AI Inference Processor Vendor/Product Directory … Over 80 design choices from over 80 vendors for your next AI cloud or edge-based inference system
Mark Stansberry, The StatisticsMatrix, March 5, 2026
Connect with Mark Stansberry On LinkedIn ..
There are numerous types of AI chips, chiplets, IP cores and modules on the market. Many, but by far not all, are classified as AI inference processors (Edge and Cloud based). Inference processors are sometimes classified as AI accelerators. Some AI processors can perform inference and model training (learning) operations.
The vendor product directory contains companies that have indicated either directly or indirectly that they offer AI inference processors for either edge devices or the cloud (data centers).
These companies, which presently number about 80, offer numerous AI processors that can be used to design edge-based or cloud-based AI inference systems (AI servers).
The directory includes the company and the name of the AI inference processor offered and the corresponding link that elaborates on the specific product or product line. The location and the stock exchange that the company is traded on is also listed.
Abacus Semiconductor>USA>CA>San Jose
AI Inference Processor
https://abacus-semi.com/products.html
Advanced Micro Devices>USA>CA>Sunnyvale
Public>NYSE>AMD
AI Cloud and PC Inference Processor<EPVC 9005
https://www.amd.com/en/products/processors/server/epyc/ai/9005-inference.html
Advanced Micro Devices>USA>CA>Sunnyvale
Public>NYSE>AMD
AI Edge Inference Processors>FPGA
https://www.amd.com/en/products/adaptive-socs-and-fpgas/versal/ai-edge-series.html
Advanced Micro Devices>Untether AI>Canada>Ontario>Toronto
AI Neural Net Inference Processor
www.untether.ai
Untether AI Shuts Down, Engineering Team Joins AMD – EE Times
Advanced Micro Devices (Xilinx)
AI FPGA Inference Processors
https://www.xilinx.com/content/dam/xilinx/publications/solution-briefs/xilinx-versal-ai-compute-solution-brief.pdf
Alibaba Group Holdings>China>Hangzhou
Public>NYSE:BABA
T-Head Unit>China
AI Inference Processor>Hanguang 800
https://www.datacenterdynamics.com/en/news/alibaba-develops-ai-inferencing-chip-as-exports-of-nvidia-h20-continue-to-stall-report/
Amazon>Perceive>USA>CA>San Jose
AI Inference Processor>Erge Edge Inference Processor
https://www.geekwire.com/2024/amazon-to-acquire-perceive-for-80m-from-xperi-expanding-its-ai-technology-for-edge-devices/
Anaflash>USA>CA>Sunnyvale
AI Edge Inference Processor
https://www.anaflash.com/product
https://www.sagence-ai.com/
Andes Technology>Taiwan>Hsinchu
Public>TWSE>6533.TW
AI Edge Inference Processor>AndesAIR AnDLA™ I350>Deep Learning Accelerator (DLA)>Convolution>Pooling >Activation>Concatenation>Channel Padding
https://www.andestech.com/en/products-solutions/andesaire-ai/andla-i350/
Applied Materials Inc>USA>CA>Santa Clara
Public>NASDAQ>AMAT
Acquisitions: Think Silicon
Podcast: Think Silicon: From Patras, Greece to its acquisition by Applied Materials | Endeavor Greece
ARM Holdings>UK>Cambridge
Public>NASDAQ>ARM
AI Edge Inference Processor>Ethos U55>CNN>RNN>NPU
https://www.arm.com/products/silicon-ip-cpu/ethos/ethos-u55
Aspinity>USA>PA>Pittsburgh
AI Inference Processor>Analog Machine Learning Chip>AML100>>Applications>Sensors>Glass Breaking>Smoke>DashCAM
https://www.aspinity.com/AML100/
Axelera AI>The Netherlands>Eindhoven
AI Edge Inference Processor>Metis
Applications>Industrial Manufacturing>Retail>Security>Healthcare>Smart Cities>Robotics>Agriculture>Computer Vision>High Performance Compute>Space
https://axelera.ai/ai-accelerators
Biren Technology>China>Shanghai
Public>HKEX>6082.HK
AI Edge Inference Processor>GPU
https://www.birentech.com/about/
Bitmain>China>Beijing
AI Training Processor>AI Edge Inference Processor >> Applications>Vision Processing>Bitcoin Mining
https://www.bitmain.com/news-detail/ubitus-contracted-with-bitmain-teaming-up-to-explore-cloud-ai-computing-market-44
Bittware(Molex)>USA>NH>Concord
AI Inference Processor>AI Groq Processor>AI Data Center Processor>AI ML AI Edge Processor
https://www.bittware.com/solutions/ai-ml/
Molex>Public>NASDAQ>MOLX>IL> Lisle
Acquired: Molex Announces Acquisition of BittWare | Molex
Blaize Holdings Inc>USA>CA>El Dorado Hills
Public>NASDAQ>BZAIW
Graphic Streaming Processor Unit: AI Inference Edge Processor
https://www.blaize.com/products/
Cadence Design Systems>USA>CA>San Jose
Public>NASDAQ>CDNS
AI Inference Processor IP Cores>NeoNPU>NeureEdge AI Coprocessor
https://www.cadence.com/en_US/home/tools/silicon-solutions/ai-ip-platform.html
CEA-LETI>France>Grenoble
AI Inference Processor Memory>Nonvolatile>Ferroelectric>Memristor
https://www.leti-cea.com/cea-tech/leti/english/Pages/What’s-On/Press%20release/French-Team-led-by-CEA-Leti-Develops-First-Hybrid-Memory-Technology-Enabling-On-Chip-AI-Learning-and-Inference.aspx
Cerebras Systems>USA>CA>Sunnyvale
AI Inference Processor>Wafer-Scale Engine (WSE-3)>LLM Inference Processor>125 Petaflops>Innovation
https://www.cerebras.ai/inference
Chinese Academy of Sciences>China>Beijing
AI Inference Edge Processor> 3D Vector-Matrix Multiplication> Energy Efficiency>8.32 TOPs per watt> Bit Density is 58.2 bit µm–2>MRI Edge Detection>CNN>CIM>Memristor
https://www.nature.com/articles/s41928-022-00795-x
d-Matrix>USA>CA>Santa Clara
AI Inference Data Center Processor>Digital CIM>Corsair>Chiplet
https://www.d-matrix.ai/
DeePhi Tech (AMD>Xilinx)>China>Beijing
AI Inference Processor>Accelerated Speech Recognition Processor>Xilinx FPGA Based
Xilinx Invests in Machine Learning Pioneer DeePhi Tech
China AI start-up DeePhi Tech sold to Xilinx | AVCJ
AWS Marketplace: DeePhi Tech
DEEPX>Korea>Gyeonggi-do
AI Edge Inference Processor>Inferences Per Second per Watt>DX-M1
https://deepx.ai/products/dx-m1/
Digital Media Professionals Inc>Japan>Tokyo
Public>TSE>3652 Configurable AI Edge Inference Processor IP Core>Image>Camera https://www.dmprof.com/en/products-and-services/ai-products/hardware/edge-ai-soc/di1-2/
eASICS (Keysight)>Belgium>Leuven
AI Inference Processor>Neural Network Inference Engine>Configurable> nearbAI IP Cores
https://careers.keysight.com/easics
EdgeCortix>Japan>Kawasaki-shi
AI Edge Inference Processor>SAKURA-II Accelerator>Dynamic Neural Accelerator Processor Architecture (DNA)
Applications>Defense>Aerospace>Video Streaming
https://www.edgecortix.com/en/products
Model Cat (eta compute)>USA>CA>Sunnyvale
AI Edge Inference Processor>TENSAI
AI Edge Inference Chip Based Model Development>Aptos> Neural Spiking Network
https://www.silabs.com/blog/silicon-labs-and-eta-compute-simplify-edge-ml-development
https://modelcat.ai/blog Inference at 100-microwatt
Etched>USA>CA>San Jose
AI Inference Processor>Etched Sohu Chip
https://www.etched.com/careers
Expedera Inc>USA>CA>Santa Clara
AI Edge Inference Processor IP Cores>NPU>OriginTM E1 LittleNPU
https://www.expedera.com/
Flex Logix Technologies>USA>CA>Mountain View
AI Inference IP Cores>InferX AI Technology>eFPGAs
https://assets.flex-logix.com/resources/FL_X1M%20Product%20Brief_226.pdf
Fractile AI>UK>London
AI Inference Processor>AI Large Language Model (LLM)>Compute in Memory (CIM) Inference Processor
https://www.fractile.ai/
FuriosaAI>Korea>Seoul
AI Inference Processor>RNGD Cloud and Data Center Accelerator>LLM>Tensor Processor Unit https://furiosa.ai/rngd
Google>USA>CA>Mountain View
Public>NASDAQ>GOOG
AI Inference Processor>Tensor Processing Unit>Ironwood>Trillium>Cloud TPU
https://cloud.google.com/tpuGoogle
Graphcore Limited>UK>Bristol
AI Cloud Processor>AI Inference Processors>Colossus MK2 IPU Processor>Training and Inference Processor
https://www.graphcore.ai/products/ipu
https://www.graphcore.ai/posts/graphcore-joins-softbank-group-to-build-next-generation-of-ai-compute
https://docs.graphcore.ai/projects/ai-float-white-paper/en/latest/index.html
GreenWaves Technologies>France>Grenoble
Technology assets acquired
AI Edge Inference Processor
https://www.linkedin.com/posts/greenwaves-technologies_weunfortunately-got-caught-in-a-perfect-activity-7313159925101166594-XnkV?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAzvvJ4BXBHbVU4e2rU_m43kQHxClvPgbfE
Groq>USA>CA>Mountain View
AI Inference Processor>AI Cloud Inference Processor>LLM>Language Processing Unit
https://groq.com/lpu-architecture
GSI Technology>USA>CA>Sunnyvale
Public>NASDAQ>GSIT
AI Inference Processor>Associative Processor>Innovation
https://gsitechnology.com/revolutionizing-ai-inference-with-gsi-apu/
Habana Labs (Intel)>Israel>Tel Aviv
AI Inference Processor>Gaudi AI Acceleration Platform
https://habana.ai/
Hailo>Israel>Tel Aviv
AI Edge Inference Processor> Haillo-15, Hailo-8 AI Accelerators>Vision Processors>VPU
https://hailo.ai/
IBM Corp>USA>New York
AI Inference Processor>IBM AI Mainframe>Spyre Accelerator>Telum II>North Pole>
https://hc2024.hotchips.org/assets/program/conference/day1/04_HC2024.IBM.CBerry.final.pdf
Intel Corp>USA>CA>Santa Clara
Public>NASDAQ>INTC
AI Inference Processor
https://www.intel.com/content/www/us/en/developer/topic-technology/artificial-intelligence/enterprise-inference.html
Kalray S.A>France>Montbonnot
Public>Euronext>ALKAL.PA
AI Edge Inference Processor > MPPA DPU Coolidge Processor
https://www.kalrayinc.com/wp-content/uploads/2023/10/WP_Kalray_MPPA_DPU_Coolidge_june2022.pdf
Kinara Inc>USA>CA>Santa Clara
AI Edge Inference Processor>Ara-2 AI processor>Edge AI acceleration
CNN Transformers>Vision Transformers
https://staging.kinara.ai/products/kinara-ara-2/
Kortiq GmbH>Germany>Munich
http://www.kortiq.com/
AI Edge Inference Processor IP Core>FPGA based Neural Network Engine IP Core>AI Vision Processor>AIScaleCDP2 IP Core
https://www.linkedin.com/company/kortiq/about/
Lightmatter>USA>MA>Boston
Envise 4S>4U Server >With 16 Envise Chips>Envise Inference System
https://lightmatter.co/products/envise/
https://www.nature.com/articles/s41586-025-08854-x
Luminous Computing (Closed)>USA>CA>Mountain View
AI Photonics Inference Processor>Supercomputer
Macronix>Taiwan>Hsinchu
Public>TWSE>2337.TW
Analog Compute in Memory>Compute in NAND Flash Memory>Analog Computing in Memory (CIM) Technique for General Matrix Multiplication (GEMM) to Support Deep Neural Network (DNN) and Cosine Similarity Search Computing using 3D AND type NOR Flash Devices
https://ieeexplore.ieee.org/abstract/document/10019495
Mediatek>Taiwan>Hsinchu
Public>TWSE>2454.TW
AI Inference Processor>NPU
https://www.mediatek.com/tek-talk-blogs/mediatek-npus-neuropilot-and-litert-are-ready-to-bring-power-ai-in-millions-of-devices
Meta>USA>CA>Menlo Park
Public>NASDAQ>META
AI Inference Processor
Meta Training and Inference Accelerator MTIA
https://ai.meta.com/blog/meta-training-inference-accelerator-AI-MTIA/
Microsoft>USA>Washington>Redmond
Public>NASDAQ>MSFT
AI Inference Processors> Maia 200
https://blogs.microsoft.com/blog/2026/01/26/maia-200-the-ai-accelerator-built-for-inference/
Myrtle AI>Myrtle.ai>UK>Cambridge
AI Inference FPGA Software for FPGA Based AI Inference Chips >Volla
Applications>Speech Processing>Fintech
https://www.prnewswire.com/news-releases/myrtleai-enables-microsecond-ml-inference-latencies-running-vollo-on-napatech-smartnics-302523069.html
Mythic>USA>CA>Palo Alto
AI Edge Processor>Analog Matrix Processor> Analog Compute Engine
https://mythic.ai/product/
Napatech AS>Denmark>Soeborg
Public>OSLO: NAPA.OL
AI Edge Processor Design>AMD-Xillinx FGPA Based>SmartNIC>Altera FPGA Based
https://www.xilinx.com/content/dam/xilinx/publications/solution-briefs/amd-napatech-solution-brief.pdf
https://www.prnewswire.com/news-releases/napatech-optimizes-datacenter-infrastructure-for-artificial- intelligence-302255806.html
https://www.napatech.com/products/
Neuchips Corp>Taiwan>Zhubei City
AI Inference Processor>Raptor n3000 Accelerator ASIC>Large Language Models
https://www.neuchips.ai/raptor-n3000
Red Hat (Neural Magic)>USA>NC>Raleigh
Public>NYSE>RHT
AI Inference Software>AI Algorithms>Compound Sparsity>CPU
Based Inference at GPU Speeds>Inference>Innovation
Red Hat Completes Acquisition of Neural Magic to Fuel Optimized Generative AI Innovation Across the Hybrid CloudRed Hat AI Factory with NVIDIA Accelerates the Path to Scalable Production AI
Optimizing YOLO26 Inferences with Neural Magic’s DeepSparse Engine
https://docs.ultralytics.com/integrations/neural-magic/
NeuReality
AI Inference Processor>N41>Computer Server on a Chip
https://www.neureality.ai/chip
NVIDIA>USA>CA>Santa Clara Public>NASDAQ>NVDA
AI Inference Platform>AI Edge Inference Processor>Deep Learning Accelerator
https://nationaltoday.com/us/ca/san-jose/news/2026/03/02/nvidia-developing-top-secret-ai-inference-chip-for-next-month-debut/
https://www.nvidia.com/en-us/search/?q=AI%20Inference%20Chips&page=1
Perceive(Amazon)>USA>CA>San Jose
AI Inference Processor>Erge Edge Inference Processor
https://www.geekwire.com/2024/amazon-to-acquire-perceive-for-80m-from-xperi-expanding-its-ai-technology-for-edge-devices/
POLYN Technology>Israel>Caesarea
Analog Neuromorphic Inference Processor>Mixed Signal>NeuroVoice>NeuroSense>Vibrosense>Neuromorphic Analog Signal Processing (NASP) Unit
https://polyn.ai/polyn-technology-announces-first-silicon-implemented-nasp-chip/
Positron AI>USA>Nevada>Reno
AI Inference Processor>Memory Bound Inference>Innovation>Asimov Processor
https://www.positron.ai/asimov
Quadric Inc>USA>CA>Burlingame
AI Edge Inference Processor>Chimera GPNPU
Applications>ML Inference>Vision Processor>CNN
https://quadric.ai/press-release/quadric-raises-30m-series-c-funding
Qualcomm>USA>CA>San Diego
Public>NASDAQ>QCOM .. AI Edge Inference Processor> Dragonwing™ QCS405 System-on-Chip> Audio AI Processor > Qualcomm Cloud AI 100
https://www.qualcomm.com/content/dam/qcomm-martech/dm-assets/documents/Prod-Brief-QCOM-Cloud-AI-100-Ultra.pdf
Rambus>USA>CA>San Jose
Public>NASDAQ>RMBS
AI Inference Processor DDR6 Memory Controller
https://www.rambus.com/blogs/deepx-rambus-and-samsung-foundry-collaborate-to-enable-efficient-edge-inferencing-applications/
Rebellions>Korea>Seoul
AI Inference Processor>Data Center Processor>Rebel Chiplet>LLM>UCIe
https://rebellions.ai/rebellions-product/rebel-quad/
Reduced Energy Microsystems>Asynchronous CNN Inference
https://techcrunch.com/2017/03/16/reduced-energy-microsystems-pits-startup-chip-chops-against-industry-giants/
Reexen>China>Shenzhen
AI Edge Inference Sensor>Analog>Mixed Signal>Neuromorphic>Near Edge AI Sensor Inference Processor>ADA 100>Wearables
https://reexen.com/newsinfo/785817.html
Renasas Electronics Corp>Japan>Tokyo .Public>.TSE> 6723.T
AI Edge Inference Processor>RA8P1 AI-Accelerated MCU with NPU
https://www.renesas.com/en/blogs/enable-high-performance-low-power-inference-your-edge-ai-applications
Saqence AI>USA>CA>Santa Clara
AI Inference Processor > Analog In Memory Compute>Analog CIM>https://www.sagence-ai.com/
SambaNova Systems>USA>CA>Palo Alto
AI Inference and Training Processor> Reconfigurable Dataflow Processing Unit RDPU>LLM>Software Defined Hardware
https://sambanova.ai/press/sambanova-unveils-fastest-chip-for-agentic-ai-collaborates-with-intel-and-raises-350m?hs_amp=true
Sapeon Korea>Korea> Kyonggi-doAI Inference Processor>Data Center Inference Processor>X330
https://rebellions.ai/newsroom/rebellions-and-sapeon-korea-complete-merger-launching-koreas-first-ai-chip-unicorn/
SemiBrain>Anaflash Subsidiary>South Korea
SEMIFIVE>Korea>Seoul
Public>KOSDAQ>490470.KQ
AI Inference Processor Platform>Quickturn AI Processor Design Services>Data Center>AI Vision Processor>ADAS>Image Recognition
https://semifive.com/solutions/soc-platform/ai-inference-platform
SiFive Inc>USA>CA>San Mateo
AI Inference Processor Kernel>Xnnpack>RISC-V
https://www.sifive.com/blog/sifive-accelerates-risc-v-vector-integration-in-xnnpack-for-optimized-ai-inference
Silicom>Israel>Kfar Sava
Public>NASDAQ>SILC
FPGA Based Programmable AI Inference Cards
https://www.silicom-usa.com/solutions/fpga-based-solutions/applications/fpga/
SiMa.ai>USA>CA>San Jose
AI Inference Edge Processor>ML>MLSoC>.https://sima.ai/mlsoc/
Supergate>Korea>Seoul
AI Edge Inference Processor>Supergate NPU>Vision Analytics> CCTV Processing>Video Stream Processing
https://supergate.cc/product/pynpu/
T-Head>Unit of Alibaba>China
Public Unit
AI Inference Processor>HuaHon
https://www.t-head.cn/product?id=6
Taalas>Canada>Ontario>Toronto Innovation
AI Inference Processor>Silicon Hardened AI Software Chip
AI Inference Software to Hardware Chips>Innovation
https://taalas.com/products/
https://www.forbes.com/sites/karlfreund/2026/02/19/taalas-launches-hardcore-chip-with-insane-ai-inference-performance/
Tachyum>USA>CA>Sunnyvale
AI Inference Processor>AI Training Processor>ML>DeepSeek>Prodigy
Applications>Finance>Fraud>Cyberattack
https://www.tachyum.com/media/press-releases/2025/06/03/tachyum-radically-cuts-the-cost-of-deepseek-by-quantizing-it-to-2-bits/As
Tenstorrent>Canada>Ontario>Toronto
AI Edge Inference Processor IP Cores>Tensix Processor>Wormhole>Blackhole
https://tenstorrent.com/ip
Think Silicon S.A. (Applied Materials)>Greece>Patras
AI Inference Processor>Vision Processor Unit (VPU) ..NEMA xNNVision Processing Unit>>VPU>CNN
https://think-silicon.com
TSMC>Taiwan>Hsinchu
Public>NYSE>TSMC
N40 RRAM Process>Compute-In-Memory (CIM)>Resistive Random Access <emory (RRAM).(MAC) workloads in deep neural networks (DNNs) inference>RRAM-based CIM Macro> With error correction code (ECC), called MAC 59.1 TOPS/W>70.9 GOPS/mm 2
A 40-nm 118.44-TOPS/W Voltage-Sensing Compute-in-Memory RRAM Macro With Write Verification and Multi-Bit Encoding | IEEE Journals & Magazine | IEEE Xplore
Untether AI(AMD)>Canada>Ontario>Toronto
AI Neural Net Inference Processor
AI FPGA Inference Processors
www.untether.ai
Untether AI Shuts Down, Engineering Team Joins AMD – EE Times
Xilinx(Advanced Micro Devices)
AI FPGA Inference Processors
https://www.xilinx.com/content/dam/xilinx/publications/solution-briefs/xilinx-versal-ai-compute-solution-brief.pdf
XMOS>UK>Bristol
AI Inference Processors>AI Inference Data Center Coprocessor>AI Server Coprocessor>AI Edge Inference Coprocessor
https://www.xmos.com/aihttps://www.xmos.com/xcore-ai