In-Memory Computing
Processing-in-memory architectures that compute directly in storage to eliminate data movement bottlenecks.
Synthara
CH · Zurich, SwitzerlandComputeRAM in-memory computing for embedded applications. Backed by Hermann Hauser and Pat Gelsinger.
Axelera AI
NL · Eindhoven, NetherlandsEuropean edge AI chips. Metis AIPU, in-memory computing.
MetisX (XCENA)
KR · Seongnam, South KoreaMemory-centric data domain specific architecture based on CXL. Won "Most Innovative Memory Technology" at FMS 2025.
RAAAM Memory
IL · IsraelAnalog memory for AI. In-memory computing technology.
Vertical Compute
BE · Louvain-La-Neuve, Belgiumimec spinoff developing vertically integrated memory and compute for AI accelerators with chiplet-based solutions.
Crossbar Inc.
US · Santa Clara, CAReRAM pioneer. Microsemi partnership for embedded non-volatile memory.
Avalanche Technology
US · Fremont, CAPerpendicular STT-MRAM (SPMEM) with proprietary spin current and voltage switching for aerospace and defense.
Syntiant
US · Irvine, CANDP100/120/250 Neural Decision Processors. 100M+ units deployed, at-memory architecture.
XConn Technologies
US · San Jose, CAFirst hybrid switch supporting both CXL 2.0 and PCIe Gen5 on single chip for memory pooling.
Rain AI
US · Redwood City, CANeuromorphic in-memory chips. $150M Series B collapsed Q1 2025 (investor pullout). Bridge raised ($3M only). In acqui-hire process — NVIDIA, OpenAI, Microsoft interested. Initial NPU shipments to select partners Oct 2024.
Panmnesia
KR · South KoreaCXL memory pooling. $60M Series A, Samsung collaboration.
Axelera AI
NL · Eindhoven, NetherlandsDigital In-Memory Computing (D-IMC). Europa AIPU (Oct 2025): 629 TOPS INT8, 64GB memory. $250M+ Series B (Feb 2026) — Europe's largest AI semiconductor round. 500+ global customers across defense, industrial, robotics, security.
d-Matrix
US · Santa Clara, CAIn-memory computing. Corsair platform: 10× GPU performance, 3× lower cost, 3–5× better energy efficiency. Includes Corsair accelerators, JetStream networking ASICs, Aviator software. $275M Series C (Nov 2025, $2B valuation, $450M total). Temasek, QIA, Microsoft M12 investors.
Anabrid
DE · Frankfurt, GermanyModern analog computers on a chip. Hybrid analog-digital computing 1000x more energy efficient. SPRIND-backed.
Nanya Technology
TW · Taipei, TaiwanTaiwan's #4 global DRAM manufacturer. Developing customized HBM for edge AI including AI PCs and robotics.
GSI Technology
US · Sunnyvale, CAAssociative Processing Unit (APU) delivering GPU-class AI performance with 98% lower energy via compute-in-SRAM.
Netsol
KR · Gyeonggi, South KoreaSTT-MRAM and SRAM solutions for industrial automation, medical devices, and IoT. Mass production since 2023.
TetraMem
US · United StatesAnalog in-memory computing chips. ReRAM-based AI accelerators.
Weebit Nano
AU · Hod HaSharon, IsraelReRAM commercialization. Multiple foundry collaborations for embedded NVM.
LightIC
CN · ChinaPhotonic in-memory computing using phase-change materials.
Everspin Technologies
US · Chandler, AZMarket leader in discrete MRAM. Industrial high-density STT-MRAM expansion (Oct 2024).
MemComputing
US · San Diego, CAPhysics-based computing using memcomputing processors for complex optimization problems.
Metisix
DE · Munich, GermanyNext-generation memory solutions for AI and HPC. Novel memory architectures for data-intensive computing.
UPMEM
FR · Grenoble, FranceProcessing-in-Memory acceleration with hundreds of programmable coprocessors integrated into DRAM. Acquired by Qualcomm.
Fractile
GB · United KingdomIn-memory processing. $15M seed funding.
Neo Semiconductor
US · United StatesAnalog in-memory computing. Early stage development.
Blueshift Memory
GB · Cambridge, UKNear-data processing memory architecture. Reduces data movement for AI and analytics workloads.
EnCharge AI
US · Santa Clara, CAAnalog in-memory computing. $100M Series B (2025), DARPA-backed.
Crossbar Inc
US · Santa Clara, CAReRAM technology. Non-volatile memory and in-memory computing.
MemVerge
US · San Jose, CABig Memory Computing software including Memory Machine products for CXL memory pooling and tiering.
Untether AI
CA · Toronto, CanadaAt-memory architecture. speedAI240 with 2 PetaOps. Acquired by AMD (2025).
4DS Memory
AU · Perth, AustraliaNon-filamentary interface switching ReRAM for high-density gigabyte storage devices.
HyperCIM
US · San Jose, CACompute-in-memory AI accelerators for edge inference. Ultra-low power neural network processing.
ZeroPoint Technologies
SE · Gothenburg, SwedenMemory compression IP. Reducing memory footprint for AI workloads.
Etron Technology
TW · Hsinchu, TaiwanFabless IC design pioneering KGDM bare-die memories. JV with Nanya for ultra-high-bandwidth AI memory.