All IPs > Processor > AI Processor
The AI Processor category within our semiconductor IP catalog is dedicated to state-of-the-art technologies that empower artificial intelligence applications across various industries. AI processors are specialized computing engines designed to accelerate machine learning tasks and perform complex algorithms efficiently. This category includes a diverse collection of semiconductor IPs that are built to enhance both performance and power efficiency in AI-driven devices.
AI processors play a critical role in the emerging world of AI and machine learning, where fast processing of vast datasets is crucial. These processors can be found in a range of applications from consumer electronics like smartphones and smart home devices to advanced robotics and autonomous vehicles. By facilitating rapid computations necessary for AI tasks such as neural network training and inference, these IP cores enable smarter, more responsive, and capable systems.
In this category, developers and designers will find semiconductor IPs that provide various levels of processing power and architectural designs to suit different AI applications, including neural processing units (NPUs), tensor processing units (TPUs), and other AI accelerators. The availability of such highly specialized IPs ensures that developers can integrate AI functionalities into their products swiftly and efficiently, reducing development time and costs.
As AI technology continues to evolve, the demand for robust and scalable AI processors increases. Our semiconductor IP offerings in this category are designed to meet the challenges of rapidly advancing AI technologies, ensuring that products are future-ready and equipped to handle the complexities of tomorrow’s intelligence-driven tasks. Explore this category to find cutting-edge solutions that drive innovation in artificial intelligence systems today.
Brainchip's Akida Neural Processor IP represents a groundbreaking approach to edge AI processing by employing a neuromorphic design that mimics natural brain function for efficient and accurate data processing directly on the device. This IP stands out due to its event-based processing capability, which significantly reduces power consumption while providing high-speed inferencing and on-the-fly learning. Akida's architecture is designed to operate independently of traditional cloud services, thereby enhancing data privacy and security. This localized processing approach enables real-time systems to act on immediate sensor inputs, offering instantaneous reactions. Additionally, the architecture supports flexible neural network configurations, allowing it to adapt to various tasks by tailoring the processing nodes to specific application needs. The Akida Neural Processor IP is supported by Brainchip's MetaTF software, which simplifies the creation and deployment of AI models by providing tools for model conversion and optimization. Moreover, the platform's inherent scalability and customization features make it versatile for numerous industry applications, including smart home devices, automotive systems, and more.
The 2nd Generation Akida platform is a substantial advancement in Brainchip's neuromorphic processing technology, expanding its efficiency and applicability across more complex neural network models. This advanced platform introduces support for Temporal Event-Based Neural Nets and Vision Transformers, aiming to enhance AI performance for various spatio-temporal and sensory applications. It's designed to drastically cut model size and required computations while boosting accuracy. Akida 2nd Generation continues to enable Edge AI solutions by integrating features that improve energy efficiency and processing speed while keeping model storage requirements low. This makes it an ideal choice for applications that demand high-performance AI in Edge devices without needing cloud connectivity. Additionally, it incorporates on-chip learning, which eliminates the need to send sensitive data to the cloud, thus enhancing security and privacy. The platform is highly flexible and scalable, accommodating a wide array of sensory data types and applications, from real-time robotics to healthcare monitoring. It's specifically crafted to run independently of the host CPU, enabling efficient processing in compact hardware setups. With this generation, Brainchip sets a new standard for intelligent, power-efficient solutions at the edge.
The NMP-750 is engineered as a performance accelerator tailored for edge computing applications that demand robust processing power and versatility. It is ideally deployed in environments such as automotive, AMR and UAV systems, AR/VR applications, as well as smart infrastructure projects like smart buildings, factories, and cities. Its design principles aim to enhance security and surveillance systems while supporting advanced telecommunications solutions. This comprehensive IP can attain up to 16 TOPS, thereby addressing needs for high throughput and efficiency in data processing tasks. The NMP-750 includes up to 16 MB of local memory, utilizing either RISC-V or Arm Cortex-R or A 32-bit CPUs to manage operational complexity through three 128-bit AXI4 interfaces for host, CPU, and data processes. This infrastructure not only ensures rapid data-handling capabilities but also optimizes system-level operations for various emerging technologies. Ideal for managing multi-camera stream processing and enhancing spectral efficiency, it is equally suited for mobility and autonomous control systems—key to future smart city and factory applications. The NMP-750's support for comprehensive automation and data analytics offers companies the potential to develop cutting-edge technologies, driving industry standards across domains.
ADAS and Autonomous Driving technology by KPIT focuses on advancing L3+ autonomy, providing scalable and safe autonomous mobility solutions. This technology addresses fundamental challenges such as consumer safety, localized infrastructure dependencies, and comprehensive validation approaches. With the ever-evolving landscape of autonomous driving, ensuring robust AI solutions beyond mere perception is crucial for elevating autonomy levels in vehicles. By integrating innovative technology and adhering to regulatory standards, KPIT empowers automakers to offer safe and reliable autonomous vehicles that meet consumer trust and performance expectations.
The Origin E1 is a highly efficient neural processing unit (NPU) designed for always-on applications across home appliances, smartphones, and edge nodes. It is engineered to deliver approximately 1 Tera Operations per Second (TOPS) and is tailored for cost- and area-sensitive deployment. Featuring the LittleNPU architecture, the Origin E1 excels in low-power environments, making it an ideal solution for devices where minimal power consumption and area are critical. This NPU capitalizes on Expedera's innovative packet-based execution strategy, which allows it to perform parallel layer execution for optimal resource use, cutting down on latency, power, and silicon area. The E1 supports a variety of network types commonly used in consumer electronics, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and more. A significant advantage of Origin E1 is its scalability and market-leading power efficiency, achieving 18 TOPS/W and supporting standard, custom, and proprietary networks. With a robust software stack and support for popular AI frameworks like TensorFlow and ONNX, it ensures seamless integration into a diverse range of AI applications.
The Origin E8 neural processing unit (NPU) stands out for its extreme performance capabilities, designed to serve demanding applications such as high-end automotive systems and data centers. Capable of delivering up to 128 TOPS per core, this NPU supports the most advanced AI workloads seamlessly, whether in autonomous vehicles or data-intensive environments. By employing Expedera's packet-based architecture, Origin E8 ensures efficient parallel processing across layers and achieves impressive scalability without the drawbacks of increased power and area penalties associated with tiled architectures. It allows running extensive AI models that cater to both standard and custom requirements without compromising on model accuracy. The NPU features a comprehensive software stack and full support for a variety of frameworks, ensuring ease of deployment across platforms. Scalability up to PetaOps and support for resolutions as high as 8K make the Origin E8 an excellent solution for industries that demand unrivaled performance and adaptability.
The NMP-350 is a low-power and cost-effective end-point accelerator designed to cater to applications across various industries. This accelerator finds its niche in markets such as automotive, AIoT, and Industry 4.0, where efficiency and scalability are critical. With potential applications in driver authentication, digital mirrors, and personalized user experiences, it is also applicable in predictive maintenance systems, machine automation, and health monitoring. Technically, the NMP-350 boasts an impressive capacity of up to 1 TOPS (Tera Operations Per Second), supported by up to 1 MB of local memory. The system is based on a flexible architecture utilizing either RISC-V or Arm Cortex-M 32-bit CPUs, accommodating three AXI4 interfaces with 128 bits each dedicated to host, CPU, and data processes. This composition assures its capability to handle a multitude of tasks efficiently while maintaining a low power profile. Its integration into smart appliances and wearable technologies showcases its versatility, providing industry players with a robust solution for building smarter and more reliable products. As industries move towards more interconnected and intelligent systems, the NMP-350 provides the necessary technology to drive innovation forward.
The Origin E2 is a versatile, power- and area-optimized neural processing unit (NPU) designed to enhance AI performance in smartphones, edge nodes, and consumer devices. This NPU supports a broad range of AI networks such as RNNs, LSTMs, CNNs, DNNs, and others, ensuring minimal latency while optimizing for power and area efficiency. Origin E2 is notable for its adaptable architecture, which facilitates seamless parallel execution across multiple neural network layers, thus maximizing resource utilization and providing deterministic performance. With performance capabilities scalable from 1 to 20 TOPS, the Origin E2 maintains excellent efficiency up to 18 TOPS per Watt, reflecting its superior design strategy over traditional layer-based solutions. This NPU's software stack supports prevalent frameworks like TensorFlow and ONNX, equipped with features such as mixed precision quantization and multi-job APIs. It’s particularly suitable for applications that require efficient processing of video, audio, and text-based neural networks, offering leading-edge performance in power-constrained environments.
The NMP-550 stands as a performance efficiency accelerator, particularly crafted for applications that demand high computational power combined with energy efficiency. This IP is especially suited for markets including automotive, mobile devices, AR/VR, and security-focused technologies. Its applicability spares a wide spectrum, fostering innovation in driver monitoring, fleet management, and advanced image or video analytics. Along with intruder detection and compliance systems, it bolsters its utility in medical devices for enhanced diagnostic capabilities. Technologically, the NMP-550 delivers up to 6 TOPS, which provides a significant boost in data processing capability. It features up to 6 MB of local memory, ensuring swift and effective data management. The design is underpinned by a choice of RISC-V or Arm Cortex-M or A 32-bit CPUs, along with three AXI4 interfaces supporting 128 bits each, allocated for host, CPU, and data handling. Such specification allows this accelerator to proficiently tackle tasks of various computational demands with resilience and efficiency. Its design caters to cross-disciplinary needs, making it an excellent fit for drone operations, robotics, and security systems requiring real-time processing and decision-making capabilities. With the inherent ability to process substantially more data at improved efficiencies, this IP aligns well with the future of immersive and interactive application deployments.
The Chimera GPNPU stands as a powerful neural processing unit tailor-made for on-device AI computing. This processor architecture revolutionizes the landscape of SoC design, providing a unified execution pipeline that integrates both matrix and vector operations with control code typically handled by separate cores. Such integration boosts developer productivity and enhances performance significantly. The Chimera GPNPU's ability to run diverse AI models—including classical backbones, vision transformers, and large language models—demonstrates its adaptability to future AI developments. Its scalable design enables handling of extensive computational workloads reaching up to 864 TOPs, making it suitable for a wide array of applications including automotive-grade AI solutions. This licensable processor core is built with a unique hybrid architecture that combines Von Neuman and 2D SIMD matrix instructions, facilitating efficient execution of a myriad array of data processing tasks. The Chimera GPNPU has been optimized for integration, allowing seamless incorporation into modern SoC designs for high-speed and power-efficient computing. Key features include a robust instruction set tailored for ML tasks, effective memory optimization strategies, and a systematic approach to on-chip data handling, all working to minimize power usage while maximizing throughput and computational accuracy. Furthermore, the Chimera GPNPU not only meets contemporary demands of AI processing but is forward-compatible with potential advancements in machine learning models. Through comprehensive safety enhancements, it addresses stringent automotive safety requirements, ensuring reliable performance in critical applications like ADAS and enhanced in-cabin monitoring systems. This combination of performance, efficiency, and scalability positions the Chimera GPNPU as a pivotal tool in the advancement of AI-driven technologies within industries demanding high reliability and long-term support.
Cortus's High Performance RISC-V Processor represents the pinnacle of processing capability, designed for demanding applications that require high-speed computing and efficient task handling. It features the world’s fastest RISC-V 64-bit instruction set architecture, implemented in an Out-of-Order (OoO) execution core, supporting both single-core and multi-core configurations for unparalleled processing throughput. This processor is particularly suited for high-end computing tasks in environments ranging from desktop computing to artificial intelligence workloads. With integrated features such as a multi-socket cache coherent system and an on-chip vector plus AI accelerator, it delivers exceptional computation power, essential for tasks such as bioinformatics and complex machine learning models. Moreover, the processor includes coherent off-chip accelerators, such as CNN accelerators, enhancing its utility in AI-driven applications. The design flexibility extends its application to consumer electronics like laptops and supercomputers, positioning the High Performance RISC-V Processor as an integral part of next-gen technology solutions across multiple domains.
The Automotive AI Inference SoC by Cortus is a cutting-edge chip designed to revolutionize image processing and artificial intelligence applications in advanced driver-assistance systems (ADAS). Leveraging RISC-V expertise, this SoC is engineered for low power and high performance, particularly suited to the rigorous demands of autonomous driving and smart city infrastructures. Built to support Level 2 to Level 4 autonomous driving standards, this AI Inference SoC features powerful processing capabilities, enabling complex image processing algorithms akin to those used in advanced visual recognition tasks. Designed for mid to high-end automotive markets, it offers adaptability and precision, key to enhancing the safety and efficiency of driver support systems. The chip's architecture allows it to handle a tremendous amount of data throughput, crucial for real-time decision-making required in dynamic automotive environments. With its advanced processing efficiency and low power consumption, the Automotive AI Inference SoC stands as a pivotal component in the evolution of intelligent transportation systems.
The Origin E6 NPU is engineered for high-performance on-device AI tasks in smartphones, AR/VR headsets, and other consumer electronics requiring cutting-edge AI models and technologies. This neural processing unit balances power and performance effectively, delivering between 16 to 32 TOPS per core while catering to a range of AI workloads including image transformers and point cloud analysis. Utilizing Expedera’s unique packet-based architecture, the Origin E6 offers superior resource utilization and ensures performance with deterministic latency, avoiding the penalties typically associated with tiled architectures. Origin E6 supports advanced AI models such as Stable Diffusion and Transformers, providing optimal performance for both current and predicted future AI workloads. The NPU integrates seamlessly into chip designs with a comprehensive software stack supporting popular AI frameworks. Its field-proven architecture, deployed in millions of devices, offers manufacturers the flexibility to design AI-enabled devices that maximize user experience while maintaining cost efficiency.
The KL730 AI SoC is equipped with a state-of-the-art third-generation reconfigurable NPU architecture, delivering up to 8 TOPS of computational power. This innovative architecture enhances computational efficiency, particularly with the latest CNN networks and transformer applications, while reducing DDR bandwidth demands. The KL730 excels in video processing, offering support for 4K 60FPS output and boasts capabilities like noise reduction, wide dynamic range, and low-light imaging. It is ideal for applications such as intelligent security, autonomous driving, and video conferencing.
Akida IP is Brainchip's pioneering neuromorphic processor technology that mimics the human brain's function to efficiently analyze sensor inputs directly at the acquisition point. This digital processor achieves outstanding performance while significantly lowering power consumption and maintaining high processing precision. Edge AI tasks are handled locally on the chip, minimizing latency and enhancing privacy through reduced cloud dependence. The scalable architecture supports up to 256 nodes interconnected via a mesh network, each featuring Neural Network Layer Engines that can be tailored for convolutional or fully connected operations. The event-based processing technology of Akida leverages the natural data sparsity found in activations and weights, cutting down the number of operations by significant margins, thus saving power and improving performance. As a highly adaptable platform, Akida supports on-chip learning and various quantization options, ensuring customized AI solutions without costly cloud retraining. This approach not only secures data privacy but also lowers operational costs, offering edge AI solutions with unprecedented speed and efficiency. Akida's breakthrough capabilities address core issues in AI processing, such as data movement, by instantiating neural networks directly in hardware. This leads to reduced power consumption and increased processing speed. Furthermore, the IP is supported by robust tools and an environment conducive to easy integration and deployment, making it highly attractive to industries seeking efficient, scalable AI solutions.
The eSi-Floating Point component provides robust floating point capabilities to eSi-RISC embedded processor cores. This feature is crucial for applications requiring high precision and complex arithmetic processing, such as digital signal processing and scientific computations. The component supports both single and double-precision floating point operations, adhering to the IEEE-754 standard. Designed for efficiency, eSi-Floating Point optimizes resource use while maximizing computational performance, making it suitable for resource-constrained environments without sacrificing precision. This component's architecture enables significant performance improvements in data processing tasks, allowing for enhanced data throughput and reduced computational time. eSi-Floating Point integrates seamlessly with the eSi-RISC architecture, providing a unified system solution that elevates processing capabilities without extensive redesigns. Its use in applications demanding precision calculation and high-speed processing emphasizes its value in fields such as audio processing, high-accuracy sensor hubs, and control systems.
The Neural Processing Unit (NPU) from OPENEDGES offers a state-of-the-art deep learning accelerator, optimized for edge computing with advanced mixed-precision computation. Featuring a powerful network compiler for efficient memory usage, it handles complex neural network operations while minimizing DRAM traffic. Its layered architecture supports modern algorithmic needs, including transformers, allowing for parallel processing of neural layers. The NPU provides significant improvements in compute density and energy efficiency, targeting applications from automotive to surveillance, where high-speed, low-power processing is critical.
NeuroMosAIc Studio is a comprehensive software platform designed to accelerate AI development and deployment across various domains. This platform serves as an essential toolkit for transforming neural network models into hardware-optimized formats specific for AiM Future's accelerators. With broad functionalities including conversion, quantization, compression, and optimization of neural networks, it empowers AI developers to enhance model performance and efficiency. The studio facilitates advanced precision analysis and adjustment, ensuring models are tuned to operate optimally within hardware constraints while maintaining accuracy. Its capability to generate C code and provide runtime libraries aids in seamless integration within target environments, enhancing the capability of developers to leverage AI accelerators fully. Through this suite, companies gain access to an array of tools including an NMP compiler, simulator, and support for NMP-aware training. These tools allow for optimized training stages and quantization of models, providing significant operational benefits in AI-powered solutions. NeuroMosAIc Studio, therefore, contributes to reducing development cycles and costs while ensuring top-notch performance of deployed AI applications.
The Metis AIPU PCIe AI Accelerator Card represents a powerful computing solution for high-demand AI applications. This card, equipped with a single Metis AI Processing Unit, delivers extraordinary processing capabilities, reaching up to 214 Tera Operations Per Second (TOPS). Designed to handle intensive computing tasks, it is particularly suited for applications requiring substantial computational power and rapid data processing, such as real-time video analytics and AI-driven operations in various industrial and retail environments. This accelerator card integrates seamlessly into PCIe slots, providing developers with an easy-to-deploy solution enhanced by Axelera AI's Voyager Software Development Kit. The kit simplifies the deployment of neural networks, making it a practical tool for both seasoned developers and newcomers to AI technology. The card's power efficiency is a standout feature, aimed at reducing operational costs while ensuring optimal performance. With its innovative architecture, the Metis AIPU PCIe AI Accelerator Card not only meets but exceeds the needs of modern AI applications, ensuring users can harness significant processing power without the overheads associated with traditional systems.
The ULYSS MCU range from Cortus is a powerful suite of automotive microcontrollers designed to address the complex demands of modern automotive applications. These MCUs are anchored by a highly optimized 32/64-bit RISC-V architecture, delivering impressive performance levels from 120MHz to 1.5GHz, making them suitable for a variety of automotive functions such as body control, safety systems, and infotainment. ULYSS MCUs are engineered to accommodate extensive application domains, providing reliability and efficiency within harsh automotive environments. They feature advanced processing capabilities and are designed to integrate seamlessly into various automotive systems, offering developers a versatile platform for building next-generation automotive solutions. The ULYSS MCU family stands out for its scalability and adaptability, enabling manufacturers to design robust automotive electronics tailored to specific needs while ensuring cost-effectiveness. With their support for a wide range of automotive networking and control applications, ULYSS MCUs are pivotal in the development of reliable, state-of-the-art automotive systems.
The Low Power RISC-V CPU IP from SkyeChip is crafted to deliver efficient computation with minimal power consumption. Featuring the RISC-V RV32 instruction set, it supports a range of functions with full standard compliance for instruction sets and partial support where necessary. Designed exclusively for machine mode, it incorporates multiple vectorized interrupts and includes comprehensive debugging capabilities. This CPU IP is well-suited for integration into embedded systems where power efficiency and processing capability are crucial.
The MIPI CSI-2 Receiver IP by Arasan is an integral component for high-speed data transmission in camera applications. This IP facilitates the camera sensor to processor interface, adhering to the MIPI specification standard. It supports performance enhancements with high bandwidth, allowing seamless capture of images and video. The CSI-2 Receiver IP is engineered to handle significant data loads, enabling real-time processing of high-resolution images. Its design includes features for error correction and auto-calibration, contributing to reliability and data integrity. Compliance with the MIPI specification ensures broad compatibility with a range of camera sensors and microcontrollers. Configurability is a key advantage, with support for various data formats and transmission modes tailored for specific user needs. This adaptability simplifies the integration process across different applications, from smartphones and tablets to automotive camera systems, ensuring consistent performance across the board.
The RWM6050 Baseband Modem represents a leap in cost-effectiveness and power efficiency for applications requiring high bandwidth and capacity in mmWave technology. Designed in partnership with Renesas, this modem can effectively pair with mmWave RF chipsets to fulfill various access and backhaul market needs. With a flexible channel structure and modulation coding, it ensures scalability for multi-gigabit data transmission. The modem's platform is designed for high configurability, and includes subsystems for beamforming and digital signal processing. It stands out for its real-time programmable scheduler and integrated network synchronization, boosting throughput for numerous demanding applications. The RWM6050 combines power with efficiency, facilitating seamless and substantial data flow over several hundred meters with dual modem support providing redundancy and resilience.
The KL630 AI SoC embodies next-generation AI chip technology with a pioneering NPU architecture. It uniquely supports Int4 precision and transformer networks, offering superb computational efficiency combined with low power consumption. Utilizing an ARM Cortex A5 CPU, it supports a range of AI frameworks and is built to handle scenarios from smart security to automotives, providing robust capability in both high and low light conditions.
Altek's AI Camera Module exemplifies innovation in the realm of smart imaging solutions, designed to serve as a critical component in AI recognition and video processing systems. This module integrates advanced image processing capabilities, enabling it to deliver superior high-resolution images that are indispensable for AI-driven applications. With an expert blend of lens design and software integration, the module achieves optimal performance in AI and IoT contexts. This modular solution is highly adaptable, supporting edge computing to meet real-time data processing needs. It can cater to high-resolution demands such as 2K and 4K video quality, enhancing detail and clarity for surveillance or autonomous platforms. Its rich functionality spans a range of use cases, from facial recognition and tracking to complex video analytics, ensuring clients have a flexible solution that fits into various AI ecosystems. Altek’s AI Camera Module is designed for seamless integration, offering capabilities that span across consumer electronics, industrial applications, and smart cities. It stands out by providing robust performance and high adaptability to different environments, harnessing machine learning algorithms to improve precision and efficiency. The module's collaboration potential with global brands underlines its reliability and advanced technological framework, making it a go-to choice for organizations aiming to excel in high-end AI+IoT implementations.
The ONNC Calibrator is crafted to optimize AI System-on-Chips by employing post-training quantization (PTQ) techniques to maintain high precision, especially in architectures using fixed-point formats like INT8. By leveraging architecture-aware quantization, it ensures chips retain 99.99% accuracy, offering unparalleled precision control across diverse hardware configurations. This calibrator supports configurable bit-width architectures, allowing the balance of precision and performance to be tailored for various applications. Capable of working with different AI frameworks such as ONNX and PyTorch, the calibrator aligns seamlessly with standard PTQ workflows without needing complex retraining. Its internal AI engine autonomously determines optimal scaling factors, making it an indispensable tool in maintaining model accuracy while reducing computational demand.
The Dynamic Neural Accelerator II (DNA-II) by EdgeCortix represents a new leap in neural network processing. This IP core is exceptionally efficient and provides scalable performance by reconfiguring runtime interconnects between its computing units. Supporting both convolutional and transformer models, DNA-II is tailored for edge AI tasks that demand high parallel processing. The modular DNA-II stands out by enhancing parallelism and optimizing on-chip memory bandwidth usage. Its symbiotic relationship with software solutions like the MERA compiler boosts its efficacy in deploying neural networks across varied applications, from smart cities to automotive systems.
The PentaG RAN platform is designed for emerging 5G network infrastructure, providing a comprehensive baseband solution for 5G RAN ASICs and Open RAN systems. It is recognized for industry-leading performance and scalability, tailored to meet the rigorous demands of modern communications networks. The platform supports a variety of network configurations and use cases, ensuring the facilitation of seamless communication in both urban and rural settings.
The Prodigy Universal Processor by Tachyum Inc. is engineered to integrate the functionalities of CPUs, GPGPUs, and TPUs into a unified, efficient architecture. This makes it an ideal candidate for applications demanding high performance, such as AI, high-performance computing (HPC), and hyperscale data centers. The processor stands out with its capacity to offer unparalleled performance while maintaining reduced energy consumption and maximizing utilization rates within server environments. Prodigy processors boast multiple SKUs offering configurations like 64-bit cores per socket running at speeds beyond 5 GHz, support for numerous DDR5 memory channels, and extensive PCI Express 5.0 lanes for high data throughput. These features enable the Prodigy to handle diverse and intensive computational tasks with ease, reducing the necessity for separate heterogeneous processing units. One of the key advantages of the Prodigy architecture is its ability to execute existing x86 applications without any modifications through a robust emulation layer. This capability simplifies transitions for enterprises looking to consolidate their systems under the Prodigy umbrella, allowing for significant operational efficiencies and cost savings. The Prodigy processor thus positions itself as a future-proof choice for enterprises aiming to modernize their data processing capabilities.
Topaz FPGAs are crafted for applications that require high-performance and cost-effective solutions with a focus on low power usage. Designed for volume production, these FPGAs leverage a unique architecture that maximizes logic utilization, facilitating a broad spectrum of applications from industrial automation to consumer electronics. These FPGAs support a variety of standards such as PCIe Gen3, MIPI, and Ethernet, making them versatile for communications and data processing tasks. Their robust protocol support allows integration into systems requiring machine vision, robotics, and broadcasting capabilities. Topaz's flexible and efficient architecture also allows for seamless migration to Titanium FPGAs if enhanced performance is necessary. A notable feature of Topaz FPGAs is their commitment to longevity and reliability. Efinix ensures stable production support for Topaz FPGAs well into the future, promising long-term reliability in embedded systems that demand uninterrupted performance. This durability and adaptability make Topaz FPGAs an excellent choice for industries that revolve around innovative and evolving tech solutions.
The xcore.ai product line from XMOS represents a pioneering approach towards versatile and high-performance microcontroller solutions. Engineered to blend control, DSP, artificial intelligence, and low-latency input/output processing, the xcore.ai platform is optimized for a wide range of applications. This includes consumer electronics, industrial automation, and automotive industries where real-time data processing and robust computational power are crucial. With its advanced processing capabilities, xcore.ai facilitates the development of smart products by integrating AI functions directly into devices, making them more responsive and capable. This line of microcontrollers supports audio signal processing and voice control technologies, which are essential for modern smart home and entertainment applications. xcore.ai is uniquely designed to handle multiple data streams with precision while maintaining the low power consumption needed for sustainable product development. The product leverages XMOS's commitment to providing cycle-accurate software programmability, which allows developers to quickly adapt and customize hardware functions to meet specific needs. By fostering an environment where software and hardware seamlessly interact, xcore.ai not only supports rapid prototyping and deployment but also ensures long-term durability in demanding environments.
The CTAccel Image Processor (CIP) on Intel Agilex FPGA offers a high-performance image processing solution that shifts workload from CPUs to FPGA technology, significantly enhancing data center efficiency. Using the Intel Agilex 7 FPGAs and SoCs F-Series, which are built on the 10 nm SuperFin process, the CIP can boost image processing speed by 5 to 20 times while reducing latency by the same measure. This enhancement is crucial for accommodating the explosive growth of image data in data centers due to smartphone proliferation and extensive use of cloud storage. The Agilex FPGA's advanced features include transceiver rates up to 58 Gbps, versatile DSP blocks supporting both fixed-point and floating-point operations, and high-performance cryptographic capabilities. These features facilitate substantial performance improvements in image transcoding, thumbnail generation, and image recognition tasks, reducing total cost of ownership by enabling data centers to maintain higher compute densities with lower operational costs. Moreover, the CIP's support for mainstream image processing software such as ImageMagick and OpenCV ensures seamless integration and deployment. The FPGA's capability for remote reconfiguration allows it to adapt swiftly to custom usage scenarios without server downtimes, enhancing maintenance and operational flexibility.
Dyumnin Semiconductors' RISCV SoC is a powerful, 64-bit quad-core server-class processor tailored for demanding applications, integrating a multifaceted array of subsystems. Key features include an AI/ML subsystem equipped with a tensor flow unit for optimized AI operations, and a robust automotive subsystem supporting CAN, CAN-FD, and SafeSPI interfaces.\n\nAdditionally, it includes a multimedia subsystem comprising HDMI, Display Port, MIPI, camera subsystems, Gfx accelerators, and digital audio, offering comprehensive multimedia processing capabilities. The memory subsystem connects to various prevalent memory protocols like DDR, MMC, ONFI, NorFlash, and SD/SDIO, ensuring vast compatibility.\n\nThe RISCV SoC's design is modular, allowing for customization to meet specific end-user applications, offering a flexible platform for creating SoC solutions with bespoke peripherals. It also doubles as a test chip available as an FPGA for evaluative purposes, making it ideal for efficient prototyping and development workflows.
Hanguang 800 AI Accelerator leads the front in artificial intelligence processing, delivering robust performance for complex machine learning tasks. T-Head's Hanguang 800 excels in accelerating inference, providing computational efficiency demanded by AI-centric applications. Designed for high-speed AI workloads, the Hanguang 800 integrates sophisticated neural network computing capabilities. Its architecture is optimized to handle large volumes of data processing, making it ideal for deep learning inference which requires high parallel computing power. The design underscores T-Head's strength in innovation, aligning efficient power consumption with high processing speeds, thereby making the Hanguang 800 a competitive choice for next-gen AI solutions across industries in need of cutting-edge processing efficiencies.
iModeler is Xpeedic's innovative solution for automated PDK model generation. This tool streamlines the process of creating Process Design Kits, which are foundational for semiconductor manufacturing processes. iModeler’s capabilities in automating PDK generation reduce time and resources required, providing a significant advancement over traditional manual methods. By utilizing sophisticated algorithms, iModeler enhances accuracy in developing intricate models that are essential for advanced semiconductor fabrication. The tool supports a broad range of semiconductor processes, ensuring cross-compatibility and robustness in model output. This level of precision supports engineers in achieving optimal results in both design and manufacturing stages. With iModeler, companies can significantly boost their development productivity, enabling quicker turnarounds in the semiconductor lifecycle. For organizations looking to maintain cutting-edge competitiveness, iModeler is a game-changer, providing the necessary infrastructure to support rapid advancements in chip manufacturing technologies.
The SCR7 is tailored for high-performance data-intensive applications, offering a 64-bit RISC-V processor with exceptional computational capabilities. Its architecture includes a dual-issue 12-stage out-of-order pipeline, advanced vector processing, and comprehensive memory management features. Ideal for AI, ML, and high-performance computing environments, it supports complex instructions and is fit for intensive processing needs across sectors such as networking, video processing, and enterprise computing.
aiSim 5 is aiMotive's state-of-the-art ISO26262 ASIL-D certified simulator designed to accelerate and optimize the validation process of Advanced Driver Assistance Systems (ADAS) and automated driving (AD) software. Its core components leverage AI-based rendering and highly optimized sensor simulation to establish a new standard in automotive simulation, delivering unmatched realism and adaptiveness. This cutting-edge tool allows for extensive multisensor environments, supporting over 20 cameras, 10 radars, and numerous lidars, thereby offering an authentic, comprehensive testing platform for autonomous systems. A testament to aiSim 5's capabilities is its robust 3D asset library and versatile content pipeline. These facilitate the creation and deployment of complex, high-fidelity environments crucial for thorough ADAS and AD software validation. Additionally, the simulator provides a cloud-native UI and open SDK, giving developers ample flexibility to create custom test scenarios and seamlessly integrate them into existing toolchains. Its proprietary aiSim AIR engine plays a pivotal role, delivering high-quality virtual sensor data streams while maintaining efficient resource use. The engine supports distributed rendering and balances workload by allowing asynchronous data transfer, further elevating the simulator's performance and ensuring compliance with stringent automotive standards.
The KL520 AI SoC by Kneron marked a significant breakthrough in edge AI technology, offering a well-rounded solution with notable power efficiency and performance. This chip can function as a host or as a supplementary co-processor to enable advanced AI features in diverse smart devices. It is highly compatible with a range of 3D sensor technologies and is perfectly suited for smart home innovations, facilitating long battery life and enhanced user control without reliance on external cloud services.
The NeuroSense AI Chip, an ultra-low power neuromorphic frontend, is engineered for wearables to address the challenges of power efficiency and data accuracy in health monitoring applications. This tiny AI chip is designed to process data directly at the sensor level, which includes tasks like heart rate measurement and human activity recognition. By performing computations locally, NeuroSense minimizes the need for cloud connections, thereby ensuring privacy and prolonging battery life. The chip excels in accuracy, significantly outperforming conventional algorithm-based solutions by offering three times better heart rate accuracy. This is achieved through its ability to reduce power consumption to below 100µW, allowing users to experience extended device operation without frequent recharging. The NeuroSense supports a simple configuration setup, making it suitable for integration into a variety of wearable devices such as fitness trackers, smartwatches, and health monitors. Its capabilities extend to advanced features like activity matrices, enabling devices to learn new human activities and classify tasks according to intensity levels. Additional functions include monitoring parameters like oxygen saturation and arrhythmia, enhancing the utility of wearable devices in providing comprehensive health insights. The chip's integration leads to reduced manufacturing costs, a smaller IC footprint, and a rapid time-to-market for new products.
The Semidynamics Vector Unit is a powerful processing element designed for applications requiring complex parallel computations such as those found in machine learning and AI workloads. Its remarkable configurability allows it to be adapted for different data types ranging from 8-bit integers to 64-bit floating-point numbers, supporting standards up to RVV 1.0. The unit can perform a wide array of operations due to its included arithmetic units for addition, subtraction, and complex tasks like multiplication and logic operations. PHased to deliver exceptional performance, the Vector Unit leverages a cross-vector-core network that ensures high bandwidth connectivity among its vector cores, capable of scaling up to 32 cores. This feature helps maximize operational efficiency, allowing tasks to be distributed across multiple cores for optimized performance and power efficiency. Its design caters to extensive data path configurations, allowing users to choose from DLEN options ranging from 128 bits to an impressive 2048 bits in width. Moreover, this Vector Unit supports flexible hardware setups by aligning vector register lengths (VLEN) with the data path requirements, offering up to an 8X ratio between VLEN and DLEN. This capability enhances its adaptability, allowing it to absorb memory latencies effectively, making it particularly suitable for AI inferencing tasks that require rapid iteration and heavy computational loads. Its integration with existing Semidynamics technologies like the Tensor Unit ensures a seamless performance boost across hardware configurations.
The EMSA5-FS, part of CAST's burgeoning suite of RISC-V solutions, stands out with its emphasis on functional safety, geared towards applications requiring rigorous reliability. This 32-bit embedded processor core adheres to rigorous safety standards, making it optimal for sectors such as automotive and industrial automation where system failures could have critical consequences. Functional safety is paramount with the EMSA5-FS, ensuring operations are secured against potential faults. This core includes specific features that enhance its reliability and help mitigate the risks inherent in functional safety applications. It supports both safe execution and error handling, conforming to high compliance levels needed for certification in critical environments. By offering a robust solution designed for tasks that cannot afford failure, the EMSA5-FS extends CAST's footprint into the rapidly expanding field of safety-critical systems. It provides manufacturers with the assurance of performance stability under the most demanding operational conditions, thereby supporting the creation of cutting-edge, compliant technologies.
The hypr_risc Radar DSP Accelerator from NOVELIC is a highly configurable digital signal processor connected to a custom RISC-V-based core. Engineered for speed, it is optimized for high-speed advanced driver-assistance systems (ADAS) applications where fast processing is critical. It handles an array of signal processing tasks, from basic object range assessment to complex imaging, and can be tailored to match any frontend.
The J1 core cell is a remarkably small and efficient audio decoder that manages Dolby Digital, AC-3, and MPEG audio decompression. With a design that occupies only 1.0 sqmm of silicon area using 0.18u CMOS technology, it delivers a robust solution for decoding 5.1 channel dolby bitstreams and supports data rates up to 640kb/s. The J1 produces high-quality stereo outputs, both normal and Pro-Logic compatible, from Dolby Digital and MPEG-encoded audio, ideal for set-top boxes and DVD applications.
Kneron's KL530 introduces a modern heterogeneous AI chip design featuring a cutting-edge NPU architecture with support for INT4 precision. This chip stands out with its high computational efficiency and minimized power usage, making it ideal for a variety of AIoT and other applications. The KL530 utilizes an ARM Cortex M4 CPU, bringing forth powerful image processing and multimedia compression capabilities, while maintaining a low power footprint, thus fitting well with energy-conscious devices.
So-Logic's Machine Learning Cores are pioneering in bringing hardware acceleration to the forefront of learning systems. These cores empower devices to implement sophisticated machine learning algorithms directly in hardware, markedly enhancing performance and efficiency. The distinctive advantage of these cores lies in their ability to handle vast amounts of data in real-time, making them suitable for advanced computing applications requiring rapid and autonomous decision-making. By integrating machine learning functionalities into FPGAs, So-Logic facilitates powerful, flexible, and adaptable systems that can keep pace with the growing demands of AI-driven applications. These cores are crafted with a focus on optimizing algorithm execution, thereby reducing latency and increasing throughput in machine learning tasks. Delivering each core with comprehensive support materials, So-Logic ensures the ease of implementation. This includes detailed documentation, example applications, and a suite of validation tests to integrate machine learning capabilities seamlessly within broader system architectures. The cores not only enhance computational efficiency but also drive innovation in areas such as adaptive control systems, predictive maintenance, and intelligent IoT devices.
The Metis AIPU M.2 Accelerator Module by Axelera AI is a cutting-edge AI acceleration tool designed for edge applications. Its compact form factor, combined with powerful AI processing technology, enables real-time data processing and analysis. Equipped with 512MB of dedicated LPDDR4x memory, this module is capable of handling multiple data streams simultaneously. Its breakthrough digital in-memory compute architecture facilitates remarkable energy efficiency, consuming far less power than traditional GPUs while maintaining top-notch performance standards. This module is ideal for applications requiring high-speed computation, such as computer vision tasks involving multi-channel video analytics and quality inspections, thereby enhancing operational efficiency and reducing latency in decision-making processes. Whether deployed in retail, security, or industrial settings, the Metis AIPU M.2 Accelerator Module provides users with significant performance gains at a lower cost, facilitating seamless integration into existing systems. With a practical design for next-generation form factor M.2 sockets, this accelerator module opens the way for innovative AI-enabled solutions in diverse contexts, promising scalability and adaptability to future technological advancements.
The RAIV is a flexible and high-performing General Purpose GPU (GPGPU), fundamental for industries experiencing rapid transformation due to the fourth industrial revolution—autonomous vehicles, IoT, and VR/AR sectors. Built with a SIMT (Single Instruction Multiple Threads) architecture, the RAIV enhances AI workloads with high-speed processing capabilities while maintaining a low-cost construct. This semiconductor IP supports diverse machine learning and neural network applications, optimizing high-speed calculations across multiple threads. Its high scalability allows tailored configurations in core units, effectively balancing performance with power efficiency dependent on application needs. The RAIV is equipped to handle 3D graphics processing and AI integration for edge computing devices, reinforcing its place in advanced technological development. Additionally, the RAIV's support for OpenCL offers compatibility across various heterogeneous computing platforms, facilitating versatile system configurations. Its optimal performance in AI tasks is further extended for use in metaverse applications, presenting a comprehensive solution that unifies graphics acceleration with AI-enhanced computational operations.
The SAKURA-II AI Accelerator by EdgeCortix is a high-efficiency device designed for demanding edge applications. This cutting-edge accelerator is tailored for fast, real-time, single-batch AI inferencing with low energy consumption and minimal footprint. The hardware enables users to process complex generative AI models such as Llama 2 and Stable Diffusion. By supporting multi-billion parameter models, it excels in areas like Vision, Language, and Audio AI applications. Its edge efficiency is further augmented by its high utilization of AI compute, surpassing many competing solutions. This ensures superior functionality across vast applications.
Designed to meet the future needs of AI technology, the SiFive Intelligence family introduces AI dataflow processors with scalable vector compute capabilities. The X280 model emphasizes high-performance scalar and vector computing suitable for AI workloads, data flow management, and complex processing tasks. By integrating SiFive Matrix Engine technology, the X280 enhances compute capabilities with a 512-bit vector length ensuring efficient computation flows. The platform is scalable, supporting integrations from entry-level to high-performance needs, whilst maintaining a focus on power efficiency and footprint reduction.
The Tianqiao-90 is a robust RISC-V CPU core designed for high-end applications, including data centers and advanced computation scenarios. It incorporates leading-edge design features such as superscalar and deep out-of-order execution, making it suitable for performance-intensive environments. The architecture supports standard RISC-V RV64GCBH extensions and has undergone substantial optimizations for performance and frequency, achieving SPECint2006 scores of 9.4 per GHz. Developed with multi-core scalability in mind, Tianqiao-90 allows configurations ranging from single-core to quad-core, enhancing its versatility for various applications. This CPU core simplifies SoC development processes and can be widely utilized across sectors like machine learning, mobile devices, and network communications. Its adaptability for memory coherence makes it ideal for multi-core systems. With a fabrication process at TSMC's 12nm node, the Tianqiao-90 offers exemplary efficiency in power consumption and area effectiveness, crucial for enterprise-level computation. Its high-frequency operation, reaching up to 2GHz, provides the necessary power for demanding computing tasks, ensuring swift and reliable performance in all deployed scenarios.