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.
The KL730 is a third-generation AI chip that integrates advanced reconfigurable NPU architecture, delivering up to 8 TOPS of computing power. This cutting-edge technology enhances computational efficiency across a range of applications, including CNN and transformer networks, while minimizing DDR bandwidth requirements. The KL730 also boasts enhanced video processing capabilities, supporting 4K 60FPS outputs. With expertise spanning over a decade in ISP technology, the KL730 stands out with its noise reduction, wide dynamic range, fisheye correction, and low-light imaging performance. It caters to markets like intelligent security, autonomous vehicles, video conferencing, and industrial camera systems, among others.
The Akida 2nd Generation represents a leap forward in the realm of AI processing, enhancing upon its predecessor with greater flexibility and improved efficiency. This advanced neural processor core is tailored for modern applications demanding real-time response and ultra-low power consumption, making it ideal for compact and battery-operated devices. Akida 2nd Generation supports various programming configurations, including 8-, 4-, and 1-bit weights and activations, thus providing developers with the versatility to optimize performance versus power consumption to meet specific application needs. Its architecture is fully digital and silicon-proven, ensuring reliable deployment across diverse hardware setups. With features such as programmable activation functions and support for sophisticated neural network models, Akida 2nd Generation enables a broad spectrum of AI tasks. From object detection in cameras to sophisticated audio sensing, this iteration of the Akida processor is built to handle the most demanding edge applications while sustaining BrainChip's hallmark efficiency in processing power per watt.
Addressing the need for high-performance AI processing, the Metis AIPU PCIe AI Accelerator Card from Axelera AI offers an outstanding blend of speed, efficiency, and power. Designed to boost AI workloads significantly, this PCIe card leverages the prowess of the Metis AI Processing Unit (AIPU) to deliver unparalleled AI inference capabilities for enterprise and industrial applications. The card excels in handling complex AI models and large-scale data processing tasks, significantly enhancing the efficiency of computational tasks within various edge settings. The Metis AIPU embedded within the PCIe card delivers high TOPs (Tera Operations Per Second), allowing it to execute multiple AI tasks concurrently with remarkable speed and precision. This makes it exceptionally suitable for applications such as video analytics, autonomous driving simulations, and real-time data processing in industrial environments. The card's robust architecture reduces the load on general-purpose processors by offloading AI tasks, resulting in optimized system performance and lower energy consumption. With easy integration capabilities supported by the state-of-the-art Voyager SDK, the Metis AIPU PCIe AI Accelerator Card ensures seamless deployment of AI models across various platforms. The SDK facilitates efficient model optimization and tuning, supporting a wide range of neural network models and enhancing overall system capabilities. Enterprises leveraging this card can see significant improvements in their AI processing efficiency, leading to faster, smarter, and more efficient operations across different sectors.
The Akida IP is an advanced processor core designed to mimic the efficient processing characteristics of the human brain. Inspired by neuromorphic engineering principles, it delivers real-time AI performance while maintaining a low power profile. The architecture of the Akida IP is sophisticated, allowing seamless integration into existing systems without the need for continuous external computation. Equipped with capabilities for processing vision, audio, and sensor data, the Akida IP stands out by being able to handle complex AI tasks directly on the device. This is done by utilizing a flexible mesh of nodes that efficiently distribute cognitive computing tasks, enabling a scalable approach to machine learning applications. Each node supports hundreds of MAC operations and can be configured to adapt to various computational requirements, making it a versatile choice for AI-centric endeavors. Moreover, the Akida IP is particularly beneficial for edge applications where low latency, high efficiency, and security are paramount. With capabilities for event-based processing and on-chip learning, it enhances response times and reduces data transfer needs, thereby bolstering device autonomy. This solidifies its position as a leading solution for embedding AI into devices across multiple industries.
The Yitian 710 Processor is a landmark server chip released by T-Head Semiconductor, representing a breakthrough in high-performance computing. This chip is designed with cutting-edge architecture that utilizes advanced Armv9 structure, accommodating a range of demanding applications. Engineered by T-Head's dedicated research team, Yitian 710 integrates high efficiency and bandwidth properties into a unique 2.5D package, housing two dies and a staggering 60 billion transistors. The Yitian 710 encompasses 128 Armv9 high-performance cores, each equipped with 64KB L1 instruction cache, 64KB L1 data cache, and 1MB L2 cache, further amplified by a collective on-chip system cache of 128MB. These configurations enable optimal data processing and retrieval speeds, making it suitable for data-intensive tasks. Furthermore, the memory subsystem stands out with its 8-channel DDR5 support, reaching peak bandwidths of 281GB/s. In terms of connectivity, the Yitian 710's I/O system includes 96 PCIe 5.0 channels with a bidirectional theoretical total bandwidth of 768GB/s, streamlining high-speed data transfer critical for server operations. Its architecture is not only poised to meet the current demands of data centers and cloud services but also adaptable for future advancements in AI inference and multimedia processing tasks.
The AI Camera Module from Altek is a versatile, high-performance component designed to meet the increasing demand for smart vision solutions. This module features a rich integration of imaging lens design and combines both hardware and software capacities to create a seamless operational experience. Its design is reinforced by Altek's deep collaboration with leading global brands, ensuring a top-tier product capable of handling diverse market requirements. Equipped to cater to AI and IoT interplays, the module delivers outstanding capabilities that align with the expectations for high-resolution imaging, making it suitable for edge computing applications. The AI Camera Module ensures that end-user diversity is meaningfully addressed, offering customization in device functionality which supports advanced processing requirements such as 2K and 4K video quality. This module showcases Altek's prowess in providing comprehensive, all-in-one camera solutions which leverage sophisticated imaging and rapid processing to handle challenging conditions and demands. The AI Camera's technical blueprint supports complex AI algorithms, enhancing not just image quality but also the device's interactive capacity through facial recognition and image tracking technology.
Chimera GPNPU is engineered to revolutionize AI/ML computational capabilities on single-core architectures. It efficiently handles matrix, vector, and scalar code, unifying AI inference and traditional C++ processing under one roof. By alleviating the need for partitioning AI workloads between different processors, it streamlines software development and drastically speeds up AI model adaptation and integration. Ideal for SoC designs, the Chimera GPNPU champions an architecture that is both versatile and powerful, handling complex parallel workloads with a single unified binary. This configuration not only boosts software developer productivity but also ensures an enduring flexibility capable of accommodating novel AI model architectures on the horizon. The architectural fabric of the Chimera GPNPU seamlessly blends the high matrix performance of NPUs with C++ programmability found in traditional processors. This core is delivered in a synthesizable RTL form, with scalability options ranging from a single-core to multi-cluster designs to meet various performance benchmarks. As a testament to its adaptability, the Chimera GPNPU can run any AI/ML graph from numerous high-demand application areas such as automotive, mobile, and home digital appliances. Developers seeking optimization in inference performance will find the Chimera GPNPU a pivotal tool in maintaining cutting-edge product offerings. With its focus on simplifying hardware design, optimizing power consumption, and enhancing programmer ease, this processor ensures a sustainable and efficient path for future AI/ML developments.
aiWare is a high-performance NPU designed to meet the rigorous demands of automotive AI inference, providing a scalable solution for ADAS and AD applications. This hardware IP core is engineered to handle a wide array of AI workloads, including the most advanced neural network structures like CNNs, LSTMs, and RNNs. By integrating cutting-edge efficiency and scalability, aiWare delivers industry-leading neural processing power tailored to automobile-grade specifications.\n\nThe NPU's architecture emphasizes hardware determinism and offers ISO 26262 ASIL-B certification, ensuring that aiWare meets stringent automotive safety standards. Its efficient design also supports up to 256 effective TOPS per core, and can scale to handle thousands of TOPS through multicore integration, minimizing power consumption effectively. The aiWare's system-level optimizations reduce reliance on external memory by leveraging local memory for data management, boosting performance efficiency across varied input data sizes and complexities.\n\naiWare’s development toolkit, aiWare Studio, is distinguished by its innovative ability to optimize neural network execution without the need for manual intervention by software engineers. This empowers ai engineers to focus on refining NNs for production, significantly accelerating iteration cycles. Coupled with aiMotive's aiDrive software suite, aiWare provides an integrated environment for creating highly efficient automotive AI applications, ensuring seamless integration and rapid deployment across multiple vehicle platforms.
SAKURA-II AI Accelerator represents EdgeCortix's latest advancement in edge AI processing, offering unparalleled energy efficiency and extensive capabilities for generative AI tasks. This accelerator is designed to manage demanding AI models, including Llama 2, Stable Diffusion, DETR, and ViT, within a slim power envelope of about 8W. With capabilities extending to multi-billion parameter models, SAKURA-II meets a wide range of edge applications in vision, language, and audio. The SAKURA-II's architecture maximizes AI compute efficiency, delivering more than twice the utilization of competitive solutions. It boasts remarkable DRAM bandwidth, essential for large language and vision models, while maintaining low power consumption. The hardware supports real-time Batch=1 processing, demonstrating its edge in performance even in constrained environments, making it a choice solution for diverse industrial AI applications. With 60 TOPS (INT8) and 30 TFLOPS (BF16) in performance metrics, this accelerator is built to exceed expectations in demanding conditions. It features robust memory configurations supporting up to 32GB of DRAM, ideal for processing intricate AI workloads. By leveraging sparse computing techniques, SAKURA-II optimizes its memory and bandwidth usage effectively, ensuring reliable performance across all deployed applications.
xcore.ai is XMOS Semiconductor's innovative programmable chip designed for advanced AI, DSP, and I/O applications. It enables developers to create highly efficient systems without the complexity typical of multi-chip solutions, offering capabilities that integrate AI inference, DSP tasks, and I/O control seamlessly. The chip architecture boasts parallel processing and ultra-low latency, making it ideal for demanding tasks in robotics, automotive systems, and smart consumer devices. It provides the toolset to deploy complex algorithms efficiently while maintaining robust real-time performance. With xcore.ai, system designers can leverage a flexible platform that supports the rapid prototyping and development of intelligent applications. Its performance allows for seamless execution of tasks such as voice recognition and processing, industrial automation, and sensor data integration. The adaptable nature of xcore.ai makes it a versatile solution for managing various inputs and outputs simultaneously, while maintaining high levels of precision and reliability. In automotive and industrial applications, xcore.ai supports real-time control and monitoring tasks, contributing to smarter, safer systems. For consumer electronics, it enhances user experience by enabling responsive voice interfaces and high-definition audio processing. The chip's architecture reduces the need for exterior components, thus simplifying design and reducing overall costs, paving the way for innovative solutions where technology meets efficiency and scalability.
The Metis AIPU M.2 Accelerator Module by Axelera AI is a compact and powerful solution designed for AI inference at the edge. This module delivers remarkable performance, comparable to that of a PCIe card, all while fitting into the streamlined M.2 form factor. Ideal for demanding AI applications that require substantial computational power, the module enhances processing efficiency while minimizing power usage. With its robust infrastructure, it is geared toward integrating into applications that demand high throughput and low latency, making it a perfect fit for intelligent vision applications and real-time analytics. The AIPU, or Artificial Intelligence Processing Unit, at the core of this module provides industry-leading performance by offloading AI workloads from traditional CPU or GPU setups, allowing for dedicated AI computation that is faster and more energy-efficient. This not only boosts the capabilities of the host systems but also drastically reduces the overall energy consumption. The module supports a wide range of AI applications, from facial recognition and security systems to advanced industrial automation processes. By utilizing Axelera AI’s innovative software solutions, such as the Voyager SDK, the Metis AIPU M.2 Accelerator Module enables seamless integration and full utilization of AI models and applications. The SDK offers enhancements like compatibility with various industry tools and frameworks, thus ensuring a smooth deployment process and quick time-to-market for advanced AI systems. This product represents Axelera AI’s commitment to revolutionizing edge computing with streamlined, effective AI acceleration solutions.
The Talamo SDK from Innatera serves as a comprehensive software development toolkit designed to maximize the capabilities of its Spiking Neural Processor (SNP) lineup. Tailored for developers and engineers, Talamo offers in-depth access to configure and deploy neuromorphic processing solutions effectively. The SDK supports the development of applications that utilize Spiking Neural Networks (SNNs) for diverse sensory processing tasks. Talamo provides a user-friendly interface that simplifies the integration of neural processing capabilities into a wide range of devices and systems. By leveraging the toolkit, developers can customize applications for specific use cases such as real-time audio analysis, touch-free interactions, and biometric data processing. The SDK comes with pre-built models and a model zoo, which helps in rapidly prototyping and deploying sensor-driven solutions. This SDK stands out by offering enhanced tools for developing low-latency, energy-efficient applications. By harnessing the temporal processing strength of SNNs, Talamo allows for the robust development of applications that can operate under strict power and performance constraints, enabling the creation of intelligent systems that can autonomously process data in real-time.
The KL630 is a pioneering AI chipset featuring Kneron's latest NPU architecture, which is the first to support Int4 precision and transformer networks. This cutting-edge design ensures exceptional compute efficiency with minimal energy consumption, making it ideal for a wide array of applications. With an ARM Cortex A5 CPU at its core, the KL630 excels in computation while maintaining low energy expenditure. This SOC is designed to handle both high and low light conditions optimally and is perfectly suited for use in diverse edge AI devices, from security systems to expansive city and automotive networks.
The Hanguang 800 AI Accelerator by T-Head Semiconductor is a powerful AI acceleration chip designed to enhance machine learning tasks. It excels in providing the computational power necessary for intensive AI workloads, effectively reducing processing times for large-scale data frameworks. This makes it an ideal choice for organizations aiming to infuse AI capabilities into their operations with maximum efficiency. Built with an emphasis on speed and performance, the Hanguang 800 is optimized for applications requiring vast amounts of data crunching. It supports a diverse array of AI models and workloads, ensuring flexibility and robust performance across varying use cases. This accelerates the deployment of AI applications in sectors such as autonomous driving, natural language processing, and real-time data analysis. The Hanguang 800's architecture is complemented by proprietary algorithms that enhance processing throughput, competing against traditional processors by providing significant gains in efficiency. This accelerator is indicative of T-Head's commitment to advancing AI technologies and highlights their capability to cater to specialized industry needs through innovative semiconductor developments.
The KL520 marks Kneron's foray into the edge AI landscape, offering an impressive combination of size, power efficiency, and performance. Armed with dual ARM Cortex M4 processors, this chip can operate independently or as a co-processor to enable AI functionalities such as smart locks and security monitoring. The KL520 is adept at 3D sensor integration, making it an excellent choice for applications in smart home ecosystems. Its compact design allows devices powered by it to operate on minimal power, such as running on AA batteries for extended periods, showcasing its exceptional power management capabilities.
The KL530 represents a significant advancement in AI chip technology with a new NPU architecture optimized for both INT4 precision and transformer networks. This SOC is engineered to provide high processing efficiency and low power consumption, making it suitable for AIoT applications and other innovative scenarios. It features an ARM Cortex M4 CPU designed for low-power operation and offers a robust computational power of up to 1 TOPS. The chip's ISP enhances image quality, while its codec ensures efficient multimedia compression. Notably, the chip's cold start time is under 500 ms with an average power draw of less than 500 mW, establishing it as a leader in energy efficiency.
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 Maverick-2 Intelligent Compute Accelerator represents the pinnacle of Next Silicon's innovative approach to computational resources. This state-of-the-art accelerator leverages the Intelligent Compute Architecture for software-defined adaptability, enabling it to autonomously tailor its real-time operations across various HPC and AI workloads. By optimizing performance using insights gained through real-time telemetry, Maverick-2 ensures superior computational efficiency and reduced power consumption, making it an ideal choice for demanding computational environments.\n\nMaverick-2 brings transformative performance enhancements to large-scale scientific research and data-heavy industries by dispensing with the need for codebase modifications or specialized software stacks. It supports a wide range of familiar development tools and frameworks, such as C/C++, FORTRAN, and Kokkos, simplifying the integration process for developers and reducing time-to-discovery significantly.\n\nEngineered with advanced features like high bandwidth memory (HBM3E) and built on TSMC's 5nm process technology, this accelerator provides not only unmatched adaptability but also an energy-efficient, eco-friendly computing solution. Whether embedded in single-die PCIe cards or dual-die OCP Accelerator Modules, the Maverick-2 is positioned as a future-proof solution capable of evolving with technological advancements in AI and HPC.
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)
The WiseEye2 AI Solution by Himax is a highly efficient processor tailored for AI applications, combining an ultra-low power CMOS image sensor with the HX6538 microcontroller. Designed for battery-powered devices requiring continuous operation, it significantly lowers power consumption while boosting performance. This processor leverages the Arm Cortex M55 CPU and Ethos U55 NPU to enhance inference speed and energy efficiency substantially, allowing execution of complex models with precision. Perfectly suited for applications like user-presence detection in laptops, WiseEye2 heightens security through sophisticated facial recognition and adaptive privacy settings. It automatically wakes or locks the device based on user proximity, thereby conserving power and safeguarding sensitive information. WiseEye2’s power management and neural processing units ensure constant device readiness, augmenting AI capabilities ranging from occupancy detection to smart security. This reflects Himax's dedication to providing versatile AI solutions that anticipate and respond to user needs seamlessly. Equipped with sensor fusion and sophisticated security engines, WiseEye2 maintains privacy and operational efficiency, marking a significant step forward in the integration of AI into consumer electronics, especially where energy conservation is pivotal.
Altek's 3D Imaging Chip is a breakthrough in the field of vision technology. Designed with an emphasis on depth perception, it enhances the accuracy of 3D scene capturing, making it ideal for applications requiring precise distance gauging such as autonomous vehicles and drones. The chip integrates seamlessly within complex systems, boasting superior recognition accuracy that ensures reliable and robust performance. Building upon years of expertise in 3D imaging, this chip supports multiple 3D modes, offering flexible solutions for devices from surveillance robots to delivery mechanisms. It facilitates medium-to-long-range detection needs thanks to its refined depth sensing capabilities. Altek's approach ensures a comprehensive package from modular design to chip production, creating a cohesive system that marries both hardware and software effectively. Deployed within various market segments, it delivers adaptable image solutions with dynamic design agility. Its imaging prowess is further enhanced by state-of-the-art algorithms that refine image quality and facilitate facial detection and recognition, thereby expanding its utility across diverse domains.
StarFive's Tianqiao-70 is engineered to deliver superior performance in a power-efficient package. This 64-bit RISC-V CPU core is designed for commercial-grade applications, where consistent and reliable performance is mandatory, yet energy consumption must be minimized. The core's architecture integrates low power design principles without compromising its ability to execute complex instructions efficiently. It is particularly suited for mobile applications, desktop clients, and intelligent gadgets requiring sustained battery life. The Tianqiao-70's design focuses on extending the operational life of devices by ensuring minimal power draw during both active and idle states. It supports an array of advanced features that cater to the latest computational demands. As an ideal solution for devices that combine portability with intensive processing demands, the Tianqiao-70 offers an optimal balance of performance and energy conservation. Its capability to adapt to various operating environments makes it a versatile option for developers looking to maximize efficiency and functionality.
aiSim 5 stands as a cutting-edge simulation tool specifically crafted for the automotive sector, with a strong focus on validating ADAS and autonomous driving solutions. It distinguishes itself with an AI-powered digital twin creation capability, offering a meticulously optimized sensor simulation environment that guarantees reproducibility and determinism. The adaptable architecture of aiSim allows seamless integration with existing industry toolchains, significantly minimizing the need for costly real-world testing.\n\nOne of the key features of aiSim is its capability to simulate various challenging weather conditions, enhancing testing accuracy across diverse environments. This includes scenarios like snowstorms, heavy fog, and rain, with sensors simulated based on physics, offering changes in conditions in real-time. Its certification with ISO 26262 ASIL-D attests to its automotive-grade quality and reliability, providing a new standard for testing high-fidelity sensor data in varied operational design domains.\n\nThe flexibility of aiSim is further highlighted through its comprehensive SDKs and APIs, which facilitate smooth integration into various systems under test. Additionally, users can leverage its extensive 3D asset library to establish detailed, realistic testing environments. AI-based rendering technologies underpin aiSim's data simulation, achieving both high efficiency and accuracy, thereby enabling rapid and effective validation of advanced driver assistance and autonomous driving systems.
RAIV represents Siliconarts' General Purpose-GPU (GPGPU) offering, engineered to accelerate data processing across diverse industries. This versatile GPU IP is essential in sectors engaged in high-performance computing tasks, such as autonomous driving, IoT, and sophisticated data centers. With RAIV, Siliconarts taps into the potential of the fourth industrial revolution, enabling rapid computation and seamless data management. The RAIV architecture is poised to deliver unmatched efficiency in high-demand scenarios, supporting massive parallel processing and intricate calculations. It provides an adaptable framework that caters to the needs of modern computing, ensuring balanced workloads and optimized performance. Whether used for VR/AR applications or supporting the back-end infrastructure of data-intensive operations, RAIV is designed to meet and exceed industry expectations. RAIV’s flexible design can be tailored to enhance a broad spectrum of applications, promising accelerated innovation in sectors dependent on AI and machine learning. This GPGPU IP not only underscores Siliconarts' commitment to technological advancement but also highlights its capability to craft solutions that drive forward computational boundaries.
The KL720 AI SoC is designed for optimal performance-to-power ratios, achieving 0.9 TOPS per watt. This makes it one of the most efficient chips available for edge AI applications. The SOC is crafted to meet high processing demands, suitable for high-end devices including smart TVs, AI glasses, and advanced cameras. With an ARM Cortex M4 CPU, it enables superior 4K imaging, full HD video processing, and advanced 3D sensing capabilities. The KL720 also supports natural language processing (NLP), making it ideal for emerging AI interfaces such as AI assistants and gaming gesture controls.
The Spiking Neural Processor T1 is an ultra-low power processor developed specifically for enhancing sensor capabilities at the edge. By leveraging advanced Spiking Neural Networks (SNNs), the T1 efficiently deciphers patterns in sensor data with minimal latency and power usage. This processor is especially beneficial in real-time applications, such as audio recognition, where it can discern speech from audio inputs with sub-millisecond latency and within a strict power budget, typically under 1mW. Its mixed-signal neuromorphic architecture ensures that pattern recognition functions can be continually executed without draining resources. In terms of processing capabilities, the T1 resembles a dedicated engine for sensor tasks, offering functionalities like signal conditioning, filtering, and classification independent of the main application processor. This means tasks traditionally handled by general-purpose processors can now be offloaded to the T1, conserving energy and enhancing performance in always-on scenarios. Such functionality is crucial for pervasive sensing tasks across a range of industries. With an architecture that balances power and performance impeccably, the T1 is prepared for diverse applications spanning from audio interfaces to the rapid deployment of radar-based touch-free interactions. Moreover, it supports presence detection systems, activity recognition in wearables, and on-device ECG processing, showcasing its versatility across various technological landscapes.
The RayCore MC is a revolutionary real-time path and ray-tracing GPU designed to enhance rendering with minimal power consumption. This GPU IP is tailored for real-time applications, offering a rich graphical experience without compromising on speed or efficiency. By utilizing advanced ray-tracing capabilities, RayCore MC provides stunning visual effects and lifelike animations, setting a high standard for quality in digital graphics. Engineered for scalability and performance, RayCore MC stands out in the crowded field of GPU technologies by delivering seamless, low-latency graphics. It is particularly suited for applications in gaming, virtual reality, and the burgeoning metaverse, where realistic rendering is paramount. The architecture supports efficient data management, ensuring that even the most complex visual tasks are handled with ease. RayCore MC's architecture supports a wide array of applications beyond entertainment, making it a vital tool in areas such as autonomous vehicles and data-driven industries. Its blend of power efficiency and graphical prowess ensures that developers can rely on RayCore MC for cutting-edge, resource-light graphic solutions.
The SEMIFIVE AI Inference Platform is engineered to facilitate rapid development and deployment of AI inference solutions within custom silicon environments. Utilizing seamless integration with silicon-proven IPs, this platform delivers a high-performance framework optimized for AI and machine learning tasks. By providing a strategic advantage in cost reduction and efficiency, the platform decreases time-to-market challenges through pre-configured model layers and extensive IP libraries tailored for AI applications. It also offers enhanced scalability through its support for various computational and network configurations, making it adaptable to both high-volume and specialized market segments. This platform supports complex AI workloads on scalable AI engines, ensuring optimized performance in data-intensive operations. The integration of advanced processors and memory solutions within the platform further enhances processing efficiency, positioning it as an ideal solution for enterprises focusing on breakthroughs in AI technologies.
ELFIS2 is a sophisticated visible light imaging ASIC designed to deliver superior performance under the extreme conditions typical in space. Known for its radiation hard design, it withstands both total ionizing dose (TID) and single event effects (SEU/SEL), ensuring dependable operation even when exposed to high radiation levels in outer space. This image sensor features HDR (High Dynamic Range) capabilities, which allow it to capture clear, contrast-rich images in environments with varying light intensities, without motion artifacts thanks to its Motion Artifact Free (MAF) technology. Additionally, its global shutter ensures that every pixel is exposed simultaneously, preventing distortion in high-speed imaging applications. Utilizing back-side illumination (BSI), the ELFIS2 achieves superior sensitivity and quantum efficiency, making it an ideal choice for challenging lighting conditions. This combination of advanced features makes the ELFIS2 particularly well-suited for scientific and space-based imaging applications requiring exacting standards.
The Dynamic Neural Accelerator II (DNA-II) from EdgeCortix is an advanced neural network IP core tailored for high efficiency and parallelism at the edge. Incorporating a run-time reconfigurable interconnect system between compute units, DNA-II effectively manages both convolutional and transformer workloads. This architecture ensures scalable performance beginning with 1K MACs, suitable for a wide range of SoC implementations. EdgeCortix's patented architecture significantly optimizes data paths between DNA engines, enhancing parallelism while reducing on-chip memory usage. As a core component of the SAKURA-II platform, DNA-II supports state-of-the-art generative AI models with industry-leading energy efficiency. DNA-II's design acknowledges the typical inefficiencies in IP cores, improving compute utilization and power consumption metrics substantially. By adopting innovative reconfigurable datapath technologies, EdgeCortix sets a new benchmark for low-power, high-performance edge AI applications.
The RISCV SoC - Quad Core Server Class is engineered for high-performance applications requiring robust processing capabilities. Designed around the RISC-V architecture, this SoC integrates four cores to offer substantial computing power. It's ideal for server-class operations, providing both performance efficiency and scalability. The RISCV architecture allows for open-source compatibility and flexible customization, making it an excellent choice for users who demand both power and adaptability. This SoC is engineered to handle demanding workloads efficiently, making it suitable for various server applications.
The Polar ID is an innovative biometric security system that elevates facial recognition technology through its use of sophisticated meta-optics. By capturing and analyzing the unique polarization signature of a face, Polar ID delivers a new standard in security. This system can detect spoofing attempts that incorporate 3D masks or other similar deceptive tactics, ensuring high security through accurate human authentication. Polar ID minimizes the complexity typically associated with face recognition systems by integrating essential optical functions into a single optic. Its compact design results in cost savings and reduces the space required for optical modules in devices like smartphones. Operating in near-infrared light, Polar ID can consistently deliver secure face unlock capabilities even under challenging lighting conditions, dramatically outperforming traditional systems that may fail in bright or dark environments. The platform does not rely on time-of-flight sensors or structured light projectors, which are costly and complex. Instead, Polar ID leverages the simplicity and efficiency of its single-shot identification process to deliver immediate authentication results. This makes it a potent tool for securing mobile transactions and providing safer user experiences in consumer technology.
The 2D FFT IP extends the power of the traditional FFT by enabling two-dimensional transforms, essential for image and signal processing where data is structured in matrices. With an impressive balance of speed and resource utilization, the 2D FFT handles massive data efficiently using internal or external memory interfaces to fit broad application demands. Its adaptability for FPGA and ASIC applications makes it an ideal candidate for high-performance computing tasks needing complex data manipulation.
The Prodigy Universal Processor by Tachyum is a groundbreaking innovation in the realm of computing, marked as the world's first processor that merges General Purpose Computing, High-Performance Computing, Artificial Intelligence, and various other AI disciplines into a single compact chip. This processor promises to revolutionize hyperscale data centers with its unprecedented processing capabilities and efficiency, pushing the boundaries of current computational power. With its superior performance per watt, Prodigy minimizes energy consumption while maximizing data processing abilities. Offering up to 21 times higher performance compared to its contemporaries, Prodigy stands out by providing a coherent multiprocessor architecture that simplifies the programming environment. It aims to overcome challenges like high power use and server underutilization, which have long plagued modern data centers. By addressing these core issues, it allows enterprises to manage workloads more effectively and sustainably. Furthermore, Prodigy's emulation platform broadens the scope of testing and evaluation, enabling developers to optimize their applications for better performance and low power consumption. With native support for the Prodigy instruction set architecture, the processor seamlessly integrates existing software packages, promising a smooth transition and robust application support. Through the integration of this versatile processor, Tachyum is leading the charge toward a sustainable technological future.
The Trifecta-GPU offers cutting-edge graphics processing capabilities designed for high-efficiency computing needs. This PXIe/CPCIe module excels in handling intensive tasks across various applications, including signal processing, modular test and measurement, and electronic warfare systems. Built to deliver robust performance, it incorporates advanced GPU technology to ensure rapid data throughput and high computational capability. With a focus on versatility, the Trifecta-GPU seamlessly integrates with existing hardware setups, aiding in the enhancement of system performance through its powerful data handling skills. It is particularly well-suited for environments that demand precise data analysis and execution speed, such as AI and machine learning inference tasks. Its inclusion in RADX's product lineup signifies its importance in providing comprehensive solutions tailored for demanding industrial and research applications. Moreover, this module supports various applications, empowered by its substantial memory bandwidth, and possesses innovative architecture designed to optimize processing power. The Trifecta-GPU is an integral component within RADX’s lineup designed to offer flexibility and power efficiency in equal measure, making it well-suited for future-tech applications that necessitate high-performance standards.
Specially engineered for the automotive industry, the NA Class IP by Nuclei complies with the stringent ISO26262 functional safety standards. This processor is crafted to handle complex automotive applications, offering flexibility and rigorous safety protocols necessary for mission-critical transportation technologies. Incorporating a range of functional safety features, the NA Class IP is equipped to ensure not only performance but also reliability and safety in high-stakes vehicular environments.
The TT-Ascalon™ is a high-performance RISC-V CPU designed for general-purpose control, emphasizing power and area efficiency. This processor features an Out-of-Order, superscalar architecture that adheres to the RISC-V RVA23 profile, co-developed with Tenstorrent's own Tensix IP for optimized performance. TT-Ascalon™ is highly scalable, suitable for various high-demand applications that benefit from robust computational capabilities. It's engineered to deliver unmatched performance while maintaining energy efficiency, making it ideal for operations that require reliability without compromising on speed and power efficiency.
The NS Class is Nuclei's crucial offering for applications prioritizing security and fintech solutions. This RISC-V CPU IP securely manages IoT environments with its highly customizable and secure architecture. Equipped to support advanced security protocols and functional safety features, the NS Class is particularly suited for payment systems and other fintech applications, ensuring robust protection and reliable operations. Its design follows the RISC-V standards and is accompanied by customizable configuration options tailored to meet specific security requirements.
The NX Class RISC-V CPU IP by Nuclei is characterized by its 64-bit architecture, making it a robust choice for storage, AR/VR, and AI applications. This processing unit is designed to accommodate high data throughput and demanding computational tasks. By leveraging advanced capabilities, such as virtual memory and enhanced processing power, the NX Class facilitates cutting-edge technological applications and is adaptable for integration into a vast array of high-performance systems.
The NI Class RISC-V CPU IP caters to communication, video processing, and AI applications, providing a balanced architecture for intensive data handling and processing capabilities. With a focus on high efficiency and flexibility, this processor supports advanced data crunching and networking applications, ensuring that systems run smoothly and efficiently even when managing complex algorithms. The NI Class upholds Nuclei's commitment to providing versatile solutions in the evolving tech landscape.
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.
SiFive Performance family processors are specifically engineered to deliver outstanding performance and efficiency across a wide range of applications. These processors cater to diverse market demands, including data centers, consumer electronics, and AI-driven workloads. They feature high-performance, 64-bit out-of-order cores with optional vector engines, making them ideal for heavy-duty tasks requiring maximum throughput and scalability. The series incorporates a variety of architectural features that optimize performance and energy efficiency. It includes cores scalable from three-wide to six-wide, supporting up to 256-bit vector operations, which are particularly advantageous for AI and multimedia processing applications. This optimal balance ensures that each core offers superior compute density and power efficiency. Additionally, the SiFive Performance series emphasizes flexibility, allowing users to mix and match cores to achieve the desired balance between performance and power consumption. This makes the series a perfect fit for both performance-intensive and power-sensitive applications, enabling developers to create customized solutions tailored to their specific needs.
The Origin E1 is a compact yet powerful neural processing unit (NPU) designed for low-power applications in home appliances, smartphones, and security cameras. Specially tailored for always-on functions, the E1 offers unmatched power efficiency by utilizing Expedera's innovative packet-based architecture. This allows it to execute multiple layers in parallel, ensuring the optimal balance of performance and resource utilization. Engineered to operate with minimal to no external memory, the Origin E1 is ideal for cost-effective and area-sensitive designs. The LittleNPU processor within it is fine-tuned for sophisticated neural networks needed in always-sensing applications, maintaining privacy by keeping data on-board. This makes the E1 a robust choice for devices requiring dedicated AI processing without the penalty of high power consumption or large silicon area. In terms of technical specifications, the Origin E1 boasts a performance efficiency of up to 18 TOPS per Watt, with the capability to run various network types such as CNNs, RNNs, and DNNs. Its adaptability allows clients to customize features to meet specific needs, guaranteeing efficient deployments across numerous devices. The E1 leverages a full TVM-based software stack for seamless integration, further endorsing its versatility and effectiveness.
The AON1020 belongs to the sophisticated AONSens™ Neural Network Series, cementing its role in both audio and multifunctional sensor applications. The core is delivered in a Verilog RTL format, allowing flexibility for ASIC or FPGA synthesis. Designed for efficient energy usage, the AON1020 is ideal for persistent voice command recognition, offering its capabilities with negligible power draw for always-on operations. Capable of beyond-the-basics voice control, this IP comprehensively tackles real-time responses to acoustic events and environmental cues. It seamlessly integrates with sensor systems for detecting human activities, making it suitable for contexts requiring sensors for activity updates, such as fitness monitoring or interactive toys. Maintaining pin-point accuracy under challenging sound conditions or user unpredictability, AON1020 is perfect for high-responsiveness scenarios. It supports the company's data-centric approach, embedding functions necessary for context detection and acoustic events, thus providing a holistic suite for user interaction and functional triggers in connected ecosystems.
The Neural Network Accelerator from Gyrus AI is designed to enhance edge computing capabilities through native graph processing. This innovative solution offers up to 30 TOPs/W, making it highly efficient by achieving 10-30 times lower clock-cycles than conventional processors. Its design supports low memory usage configurations, significantly reducing power consumption and enabling up to 20 times lower energy requirements. Designed for seamless integration, it supports diverse neural network models with superior utilization of over 80%, leading to enhanced performance and reduced die area by 10 to 8 times smaller than traditional designs.
The eSi-3264 stands out with its support for both 32/64-bit operations, including 64-bit fixed and floating-point SIMD (Single Instruction Multiple Data) DSP extensions. Engineered for applications mandating DSP functionality, it does so with minimal silicon footprint. Its comprehensive instruction set includes specialized commands for various tasks, bolstering its practicality across multiple sectors.
The memBrain™ Neuromorphic Memory is a groundbreaking technology tailored for AI applications. Designed to efficiently move AI processing from cloud environments to edge devices, this memory technology excels in its power efficiency, making it suitable for battery-powered, deeply embedded devices. This innovation is crucial for applications such as video and voice recognition, where low-latency processing at the edge is vital. memBrain™ technology supports the growing demand for AI at the edge, providing significant power savings and reducing the Bill of Materials (BOM) costs for deploying AI functionalities. As AI workloads proliferate, the energy-efficient characteristics of memBrain™ make it ideal for a variety of edge devices that require not only performance but also sustainability. Since its introduction in 2017, memBrain™ has been adopted by several companies looking to enhance their edge AI computing capabilities. Its neural network-inspired architecture allows it to process data in a manner akin to the human brain, enabling more efficient learning and potent AI inference directly within devices.
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.
The SEMIFIVE SoC Platform is a bespoke development environment designed to expedite the creation of custom silicon solutions by leveraging domain-specific architectures. It integrates a pre-verified IP pool, providing a robust foundation for applications requiring tailored performance and cost efficiencies. This platform significantly reduces development time and associated risks by offering a ready-to-use environment that includes silicon-proven design components. With its comprehensive set of features, the platform facilitates rapid prototyping and market deployment, employing a high degree of reusability in design and verification components. By decreasing non-recurring engineering (NRE) costs and enhancing design reliability, the SoC Platform ensures faster time-to-market, making it ideal for industries aiming for quick product turnarounds. The SoC Platform supports scalable integration with third-party IPs, allowing flexibility to meet diverse application needs. Its architecture includes support for multiple processors, memory interfaces, and connectivity solutions, providing a one-stop solution for industries across AI, IoT, and HPC domains, ensuring performance optimization and minimal risk.
The SEMIFIVE HPC Platform is crafted to meet the demanding requirements of high-performance computing applications. This platform encompasses a cohesive suite of silicon-proven IPs designed for intensive data-processing tasks across various industries. Utilizing advanced process nodes, the HPC Platform provides a scalable architecture supporting high-speed data transfer and computational efficiency, making it suitable for hyperscale and cloud server environments. Its sophisticated integration of memory technologies and processing units offers enhanced performance capabilities, tailored for high-volume data operations. This platform is not only optimized for performance but also for power efficiency, ensuring sustainable high-speed operations. Its design includes robust security features and compliance with stringent industry standards, positioning it as an essential asset for enterprises aiming to leverage HPC for competitive advantage.
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!
No credit card or payment details required.
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!