VeriSilicon's Neural Network Processor IP Embedded in Over 100 AI Chips
50 customers licensed the technology for more than 100 AI chips in 10 major market segments
Shanghai, Chnia - Nov. 12, 2021 - VeriSilicon (688521.SH), a leading Silicon Platform as a Service (SiPaaS®) company, today announced its neural network processor (the Vivante1 NPU) IP designed for artificial intelligence (AI) applications now features in more than 100 AI chips supplied by 50 licensees. These chips with built-in VeriSilicon Vivante NPUs are in 10 major market segments, including Internet of Things (IoT), wearables, smart TVs, smart home, security monitoring, servers, automotive electronics, smartphones, tablets and smart healthcare.
VeriSilicon’s Vivante NPU is a high-performance computer vision and AI processor with scalable, programmable and low-power architecture. The NPU’s performance with a single convolutional kernel can range from 0.5 TOPs to 20 TOPS, however 500 TOPs can be reached after the extension of multi-convolutional computing kernels. This range of performance enables a wide array of applications ranging from ultra-low power wearables to high-performance computing in data centers. Licensees can customize the Vivante NPU to fit different chip sizes and power budgets, making it a cost-effective premium neural network acceleration engine. The complete software stack and SDK supports mainstream deep learning frameworks, including TensorFlow, PyTorch, ONNX, Apache TVM, and IREE. The Vivante NPU also provides proven solutions from model conversion, quantization, image optimization, and online/offline compilation, to the final deployment of devices to help customers quickly launch products.
As China’s leading semiconductor IP provider and the seventh-largest globally2, VeriSilicon has six categories of in-house processor technologies, which include the Vivante NPU, graphics processor (GPU), video processor (VPU), digital signal processor (DSP), image signal processor (ISP), and display processor. VeriSilicon’s Vivante NPU can be used with VeriSilicon's other proprietary processors to enhance images, audio, and other signals in real time to improve the user experience. These combinations are now used by leading TV manufacturers in their flagship products.
“Neural network processing technology can deeply interconnect and intelligently enhance other technologies such as ISP to create disruptive applications,” said Wei-Jin Dai, Executive VP and GM of IP Division at VeriSilicon. “Based on the successful Vivante NPU, we are focusing on building a comprehensive AI ecosystem (hardware and software) providing open-source standards and machine learning frameworks support. Our open-source TIM-VX (Tensor Interface Module) project has now been widely adopted enabling leading companies to build their intelligent ecosystems."
DC9000 adopts VeriSilicon’s latest display processor V9 architecture includes local tone mapping, edge enhancement, and 3D color lookup table and color engines. It includes key innovations to minimize power, bandwidth and silicon area by optimize memory access algorithm to implement the high quality display.
1 Vivante is a brand of VeriSilicon’s IP product line
2 Data source: IPnest Design IP Report, 2021
About VeriSilicon
VeriSilicon Microelectronics (Shanghai) Co., Ltd. (VeriSilicon, 688521.SH) is committed to providing customers with platform-based, all-round, one-stop custom silicon services and semiconductor IP licensing services leveraging its in-house semiconductor IP. Under the unique "Silicon Platform as a Service" (SiPaaS) business model, depending on the comprehensive IP portfolio, VeriSilicon can create silicon products from definition to test and package in a short period of time, and provides high performance and cost-efficient semiconductor alternative products for IDM, Fabless, system vendors (OEM/ODM) and large Internet companies, etc. VeriSilicon's business covers consumer electronics, automotive electronics, computer and peripheral, industry, data processing, Internet of Things and other applications. VeriSilicon presents a variety of customized silicon solutions, including high-definition video, high-definition audio and voice, In-Vehicle Infotainment, video surveillance, IoT connectivity, data center, etc. In addition, VeriSilicon has six types of in-house processor IPs, namely GPU IP, NPU IP, VPU IP, DSP IP, ISP IP and Display Processor IP, and more than 1,400 analog and mixed signal IPs and RF IPs. Founded in 2001 and headquartered in Shanghai, China, VeriSilicon has 6 design and R&D centers in China and the United States, as well as 11 sales and customer service offices worldwide. VeriSilicon currently has more than 1,300 employees.
|
Verisilicon, Inc. Hot IP
Related News
- VeriSilicon's Vivante VIP8000 Neural Network Processor IP Delivers Over 3 Tera MACs Per Second
- BrainChip Demonstrates Company's Event-Based AI Neural Processor at Embedded Vision Summit
- Inuitive Adopts VeriSilicon's Advanced ISP IP for its Vision AI Processor
- Dolphin Design wins an Embedded Award for Tiny Raptor, its Energy-Efficient Neural Network AI Accelerator
- DSP Group Unveils DBM10 Low-Power Edge AI/ML SoC with Dedicated Neural Network Inference Processor
Breaking News
- Ubitium Debuts First Universal RISC-V Processor to Enable AI at No Additional Cost, as It Raises $3.7M
- TSMC drives A16, 3D process technology
- Frontgrade Gaisler Unveils GR716B, a New Standard in Space-Grade Microcontrollers
- Blueshift Memory launches BlueFive processor, accelerating computation by up to 50 times and saving up to 65% energy
- Eliyan Ports Industry's Highest Performing PHY to Samsung Foundry SF4X Process Node, Achieving up to 40 Gbps Bandwidth at Unprecedented Power Levels with UCIe-Compliant Chiplet Interconnect Technology
Most Popular
- Cadence Unveils Arm-Based System Chiplet
- CXL Fabless Startup Panmnesia Secures Over $60M in Series A Funding, Aiming to Lead the CXL Switch Silicon Chip and CXL IP
- Esperanto Technologies and NEC Cooperate on Initiative to Advance Next Generation RISC-V Chips and Software Solutions for HPC
- Eliyan Ports Industry's Highest Performing PHY to Samsung Foundry SF4X Process Node, Achieving up to 40 Gbps Bandwidth at Unprecedented Power Levels with UCIe-Compliant Chiplet Interconnect Technology
- Arteris Selected by GigaDevice for Development in Next-Generation Automotive SoC With Enhanced FuSa Standards
E-mail This Article | Printer-Friendly Page |