NextSilicon offers Maverick-2
Product
Maverick-2 is NextSilicon's runtime-reconfigurable dataflow accelerator for high-performance computing and AI workloads.
- Currently Offered
Maverick-2 is the production dataflow accelerator built on NextSilicon's Intelligent Compute Architecture. It profiles a running application and reconfigures its hardware at run time to accelerate the code's hot spots, so existing C, C++, Fortran, OpenMP, and Kokkos programs gain speed without porting or rewriting. NextSilicon ships the offering in a single-die PCIe Gen 5 card and a dual-die OAM form factor, both built on TSMC 5nm with HBM3E memory.
The offering targets HPC, AI, and vector-database workloads across energy, financial, life-sciences, graph, and manufacturing computing. NextSilicon pairs the accelerator with developer tools that expose run-time telemetry, chip-level state, and per-core projections to the programmer. Deployments reported by the company and its partners center on scientific and national-security computing, including evaluation of Maverick-based systems inside the Vanguard program at Sandia National Laboratories.
Maverick-2 is NextSilicon's runtime-reconfigurable dataflow accelerator for high-performance computing and AI workloads.
Broadcom designs and supplies semiconductors for networking, wireless, broadband, and industrial markets.
Rack-mounted AI inference system integrating proven compute IP with photonic interconnect and large, high-bandwidth memory.
Groq LPU is a custom inference chip with on-chip SRAM and deterministic scheduling delivering low-latency AI inference for LLMs.
Purpose-built AI supercomputers integrating GPUs, CPUs, networking, and software for enterprise training and inference.
NVIDIA B200 GPU for high-performance AI training and inference.
Broadcom designs and supplies semiconductors for networking, wireless, broadband, and industrial markets.
MatX One is a high-throughput chip for large language model training and inference.
NVIDIA Hopper is a data-center GPU architecture for AI and high-performance computing workloads.
NVIDIA Blackwell is a data-center GPU architecture for generative AI and accelerated computing.
Modular reference platform letting server makers build NVLink-connected GPU systems for hyperscale AI and HPC.
Rack-mounted AI inference system integrating proven compute IP with photonic interconnect and large, high-bandwidth memory.
TPU is Google's custom AI accelerator ASIC family delivered through Google Cloud TPU Pods for ML training and inference.
Intel Gaudi is a data-center AI accelerator product family.
