All IPs > Processor > Vision Processor
Vision processors are a specialized subset of semiconductor IPs designed to efficiently handle and process visual data. These processors are pivotal in applications that require intensive image analysis and computer vision capabilities, such as artificial intelligence, augmented reality, virtual reality, and autonomous systems. The primary purpose of vision processor IPs is to accelerate the performance of vision processing tasks while minimizing power consumption and maximizing throughput.
In the world of semiconductor IP, vision processors stand out due to their ability to integrate advanced functionalities such as object recognition, image stabilization, and real-time analytics. These processors often leverage parallel processing, machine learning algorithms, and specialized hardware accelerators to perform complex visual computations efficiently. As a result, products ranging from high-end smartphones to advanced driver-assistance systems (ADAS) and industrial robots benefit from improved visual understanding and processing capabilities.
The semiconductor IPs for vision processors can be found in a wide array of products. In consumer electronics, they enhance the capabilities of cameras, enabling features like face and gesture recognition. In the automotive industry, vision processors are crucial for delivering real-time data processing needed for safety systems and autonomous navigation. Additionally, in sectors such as healthcare and manufacturing, vision processor IPs facilitate advanced imaging and diagnostic tools, improving both precision and efficiency.
As technology advances, the demand for vision processor IPs continues to grow. Developers and designers seek IPs that offer scalable architectures and can be customized to meet specific application requirements. By providing enhanced performance and reducing development time, vision processor semiconductor IPs are integral to pushing the boundaries of what's possible with visual data processing and expanding the capabilities of next-generation products.
BrainChip's Akida Neural Processor IP is a groundbreaking development in neuromorphic processing, designed to mimic the human brain in interpreting sensory inputs. By implementing an event-based architecture, it processes only the critical data at the point of acquisition, achieving unparalleled performance with significantly reduced power consumption. This architecture enables on-chip learning, reducing dependency on cloud processing, thus enhancing privacy and security.\n\nThe Akida Neural Processor IP supports incremental learning and high-speed inference across a vast range of applications, making it highly versatile. It is structured to handle data sparsity effectively, which cuts down on operations substantially, leading to considerable improvements in efficiency and responsiveness. The processor's scalability and compact design allow for wide deployment, from minimal-node setups for ultra-low power operations to more extensive configurations for handling complex tasks.\n\nImportantly, the Akida processor uses a fully customizable AI neural processor that leverages event-based processing and an on-chip mesh network for seamless communication. The technology also features support for hybrid quantized weights and provides robust tools for integration, including fully synthesizable RTL IP packages, hardware-based event processing, and on-chip learning capabilities.
The Akida 2nd Generation is an evolution of BrainChip's innovative neural processor technology. It builds upon its predecessor's strengths by delivering even greater efficiency and a broader range of applications. The processor maintains an event-based architecture that optimizes performance and power consumption, providing rapid response times suitable for edge AI applications that prioritize speed and privacy.\n\nThis next-generation processor enhances accuracy with support for 8-bit quantization, which allows for finer grained processing capabilities and more robust AI model implementations. Furthermore, it offers extensive scalability, supporting configurations from a few nodes for low-power needs to many nodes for handling more complex cognitive tasks. As with the previous version, its architecture is inherently cloud-independent, enabling inference and learning directly on the device.\n\nAkida 2nd Generation continues to push the boundaries of AI processing at the edge by offering enhanced processing capabilities, making it ideal for applications demanding high accuracy and efficiency, such as automotive safety systems, consumer electronics, and industrial monitoring.
The KL730 is a sophisticated AI System on Chip (SoC) that embodies Kneron's third-generation reconfigurable NPU architecture. This SoC delivers a substantial 8 TOPS of computing power, designed to efficiently handle CNN network architectures and transformer applications. Its innovative NPU architecture significantly optimizes DDR bandwidth, providing powerful video processing capabilities, including supporting 4K resolution at 60 FPS. Furthermore, the KL730 demonstrates formidable performance in noise reduction and low-light imaging, positioning it as a versatile solution for intelligent security, video conferencing, and autonomous applications.
The Metis AIPU PCIe AI Accelerator Card offers exceptional performance for AI workloads demanding significant computational capacity. It is powered by a single Metis AIPU and delivers up to 214 TOPS, catering to high-demand applications such as computer vision and real-time image processing. This PCIe card is integrated with the Voyager SDK, providing developers with a powerful yet user-friendly software environment for deploying complex AI applications seamlessly. Designed for efficiency, this accelerator card stands out by providing cutting-edge performance without the excessive power requirements typical of data center equipment. It achieves remarkable speed and accuracy, making it an ideal solution for tasks requiring fast data processing and inference speeds. The PCIe card supports a wide range of AI application scenarios, from enhancing existing infrastructure capabilities to integrating with new, dynamic systems. Its utility in various industrial settings is bolstered by its compatibility with the suite of state-of-the-art neural networks provided in the Axelera AI ecosystem.
The Yitian 710 Processor is an advanced Arm-based server chip developed by T-Head, designed to meet the extensive demands of modern data centers and enterprise applications. This processor boasts 128 high-performance Armv9 CPU cores, each coupled with robust caches, ensuring superior processing speeds and efficiency. With a 2.5D packaging technology, the Yitian 710 integrates multiple dies into a single unit, facilitating enhanced computational capability and energy efficiency. One of the key features of the Yitian 710 is its memory subsystem, which supports up to 8 channels of DDR5 memory, achieving a peak bandwidth of 281 GB/s. This configuration guarantees rapid data access and processing, crucial for high-throughput computing environments. Additionally, the processor is equipped with 96 PCIe 5.0 lanes, offering a dual-direction bandwidth of 768 GB/s, enabling seamless connectivity with peripheral devices and boosting system performance overall. The Yitian 710 Processor is meticulously crafted for applications in cloud services, big data analytics, and AI inference, providing organizations with a robust platform for their computing needs. By combining high core count, extensive memory support, and advanced I/O capabilities, the Yitian 710 stands as a cornerstone for deploying powerful, scalable, and energy-efficient data processing solutions.
The Metis AIPU M.2 Accelerator Module is designed for edge AI applications that demand high-performance inference capabilities. This module integrates a single Metis AI Processing Unit (AIPU), providing an excellent solution for AI acceleration within constrained devices. Its capability to handle high-speed data processing with limited power consumption makes it an optimal choice for applications requiring efficiency and precision. With 1GB of dedicated DRAM memory, it seamlessly supports a wide array of AI pipelines, ensuring rapid integration and deployment. The design of the Metis AIPU M.2 module is centered around maximizing performance without excessive energy consumption, making it suitable for diverse applications such as real-time video analytics and multi-camera processing. Its compact form factor eases incorporation into various devices, delivering robust performance for AI tasks without the heat or power trade-offs typically associated with such systems. Engineered to problem-solve current AI demands efficiently, the M.2 module comes supported by the Voyager SDK, which simplifies the integration process. This comprehensive software suite empowers developers to build and optimize AI models directly on the Metis platform, facilitating a significant reduction in time-to-market for innovative solutions.
The AI Camera Module by Altek Corporation exemplifies cutting-edge image capture technology, integrating both hardware and software to deliver high-quality, intelligent imaging solutions. This module is built on robust AI frameworks allowing it to adapt and optimize image processing based on specific application needs. It finds use in areas where high-resolution and real-time processing are essential, such as security systems and automotive industries.<br/><br/>Equipped with versatile imaging sensors, the AI Camera Module ensures excellent picture quality even in challenging lighting conditions, thanks to its AI-driven image enhancement algorithms. It supports edge computing, which reduces latency and enhances the speed of image analysis, thus providing timely insights and data processing right on the device itself.<br/><br/>This camera module stands out for its interoperability with IoT devices, paving the way for a more interconnected and intelligent ecosystem. Its advanced features such as facial detection, motion tracking, and object recognition empower users across various domains, from consumer electronics to industrial solutions, making it an indispensable tool for modern digital infrastructures.
The AX45MP is engineered as a high-performance processor that supports multicore architecture and advanced data processing capabilities, particularly suitable for applications requiring extensive computational efficiency. Powered by the AndesCore processor line, it capitalizes on a multicore symmetric multiprocessing framework, integrating up to eight cores with robust L2 cache management. The AX45MP incorporates advanced features such as vector processing capabilities and support for MemBoost technology to maximize data throughput. It caters to high-demand applications including machine learning, digital signal processing, and complex algorithmic computations, ensuring data coherence and efficient power usage.
The KL520 was Kneron's first foray into AI SoCs, characterized by its small size and energy efficiency. This chip integrates a dual ARM Cortex M4 CPU architecture, which can function both as a host processor and as a supportive AI co-processor for diverse edge devices. Ideal for smart devices such as door locks and cameras, it is compatible with various 3D sensor technologies, offering a balance of compact design and high performance. As a result, this SoC has been adopted by multiple products in the smart home and security sectors.
The Jotunn8 represents a leap in AI inference technology, delivering unmatched efficiency for modern data centers. This chip is engineered to manage AI model deployments with lightning-fast execution, at minimal cost and high scalability. It ensures optimal performance by balancing high throughput and low latency, while being extremely power-efficient, which significantly lowers operational costs and supports sustainable infrastructures. The Jotunn8 is designed to unlock the full capacity of AI investments by providing a high-performance platform that enhances the delivery and impact of AI models across applications. It is particularly suitable for real-time applications such as chatbots, fraud detection, and search engines, where ultra-low latency and very high throughput are critical. Power efficiency is a major emphasis of the Jotunn8, optimizing performance per watt to control energy as a substantial operational expense. Its architecture allows for flexible memory allocation ensuring seamless adaptability across varied applications, providing a robust foundation for scalable AI operations. This solution is aimed at enhancing business competitiveness by supporting large-scale model deployment and infrastructure optimization.
The Chimera GPNPU from Quadric is designed as a general-purpose neural processing unit intended to meet a broad range of demands in machine learning inference applications. It is engineered to perform both matrix and vector operations along with scalar code within a single execution pipeline, which offers significant flexibility and efficiency across various computational tasks. This product achieves up to 864 Tera Operations per Second (TOPs), making it suitable for intensive applications including automotive safety systems. Notably, the GPNPU simplifies system-on-chip (SoC) hardware integration by consolidating hardware functions into one processor core. This unification reduces complexity in system design tasks, enhances memory usage profiling, and optimizes power consumption when compared to systems involving multiple heterogeneous cores such as NPUs and DSPs. Additionally, its single-core setup enables developers to efficiently compile and execute diverse workloads, improving performance tuning and reducing development time. The architecture of the Chimera GPNPU supports state-of-the-art models with its Forward Programming Interface that facilitates easy adaptation to changes, allowing support for new network models and neural network operators. It’s an ideal solution for products requiring a mix of traditional digital signal processing and AI inference like radar and lidar signal processing, showcasing a rare blend of programming simplicity and long-term flexibility. This capability future-proofs devices, expanding their lifespan significantly in a rapidly evolving tech landscape.
The Polar ID Biometric Security System by Metalenz revolutionizes smartphone biometric security with its advanced imaging capabilities that capture the full polarization state of light. This system detects unique facial polarization signatures, enabling high-precision face authentication that even sophisticated 3D masks cannot deceive. Unlike traditional systems requiring multiple optical modules, Polar ID achieves secure recognition with a single image, ideal for secure digital payments and more. Operating efficiently across various lighting conditions, from bright daylight to complete darkness, Polar ID ensures robust security without compromising user convenience. By leveraging meta-optic technology, it offers a compact, cost-effective alternative to structured light solutions, suitable for widespread deployment across millions of mobile devices.
The KL630 chip stands out with its pioneering NPU architecture, making it the industry's first to support Int4 precision alongside transformer networks. This unique capability enables it to achieve exceptional computational efficiency and low energy consumption, suitable for a wide variety of applications. The chip incorporates an ARM Cortex A5 CPU, providing robust support for all major AI frameworks and delivering superior ISP capabilities for handling low light conditions and HDR applications, making it ideal for security, automotive, and smart city uses.
xcore.ai is a powerful platform tailored for the intelligent IoT market, offering unmatched flexibility and performance. It boasts a unique multi-threaded micro-architecture that provides low-latency and deterministic performance, perfect for smart applications. Each xcore.ai contains 16 logical cores distributed across two multi-threaded processor tiles, each equipped with 512kB of SRAM and capable of both integer and floating-point operations. The integrated interprocessor communication allows high-speed data exchange, ensuring ultimate scalability across multiple xcore.ai SoCs within a unified development environment.
The Dynamic Neural Accelerator II (DNA-II) is a highly efficient and versatile IP specifically engineered for optimizing AI workloads at the edge. Its unique architecture allows runtime reconfiguration of interconnects among computing units, which facilitates improved parallel processing and efficiency. DNA-II supports a broad array of networks, including convolutional and transformer networks, making it an ideal choice for numerous edge applications. Its design emphasizes low power consumption while maintaining high computational performance. By utilizing a dynamic data path architecture, DNA-II sets a new benchmark for IP cores aimed at enhancing AI processing capabilities.
The KL530 is built with an advanced heterogeneous AI chip architecture, designed to enhance computing efficiency while reducing power usage. Notably, it is recognized as the first in the market to support INT4 precision and transformers for commercial applications. The chip, featuring a low-power ARM Cortex M4 CPU, delivers impressive performance with 1 TOPS@INT 4 computing power, providing up to 70% higher processing efficiency compared to INT8 architectures. Its integrated smart ISP optimizes image quality, supporting AI models like CNN and RNN, suitable for IoT and AIoT ecosystems.
The Hanguang 800 AI Accelerator by T-Head is an advanced semiconductor technology designed to accelerate AI computations and machine learning tasks. This accelerator is specifically optimized for high-performance inference, offering substantial improvements in processing times for deep learning applications. Its architecture is developed to leverage parallel computing capabilities, making it highly suitable for tasks that require fast and efficient data handling. This AI accelerator supports a broad spectrum of machine learning frameworks, ensuring compatibility with various AI algorithms. It is equipped with specialized processing units and a high-throughput memory interface, allowing it to handle large datasets with minimal latency. The Hanguang 800 is particularly effective in environments where rapid inferencing and real-time data processing are essential, such as in smart cities and autonomous driving. With its robust design and multi-faceted processing abilities, the Hanguang 800 Accelerator empowers industries to enhance their AI and machine learning deployments. Its capability to deliver swift computation and inference results ensures it is a valuable asset for companies looking to stay at the forefront of technological advancement in AI applications.
SAKURA-II is an advanced AI accelerator recognized for its efficiency and adaptability. It is specifically designed for edge applications that require rapid, real-time AI inference with minimal delay. Capable of processing expansive generative AI models such as Llama 2 and Stable Diffusion within an 8W power envelope, this accelerator supports a wide range of applications from vision to language processing. Its enhanced memory bandwidth and substantial DRAM capacity ensure its suitability for handling complex AI workloads, including large-scale language and vision models. The SAKURA-II platform also features robust power management, allowing it to achieve high efficiency during operations.
aiWare represents aiMotive's advanced hardware intellectual property core for automotive neural network acceleration, pushing boundaries in efficiency and scalability. This neural processing unit (NPU) is tailored to meet the rigorous demands of automotive AI inference, providing robust support for various AI workloads, including CNNs, LSTMs, and RNNs. By achieving up to 256 Effective TOPS and remarkable scalability, aiWare caters to a wide array of applications, from edge processors in sensors to centralized high-performance modules.\n\nThe design of aiWare is particularly focused on enhancing efficiency in neural network operations, achieving up to 98% efficiency across diverse automotive applications. It features an innovative dataflow architecture, ensuring minimal external memory bandwidth usage while maximizing in-chip data processing. This reduces power consumption and enhances performance, making it highly adaptable for deployment in resource-critical environments.\n\nAdditionally, aiWare is embedded with comprehensive tools like the aiWare Studio SDK, which streamlines the neural network optimization and iteration process without requiring extensive NPU code adjustments. This ensures that aiWare can deliver optimal performance while minimizing development timelines by allowing for early performance estimations even before target hardware testing. Its integration into ASIL-B or higher certified solutions underscores aiWare's capability to power the most demanding safety applications in the automotive domain.
The eSi-3264 epitomizes the pinnacle of the eSi-RISC portfolio, presenting a 32/64-bit processor furnished with SIMD extensions catering to high-performance requirements. Designed for applications demanding digital signal processing functionality, this processor capitalizes on minimal silicon usage while ensuring exceedingly low power consumption. Incorporating an extensive pipeline capable of dual and quad multiply-accumulate operations, the eSi-3264 significantly benefits applications in audio processing, sensor control, and touch interfacing. Its built-in IEEE-754 single and double-precision floating point operations promote comprehensive data processing capabilities, extending versatility across computationally intensive domains. The processor accommodates configurable caching attributes and a memory management unit to bolster performance amidst off-chip memory access. Its robust instruction repertoire, optional custom operations, and user-privilege modes ensure full control in secure execution environments, supporting diverse operational requirements with unmatched resource efficiency.
The Spiking Neural Processor T1 is a microcontroller tailored for ultra-low-power applications demanding high-performance pattern recognition at the sensor edge. It features an advanced neuromorphic architecture that leverages spiking neural network engines combined with RISC-V core capabilities. This architecture allows for sub-milliwatt power dissipation and sub-millisecond latency, enabling the processor to conduct real-time analysis and identification of embedded patterns in sensor data while operating in always-on scenarios. Additionally, the T1 provides diverse interfaces, making it adaptable for use with various sensor types.
The Intelligence X280 is engineered to provide extensive capabilities for artificial intelligence and machine learning applications, emphasizing a software-first design approach. This high-performance processor supports vector and matrix computations, making it adept at handling the demanding workloads typical in AI-driven environments. With an extensive ALU and integrated VFPU capabilities, the X280 delivers superior data processing power. Capable of supporting complex AI tasks, the X280 processor leverages SiFive's advanced vector architecture to allow for high-speed data manipulation and precision. The core supports extensive vector lengths and offers compatibility with various machine learning frameworks, facilitating seamless deployment in both embedded and edge AI applications. The Intelligence family, represented by the X280, offers solutions that are not only scalable but are customizable to particular workload specifications. With high-bandwidth interfaces for connecting custom engines, this processor is built to evolve alongside AI's progressive requirements, ensuring relevance in rapidly changing technology landscapes.
The KL720 is engineered for high efficiency, achieving up to 0.9 TOPS per Watt, setting it apart in the edge AI marketplace. Designed for real-world scenarios where power efficiency is paramount, this chip supports high-end IP cameras, smart TVs, and AI-enabled devices like glasses and headsets. Its ARM Cortex M4 CPU facilitates the processing of complex tasks like 4K image handling, full HD video, and 3D sensing, making it versatile for applications that include gaming and AI-assisted interactions.
The CTAccel Image Processor on Intel Agilex FPGA is designed to handle high-performance image processing by capitalizing on the robust capabilities of Intel's Agilex FPGAs. These FPGAs, leveraging the 10 nm SuperFin process technology, are ideal for applications demanding high performance, power efficiency, and compact sizes. Featuring advanced DSP blocks and high-speed transceivers, this IP thrives in accelerating image processing tasks that are typically computational-intensive when executed on CPUs. One of the main advantages is its ability to significantly enhance image processing throughput, achieving up to 20 times the speed while maintaining reduced latency. This performance prowess is coupled with low power consumption, leading to decreased operational and maintenance costs due to fewer required server instances. Additionally, the solution is fully compatible with mainstream image processing software, facilitating seamless integration and leveraging existing software investments. The adaptability of the FPGA allows for remote reconfiguration, ensuring that the IP can be tailored to specific image processing scenarios without necessitating a server reboot. This ease of maintenance, combined with a substantial boost in compute density, underscores the IP's suitability for high-demand image processing environments, such as those encountered in data centers and cloud computing platforms.
The PACE Photonic Arithmetic Computing Engine from Lightelligence represents a paradigm shift in computing technologies. By utilizing photonic processes, this product significantly boosts computing speeds while maintaining energy efficiency. PACE is designed to leverage the inherent capabilities of photonics to perform high-speed arithmetic calculations, which are essential for complex data processing tasks. It's an ideal solution for industries demanding rapid and intensive computational power without the typical energy overhead.<br> <br> This advanced engine is central to the development of next-generation computing environments, where performance metrics exceed traditional expectations. By converting light signals into computing potential, PACE ensures that intensive processes such as AI computations, data analyses, and real-time processing are handled more efficiently. This product is tailored for enterprises seeking to minimize latency and enhance throughput across various applications.<br> <br> PACE not only meets the requirements of current computational demands but also sets the stage for future innovations in the field. It's a promising tool for developers and researchers aiming to explore the unexplored realms of digital capabilities, fostering an era of optical computing that's faster and more efficient than ever before. This makes PACE an indispensable component in both current and upcoming technological advancements.
The RAIV General Purpose GPU (GPGPU) epitomizes versatility and cutting-edge technology in the realm of data processing and graphics acceleration. It serves as a crucial technology enabler for various prominent sectors that are central to the fourth industrial revolution, such as autonomous driving, IoT, virtual reality/augmented reality (VR/AR), and sophisticated data centers. By leveraging the RAIV GPGPU, industries are able to process vast amounts of data more efficiently, which is paramount for their growth and competitive edge. Characterized by its advanced architectural design, the RAIV GPU excels in managing substantial computational loads, which is essential for AI-driven processes and complex data analytics. Its adaptability makes it suitable for a wide array of applications, from enhancing automotive AI systems to empowering VR environments with seamless real-time interaction. Through optimized data handling and acceleration, the RAIV GPGPU assists in realizing smoother and more responsive application workflows. The strategic design of the RAIV GPGPU focuses on enabling integrative solutions that enhance performance without compromising on power efficiency. Its functionality is built to meet the high demands of today’s tech ecosystems, fostering advancements in computational efficiency and intelligent processing capabilities. As such, the RAIV stands out not only as a tool for improved graphical experiences but also as a significant component in driving innovation within tech-centric industries worldwide. Its pioneering architecture thus supports a multitude of applications, ensuring it remains a versatile and indispensable asset in diverse technological landscapes.
The 2D FFT core is engineered to deliver fast processing for two-dimensional FFT computations, essential in image and video processing applications. By utilizing both internal and external memory effectively, this core is capable of handling large data sets typical in medical imaging or aerial surveillance systems. This core leverages Dillon Engineering’s ParaCore Architect utility to maximize flexibility and efficiency. It takes advantage of a two-engine design, where data can flow between stages without interruption, ensuring high throughput and minimal memory delays. Such a robust setup is vital for applications where swift processing of extensive data grids is crucial. The architecture is structured to provide consistent, high-quality transform computations that are essential in applications where accuracy and speed are non-negotiable. The 2D FFT core, with its advanced design parameters, supports the varied demands of modern imaging technology, providing a reliable tool for developers and engineers working within these sectors.
The CTAccel Image Processor for Xilinx's Alveo U200 is a FPGA-based accelerator aimed at enhancing image processing workloads in server environments. Utilizing the powerful capabilities of the Alveo U200 FPGA, this processor dramatically boosts throughput and reduces processing latency for data centers. The accelerator can vastly increase image processing speed, up to 4 to 6 times that of traditional CPUs, and decrease latency likewise, ensuring that compute density in a server setting is significantly boosted. This performance uplift enables data centers to lower maintenance and operational costs due to reduced hardware requirements. Furthermore, this IP maintains full compatibility with popular image processing software like OpenCV and ImageMagick, ensuring smooth adaptation for existing workflows. The advanced FPGA partial reconfiguration technology allows for dynamic updates and adjustments, increasing the IP's pragmatism for a wide array of image-related applications and improving overall performance without the need for server reboots.
The MIPITM V-NLM-01 is a specialized non-local mean image noise reduction product designed to enhance image quality through sophisticated noise reduction techniques. This hardware core features a parameterized search-window size and adjustable bits per pixel, ensuring a high degree of customization and efficiency. Supporting HDMI with resolutions up to 2048×1080 at 30 to 60 fps, it is ideally suited for applications requiring image enhancement and processing.
The NeuroSense is a compact AI chip designed specifically for wearable devices, featuring neuromorphic analog signal processing technology. Its main focus lies in resolving common challenges faced by wearable tech, such as high power consumption, and limited battery life. By enabling highly accurate heart rate monitoring and activity recognition, this chip facilitates better fitness tracking without excessively draining battery resources. The NeuroSense's capability of operating independently from cloud connections addresses significant privacy concerns and data latency issues. It excels in delivering enhanced accuracy in heart rate measurements by utilizing a simple photoplethysmogram (PPG) configuration, which involves minimalistic hardware components like two LEDs and one photodiode. Through this setup, it achieves precision in bio-signal extraction far beyond conventional algorithmic methods, particularly when the wearer is in motion. Furthermore, the NeuroSense empowers wearables with advanced features like learning and recognizing user-specific activity patterns. With ultra-low power consumption and a compact size, the NeuroSense enables manufacturers to preserve space within constrained wearable designs while simultaneously enhancing battery life—solving a key concern in the realm of constantly operating smart devices.
The SiFive Performance family is an embodiment of high-efficiency computing, tailored to deliver maximum throughput across various applications. Designed with a 64-bit out-of-order architecture, these processors are equipped with up to 256-bit vector support, making them proficient in handling complex data and multimedia processing tasks critical for data centers and AI applications. The Performance cores range from 3-wide to 6-wide out-of-order models, capable of integrating up to two vector engines dedicated to AI workload optimizations. This setup provides an excellent balance of energy efficiency and computing power, supporting diverse applications ranging from web servers and network storage to consumer electronics requiring smart capabilities. Focused on maximizing performance while minimizing power usage, the Performance family allows developers to customize and optimize processing capabilities to match specific use-cases. This adaptability, combined with high efficiency, renders the Performance line a fitting choice for modern computational tasks that demand both high throughput and energy conservation.
The **Ceva-SensPro DSP family** unites scalar processing units and vector processing units under an 8-way VLIW architecture. The family incorporates advanced control features such as a branch target buffer and a loop buffer to speed up execution and reduce power. There are six family members, each with a different array of MACs, targeted at different application areas and performance points. These range from the Ceva-SP100, providing 128 8-bit integer or 32 16-bit integer MACs at 0.2 TOPS performance for compact applications such as vision processing in wearables and mobile devices; to the Ceva-SP1000, with 1024 8-bit or 256 16-bit MACs reaching 2 TOPS for demanding applications such as automotive, robotics, and surveillance. Two of the family members, the Ceva-SPF2 and Ceva-SPF4, employ 32 or 64 32-bit floating-point MACs, respectively, for applications in electric-vehicle power-train control and battery management. These two members are supported by libraries for Eigen Linear Algebra, MATLAB vector operations, and the TVM graph compiler. Highly configurable, the vector processing units in all family members can add domain-specific instructions for such areas as vision processing, Radar, or simultaneous localization and mapping (SLAM) for robotics. Integer family members can also add optional floating-point capabilities. All family members have independent instruction and data memory subsystems and a Ceva-Connect queue manager for AXI-attached accelerators or coprocessors. The Ceva-SensPro2 family is programmable in C/C++ as well as in Halide and Open MP, and supported by an Eclipse-based development environment, extensive libraries spanning a wide range of applications, and the Ceva-NeuPro Studio AI development environment. [**Learn more about Ceva-SensPro2 solution>**](https://www.ceva-ip.com/product/ceva-senspro2/?utm_source=silicon_hub&utm_medium=ip_listing&utm_campaign=ceva_senspro2_page)
CTAccel's Image Processor for AWS offers a powerful image processing acceleration solution as part of Amazon's cloud infrastructure. This FPGA-based processor is available as an Amazon Machine Image (AMI) and enables customers to significantly enhance their image processing capabilities within the cloud environment. The AWS-based accelerator provides a remarkable tenfold increase in image processing throughput and similar reductions in computational latency, positively impacting Total Cost of Ownership (TCO) by reducing infrastructure needs and improving operational efficiency. These enhancements are crucial for applications requiring intensive image analysis and processing. Moreover, the processor supports a variety of image enhancement functions such as JPEG thumbnail generation and color adjustments, making it suitable for diverse cloud-based processing scenarios. Its integration within the AWS ecosystem ensures that users can easily deploy and manage these advanced processing capabilities across various imaging workflows with minimal disruption.
aiSim 5 is a premier simulation tool tailored for ADAS (Advanced Driver Assistance Systems) and automated driving validations. As the world's first ISO26262 ASIL-D certified simulator, aiSim 5 employs state-of-the-art AI-based digital twin technology. This enhances its capability to simulate complex driving scenarios with high precision, making it an ideal environment for testing AD systems. The simulator boasts a proprietary rendering engine that provides a deterministic and high-fidelity virtual reality where sensor simulations can cover diverse climatic and operational conditions, such as snow, rain, and fog, ensuring results are reproducible and reliable.\n\naiSim 5's architecture is designed for flexibility, allowing seamless integration with existing development toolchains, and thus minimizing the need for expensive real-world testing. It features a comprehensive 3D asset library that includes detailed environments, vehicles, and scenarios, which can be customized to generate synthetic data for testing. The solution supports multi-sensor simulations, providing a rich testing ground for developers looking to refine and validate their software stacks.\n\nThanks to its modular C++ and Python APIs, aiSim 5 can be easily deployed within any System under Test (SuT) and CI/CD pipeline, enhancing its adaptability across various automotive applications. Additionally, its open SDK facilitates developer customizations, ensuring aiSim 5 remains adaptable and user-friendly. With built-in scenario randomization, users can efficiently simulate a wide array of driving conditions, making aiSim 5 a powerful tool for ensuring automotive system safety and accuracy.
The SoC Platform by SEMIFIVE facilitates the rapid development of custom silicon chips, optimized for specific applications through the use of domain-specific architectures. Paired with a pool of pre-verified IPs, it lowers the cost, mitigates risks, and speeds up the development timeline compared to traditional methods. This platform effortlessly supports a multitude of applications by providing silicon-proven infrastructure. Supporting various process technologies, this platform integrates seamlessly with existing design methodologies, offering flexibility and the possibility to fine-tune specifications according to application needs. The core of the platform's design philosophy focuses on maximizing reusability and minimizing engineering overhead, key for reducing time-to-market. Designed for simplicity and comprehensiveness, the SoC Platform offers tools and models that ensure quality and reduce integration complexity, from architecture and physical design to software support. As an end-to-end solution, it stands out as a reliable partner for enterprises aiming to bring innovative products to market efficiently and effectively.
The Tyr AI Processor Family is designed around versatile programmability and high performance for AI and general-purpose processing. It consists of variants such as Tyr4, Tyr2, and Tyr1, each offering a unique performance profile optimized for different operational scales. These processors are fully programmable and support high-level programming throughout, ensuring they meet diverse computing needs with precision. Each member of the Tyr family features distinct core configurations, tailored for specific throughput and performance needs. The top-tier Tyr4 boasts 8 cores with a peak capability of 1600 Tflops when leveraging fp8 tensor cores, making it suitable for demanding AI tasks. Tyr2 and Tyr1 scale down these resources to 4 and 2 cores, respectively, achieving proportional efficiency and power savings. All models incorporate substantial on-chip memory, optimizing data handling and execution efficiency without compromising on power use. Moreover, the Tyr processors adapt AI processes automatically on a layer-by-layer basis to enhance implementation efficiency. This adaptability, combined with their near-theory performance levels, renders them ideal for high-throughput AI workloads that require flexible execution and dependable scalability.
The RISC-V CPU IP NS Class is crafted explicitly for applications requiring heightened security and robustness, such as fintech payment systems and IoT security solutions. This processor class is equipped to support secure operations, incorporating features essential for protecting data and ensuring secure communications within devices. This processor integrates security protocols aligned with the RISC-V open standard, offering developers the ability to embed reliable security measures directly at the hardware level. Its architecture provides the foundation for developing systems where data integrity and secure processing are non-negotiable, ensuring that sensitive applications run safely and efficiently. The RISC-V CPU IP NS Class is supported by a strong ecosystem offering tools and resources to facilitate the secure application development process. With its ability to integrate seamlessly with other embedded systems, the NS Class empowers designers to create solutions that prioritize and enhance security in modern digital environments, where threats are constantly evolving.
The Chimera Software Development Kit (SDK) by Quadric empowers developers with a robust platform to create and deploy complex AI and machine learning applications efficiently. It offers tools for developing, simulating, profiling, and deploying software, perfectly adaptable for Quadric’s Chimera GPNPU. The SDK simplifies coding by allowing integration of machine learning graph code with traditional C++ code into a singular, streamlined programming flow. This SDK includes the Chimera LLVM C++ compiler which utilizes state-of-the-art compiler infrastructure tailored to Chimera's specific instruction sets, driving efficiency and optimization. The SDK is compatible with Docker environments, enabling seamless on-premises or cloud-based deployment. This flexibility supports the versatile development needs of corporates working with private proprietary models while streamlining the toolchain for increased productivity. Its Graph Compiler transcodes machine learning inference models from popular frameworks like TensorFlow and PyTorch into optimized C++ using the Chimera Compute Library. This feature ensures that even the most complex AI models are efficiently deployed, lowering computational overheads and maximizing processing potential. Hence, the Chimera SDK serves as an invaluable tool for engineers aiming to expedite the deployment of cutting-edge ML algorithms both effectively and swiftly.
The Vega eFPGA is a flexible programmable solution crafted to enhance SoC designs with substantial ease and efficiency. This IP is designed to offer multiple advantages such as increased performance, reduced costs, secure IP handling, and ease of integration. The Vega eFPGA boasts a versatile architecture allowing for tailored configurations to suit varying application requirements. This IP includes configurable tiles like CLB (Configurable Logic Blocks), BRAM (Block RAM), and DSP (Digital Signal Processing) units. The CLB part includes eight 6-input Lookup Tables that provide dual outputs, and also an optional configuration with a fast adder having a carry chain. The BRAM supports 36Kb dual-port memory and offers flexibility for different configurations, while the DSP component is designed for complex arithmetic functions with its 18x20 multipliers and a wide 64-bit accumulator. Focused on allowing easy system design and acceleration, Vega eFPGA ensures seamless integration and verification into any SoC design. It is backed by a robust EDA toolset and features that allow significant customization, making it adaptable to any semiconductor fabrication process. This flexibility and technological robustness places the Vega eFPGA as a standout choice for developing innovative and complex programmable logic solutions.
Tailored for 64-bit architectures, the RISC-V CPU IP NX Class is crafted to cater to demanding applications requiring advanced processing capabilities. It's particularly well-suited for storage solutions and the burgeoning fields of augmented reality (AR), virtual reality (VR), and artificial intelligence (AI). With its robust architecture, the NX Class processor is capable of handling intensive computational tasks efficiently. The NX Class offers an expansive feature set that ensures high performance and functionality across a diverse range of applications. Utilizing the RISC-V standards, the NX Class provides implementers with the flexibility to customize their solutions, ensuring that the processor meets specific operational and performance criteria essential for high-end applications. Developers benefit from a rich software ecosystem for the NX Class, which includes comprehensive tools and libraries that support rapid development and innovation. The processor is well-equipped to facilitate the development of next-generation products that require powerful processing cores, ensuring that integrators can deliver cutting-edge solutions to markets that demand reliability, speed, and scalability.
The Akida1000 Reference SoC is a fully functional silicon solution designed to demonstrate the capabilities of BrainChip's neuromorphic AI technology. Built to showcase the potential of the Akida architecture, this SoC integrates the neural processor with a variety of connectivity options and memory interfaces, allowing for comprehensive evaluation in real-world settings.\n\nThis reference chip is ideal for developers and businesses looking to explore AI's vast applications at the edge, offering a seamless way to prototype and build AI systems without cloud dependency. The Akida1000 is engineered to function as either an embedded accelerator or a standalone AI coprocessor, supporting a breadth of use cases requiring efficient on-device data processing.\n\nAkida1000's flexibility and high performance make it suitable for diverse industries including automotive, healthcare, and industrial sectors, facilitating predictive analytics and complex problem-solving where traditional processors might struggle. Its design includes standard interfaces like PCIe, USB, and various I/O options, making it easily integrable into existing infrastructures.
FortiPKA-RISC-V is a high-speed public key accelerator that enhances the efficiency of cryptographic tasks by offloading complex operations from the main CPU. It is particularly effective for tasks involving large integer arithmetic typical in asymmetric cryptography. The design eliminates the need for data transformations linked to Montgomery domain conversion, boosting performance significantly. The RISC-V core allows flexible integration using interfaces such as AMBA AXI4, APB, and others. It supports a wide range of cryptographic algorithms including RSA, ECDSA, and SM2, maintaining resilience against side-channel attacks through robust technological methodologies. This solution proves ideal for embedded systems in IoT, automotive, and payment systems, offering high configurability to align with specific performance and area requirements.
The AIoT Platform from SEMIFIVE is crafted to create specialized IoT and edge processing devices with efficiency and cutting-edge technology. Leveraging silicon-proven design components on Samsung's 14nm process, it streamlines the development of high-performance, power-efficient applications. This platform is equipped with dual SiFive U54 RISC-V CPUs, LPDDR4 memory, and comprehensive interfaces like MIPI-CSI and USB3.0. Targeted at consumer electronics such as wearables and smart home devices, this platform supports a wide array of IoT applications, including industrial IoT and smart security systems. Its architectural flexibility allows customization of system specifications, enabling designers to address the unique requirements of diverse IoT deployments. The AIoT platform supports applications with rigorous demands for power efficiency and cost-effectiveness, ensuring swift time-to-market and reduced development cycles. With a collaborative ecosystem of package design, board evaluation, and software, it paves the way for innovative IoT solutions that seamlessly integrate advanced technologies into everyday devices.
The RayCore MC Ray Tracing GPU is a cutting-edge GPU IP known for its real-time path and ray tracing capabilities. Designed to expedite the rendering process efficiently, this GPU IP stands out for its balance of high performance and low power consumption. This makes it ideal for environments requiring advanced graphics processing with minimal energy usage. Capitalizing on world-class ray tracing technology, the RayCore MC ensures seamless, high-quality visual outputs that enrich user experiences across gaming and metaverse applications. Equipped with superior rendering speed, the RayCore MC integrates sophisticated algorithms that handle intricate graphics computations effortlessly. This GPU IP aims to redefine the norms of graphics performance by combining agility in data processing with high fidelity in visual representation. Its real-time rendering finesse significantly enhances user interaction by offering a flawless graphics environment, conducive for both immersive gaming experiences and professional metaverse developments. The RayCore MC GPU IP is also pivotal for developers aiming to push the boundaries of graphics quality and efficiency. With an architecture geared towards optimizing both visual output and power efficiency, it stands as a benchmark for future GPU innovations in high-demand industries. The IP's ability to deliver rapid rendering with superior graphic integrity makes it a preferred choice among developers focused on pioneering graphics-intensive applications.
The Maverick-2 Intelligent Compute Accelerator is an advanced computing architecture that revolutionizes the handling of high-performance tasks through intelligent software-defined hardware. It optimizes workloads in real time, ensuring that computational resources are used efficiently to accelerate applications in high-performance computing (HPC) and artificial intelligence (AI). Designed to support demanding workloads, Maverick-2 improves developer productivity and reduces the time-to-solution significantly by bypassing the traditional need for extensive application porting. This cutting-edge system uses an adaptable architecture that dynamically adjusts to application demands, freeing developers to focus on scientific or computational breakthroughs without being encumbered by technical complexities. With support for a wide array of programming languages and frameworks, including C/C++, FORTRAN, and OpenMP, Maverick-2 provides a flexible platform that reduces the need for code rewriting or vendor-specific tools. Operating on TSMC's advanced 5nm process technology, the Maverick-2 boasts significant power efficiencies and scalability, with deployments in single or dual configurations to suit various computational needs. It combines HBM3E memory and high-bandwidth interfaces to ensure high throughput and low-latency data access, crucial for many contemporary scientific and industrial applications. Moreover, the intelligent architecture adapts over time, future-proofing investments by evolving alongside new computing needs and technologies.
The NoISA Processor by Hotwright Inc. is designed to revolutionize conventional processing paradigms by eliminating the constraints of fixed instruction set architectures. Utilizing the Hotstate machine, this processor is an advanced runtime loadable microcoded algorithmic state machine that operates based on a specialized subset of C code. Unlike traditional processors that use predefined ALUs, register files, and controllers, the NoISA Processor is highly adaptable, allowing for dynamic modifications via microcode rather than rigid instructions. The design philosophy behind the NoISA Processor is to offer a solution for applications where traditional softcore CPUs are either too resource-intensive or insufficiently fast. It excels in power efficiency, making it particularly suitable for edge computing and IoT devices where every milliwatt counts. With the ability to rapidly design small, efficient controllers or C-programmable state machines, it provides unparalleled freedom and performance without being tied to a fixed instruction set. Additionally, the NoISA Processor offers significant advantages for systolic arrays and FPGA-related applications, allowing changes in the behavior without hardware modification. This flexibility enables developers to achieve peak performance levels as the processor's behavior can be updated by simply reloading microcode, thus offering a more agile and powerful alternative to traditional processors.
The Prodigy Universal Processor from Tachyum is a groundbreaking innovation designed to revolutionize data center performance. It remarkably integrates the power of CPUs, GPGPUs, and TPUs into a single homogeneous architecture. This processor architecture supports not only general computing tasks but also excels in high-performance computing, AI, and deep machine learning applications. The Prodigy is engineered to reduce power consumption drastically while significantly boosting server utilization and space efficiency, a much-needed advantage for modern data centers. One of its most praised attributes is its unrivaled performance per watt, being able to deliver up to 18 times higher performance and six times better energy efficiency compared to its peers. The processor's design overcomes the technological bottlenecks that have traditionally hindered data center efficiency, such as excessive power usage and low server utilization rates. Its streamlined architecture simplifies programming, offering a coherent multiprocessor model that easily integrates into existing data center infrastructures. Moreover, Tachyum's Universal Processor breaks free from the constraints imposed by Moore's Law, setting new standards in computational power and energy utilization. Its innovative approach allows seamless execution of traditional and high-demand AI tasks without necessitating significant overhauls in the software environment. As such, this processor is poised to be a key player in emerging technologies, driving future developments in AI and helping propel forward-thinking organizational strategies across the globe.
The AON1020 enhances the capabilities of edge AI by addressing not only voice and audio but also various sensor applications. It extends the AONSens™ Neural Network cores line to include comprehensive sensor activities such as human motion detection. This IP delivers its AI processing engine in Verilog RTL, making it versatile for inclusion in numerous ASIC and FPGA designs, while maintaining a focus on low power consumption and high functional accuracy.
AON1100 is acclaimed as the leading edge AI chip optimized for voice and sensor applications. With a power consumption of less than 260µW, it excels in delivering near-perfect accuracy even in challenging environments. This makes it indispensably suited for continuously active devices that require both precision and energy efficiency.
The RISC-V CPU IP NI Class is tailored for AI, ADAS communications, and advanced video processing applications. Developed for performance-intensive environments, this processor is equipped to handle complex computational tasks required for sophisticated AI operations and real-time data processing in advanced driver-assistance systems (ADAS). With its focus on AI and video applications, the NI Class supports a wide range of functionalities that promote efficient processing and integration into systems that demand high computational capabilities. The flexibility of the RISC-V architecture allows developers to implement custom solutions that meet specific application criteria, ensuring efficient deployment across varied markets. The RISC-V CPU IP NI Class is supported by a dynamic ecosystem providing extensive development tools and libraries to maximize its potential in applications. This ecosystem facilitates rapid prototyping and deployment, allowing designers to stay ahead in evolving industries such as multimedia processing and automotive technology, where performance and adaptability are keys to success.
Join the world's most advanced semiconductor IP marketplace!
It's free, and you'll get all the tools you need to discover IP, meet vendors and manage your IP workflow!
Join the world's most advanced AI-powered semiconductor IP marketplace!
It's free, and you'll get all the tools you need to advertise and discover semiconductor IP, keep up-to-date with the latest semiconductor news and more!
Plus we'll send you our free weekly report on the semiconductor industry and the latest IP launches!