vision processors, with Edge AI Studio tools.
Automotive supplier with ADAS sensors and AI research for driving assistance
Valeo is a French automotive supplier that makes cameras, radar, ultrasonic sensors and lidar for driver-assistance systems. It runs valeo.ai, a research lab working on perception and AI for automated driving.
Power, cooling and monitoring infrastructure for AI data centers
Vertiv provides power systems, liquid and air cooling, and monitoring software for data centers, including prefabricated modules for high-density AI deployments.
Technology solutions provider with an AI proving ground lab
World Wide Technology integrates and supplies IT infrastructure from vendors such as NVIDIA, Cisco, Dell and HPE. It operates the Advanced Technology Center lab, where customers test AI infrastructure before purchase, and offers integration and supply chain services.
No providers match those filters.
Offer ai hardware solutions or services and not listed? Apply to join matchmAIQer
Compute is the constraint behind many AI roadmaps. This category covers accelerated compute, edge devices and the systems that manage them, whether in the cloud, in a colocation facility or on premises.
Accelerated Compute, Edge and IoT Devices and Hardware Management.
52 providers, each with an overview, products, key features, deployment options, pricing model and FAQs.
Take Find your match. It scores how well ai hardware fits the priorities you name and the readiness gaps in your answers, and shows the specialties and providers to look at first.
