The startup develops an AI-based NeuroMosAIc Processor (NMP) that integrates a RISC-V architecture for high-performance computing in semiconductor applications. Its technology enables clients to efficiently evaluate neural network performance metrics such as accuracy, memory bandwidth, and run-time using SDK solutions compatible with TensorFlow, Caffe, PyTorch, and ONNX frameworks.
Funding
$8M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
The increasing complexity and performance demands of AI models, especially in edge computing applications, often exceed the capabilities of commodity compute offerings in terms of power consumption, cost, and multi-sensor data processing. This makes it difficult to deploy advanced AI functionalities on resource-constrained devices.
Solution
AiM Future provides a family of NeuroMosAIc Processors (NMP) and a software studio designed to address the challenges of edge AI deployment. The NMP IP cores are fully configurable and pre-configured AI accelerators that enhance performance and energy efficiency for various applications. The NeuroMosAIc Studio offers a comprehensive set of tools for developers to run pre-trained machine learning models on the NMP hardware, including a hardware-aware model converter, mapper, and profiler. This combination enables the development of new device classes, ranging from battery-operated sensors to edge gateways and servers, with optimized performance and power consumption.
Target Audience
The primary target audience includes semiconductor designers, product planners, and system integrators in the IoT, consumer electronics, automotive, and edge computing markets who require efficient AI acceleration solutions for their devices.
Features
- NeuroMosAIc Processor (NMP) family with tiered performance options: NMP-350 (up to 1 TOPS), NMP-550 (up to 6 TOPS), and NMP-750 (up to 16 TOPS)
- Support for concurrent multimodal inference, enabling simultaneous processing of multiple real-time sensor data streams
- Compatibility with industry-standard CPUs, including Arm Cortex-A and Cortex-M processors
- NeuroMosAIc Studio software suite for model conversion, mapping, compilation, simulation, and profiling
- Support for industry-standard frameworks such as ONNX, PyTorch, and TensorFlow Lite
- Hardware-aware model converter, mapper, and profiler for maximizing efficiency
- Advanced quantization, pruning, and sparsity techniques for optimizing accuracy, latency, and memory size
- GAIA architecture designed to enable energy-efficient transformers and large language models (LLMs)