“Imagination’s automotive products are seamlessly designed for the performance, flexibility and safety needs of future vehicles,” says Aish Dubey, VP and GM for High Performance Computing SoC Business Division at Renesas. “With equally strong expertise in the smartphone market, our customers can greatly benefit from Imagination’s knowledge and technical capabilities. Android™ has emerged as an important feature for in-car infotainment systems, and Imagination already has in place the tools and partnerships that OEMs need to develop great automotive experiences on Android.”

The R-Car Gen 5 is a multi-domain architecture for centralised computing systems and software-defined vehicles.  IMG BXS offers R-Car customers a cost-effective way to handle multiple domains at the same time. Its hardware-based virtualisation solution, with fully secure domain isolation, freedom from interference and quality of service, allows for high performance GPU multitasking without compromising on safety. It is certifiable up to ASIL-B and offers multi-core support using advanced chiplet integration technology.

“Renesas is one of the world’s foremost automotive chip designers and the R-Car Gen 5 is an impressive solution for in-vehicle entertainment, software-defined vehicle applications and autonomy,” says Jake Kochnowicz, Chief Revenue Officer, Imagination. “The long-standing collaboration between Imagination and Renesas, culminating in the use of IMG BXS in the R-Car Gen5, is set to give automotive manufacturers a versatile computing platform with which they can design the next generation of driving experiences.”

For more information on how IMG BXS GPU and Imagination’s partnership with Renesas, read this supporting blog or visit the Imagination Booth at Embedded World, Stand 4-659. 

The post Imagination GPU Powers Renesas R-Car Gen 5 SoC appeared first on Imagination.

“The global smartphone market is experiencing a resurgence, propelled by cutting-edge AI features such as personal agents and enhanced photography,” says Peter Richardson, Partner & VP at Counterpoint Research. “However, the success of this AI-driven revolution hinges on maintaining the high standards users expect: smooth interfaces, sleek designs, and all-day battery life. As the market matures, consumers are gravitating towards premium devices that seamlessly integrate these advanced AI capabilities without compromising on essential smartphone qualities.” 

Offering up to 64 GPixel/s, 2 TFLOPS FP32 and 8 TOPS INT8 in an ultra-parallel compute engine running at 1GHz, DXTP is available in two off-the-shelf configurations and has already been licensed for use in mobile and automotive. In addition to performance and efficiency optimisations, DXTP is highly flexible, featuring fully secure GPU multitasking (via Imagination’s low-overhead, hardware-based virtualisation technology) and the ability to run graphics and compute operations at the same time with minimal overhead. 

“Zelos is integrating the advanced, power-efficient compute of Imagination DXTP into our upcoming chip,” says Cheng Chen, Technical Director at Zelos. “Imagination’s GPUs combine the performance of ultra-parallel processing with the flexibility that comes from a highly programmable architecture, making them the ideal platform for accelerating our AI models.” 

DXTP is supported by Imagination’s well-established software ecosystem and award-winning SDK and tools.  It ships with highly optimised OpenCL™ compute libraries that boost GPU utilisation for popular AI tasks, as well as reference kits for oneAPI and TensorGraph that accelerate the porting of existing code onto Imagination-based hardware. Optimised LiteRT support for Imagination GPUs will enable high performance AI on Android™.  App developers have access to the PowerVR developer tools for low-level performance analysis, debugging, trace capturing, as well as expert support through the Imagination developer forum. 

“Imagination DXTP is a real example of many small steps achieving big gains and of the resourcefulness of Imagination’s engineering teams,” says James Chapman, Chief Product Officer, Imagination.  “The set of performance and efficiency improvements inside DXTP will enable future smartphones to run the next wave of gaming and AI applications with even lower power consumption than before.”   

DXTP is the final product of Imagination’s D-Series GPU family which includes the high-performance IMG DXD with DirectX FL11_0 for the desktop market and the innovative IMG DXS for the safety-critical automotive market. 

The post Imagination takes efficiency up a level with latest D-Series GPU IP   appeared first on Imagination.

“全球智能手机市场正在复苏，推动力来自于尖端的AI功能，如个人助手和增强摄影，”Counterpoint Research的合伙人兼副总裁Peter Richardson表示。“然而，这场由AI驱动的革命的成功，依赖于保持用户期望的高标准：流畅的界面、精致的设计和全天候的电池续航。随着市场的成熟，消费者更倾向于选择能够无缝整合这些先进AI功能且不妥协手机基本性能的高端设备。”

DXTP提供高达64 GPixel/s的图形处理能力，2 TFLOPS的FP32性能和8 TOPS的INT8 AI性能，采用超并行计算引擎，工作频率为1GHz。DXTP提供两种现成的配置，且已获得在移动和汽车领域的授权使用。除了性能和效率优化，DXTP还具有高度灵活性，支持完全安全的GPU多任务处理（通过Imagination低开销的硬件虚拟化技术），能够同时运行图形和计算任务。

DXTP由Imagination成熟的软件生态系统和屡获殊荣的SDK及工具提供支持。它配备了高度优化的OpenCL™计算库，能够提高GPU在常见AI任务中的利用率，同时提供oneAPI和TensorGraph的参考套件，加速现有代码在Imagination硬件上的移植。优化的LiteRT支持将为Imagination GPU在Android™平台上的高性能AI提供支持。应用开发者还可以使用PowerVR开发者工具进行底层性能分析、调试、追踪捕获，并通过Imagination开发者论坛获得专家支持。

“Imagination DXTP是许多小步伐带来大进步的真实例证，也是Imagination工程团队聪明才智的体现，”Imagination首席产品官James Chapman表示。“DXTP内一系列性能和效率的改进将使未来的智能手机能够以比以往更低的功耗运行下一波游戏和AI应用。”

DXTP是Imagination D系列GPU家族的最终产品，D系列还包括面向桌面市场的高性能IMG DXD（支持DirectX FL11_0）和面向安全关键汽车市场的创新IMG DXS。

The post Imagination通过最新的D系列GPU IP将效率提升至新高度 appeared first on Imagination.

In addition to delivering outstanding performance for graphics and compute workloads, IMG DXS introduces a new, patented, zero-compromise approach to meeting ASIL-B requirements. Its innovation in the field of distributed functional safety means that automotive SoC designers no longer need to lay down extra silicon, repeat workloads, or implement costly and convoluted software testing to check for processing faults.   

“This shortlisting is a testimonial to the incredible engineers in Imagination who brought this product to life,” says James Chapman, Chief Product Officer, Imagination Technologies. “At a time when the car industry needs to increase the amount of compute available in vehicles quickly yet responsibly, performance-driven functional safety solutions like IMG DXS are a true game-changer.” 

Two other companies are shortlisted for this award: Eatron Technologies and Oxford RF.  The winner will be announced at the TechWorks Gala Dinner and Awards Ceremony in London on Thursday 4 December. 

More information on Imagination DXS GPU IP is available on the Imagination website along with a detailed white paper outlining its innovation in the field of distributed functional safety. 

The post Imagination DXS GPU IP recognised as game-changer for the car industry appeared first on Imagination.

All functions that a vehicle performs are graded based on potential risk according to the ISO 26262 Automotive Safety Integrity Level (ASIL). The ASIL-B standard requires processors to have diagnostic coverage of greater than 90% for single point faults.  

The Imagination DXS GPU introduced an innovation in the field of distributed functional safety that eliminates the PPA (power, performance, area) overhead of meeting ASIL-B functional safety standards. It features a patented mechanism that takes advantage of the inherent parallelism of today’s processors and the fact that no thread is ever fully utilised to identify faults within the timeframe set by the ASIL-B standard without the need to repeat workloads or duplicate silicon area. More information is available in our “Innovations in Distributed Safety” white paper.

“Imagination GPUs are widely deployed in the automotive market and expanding their presence beyond in-vehicle infotainment into the more safety-critical advanced driver assistance systems,” says James Chapman, Chief Product Officer at Imagination. “Independent certification that the Imagination DXS GPU meets the >90% diagnostic coverage expected of ISO 26262 ASIL-B components provides the necessary reassurance to safety-critical system integrators that its novel distributed safety technique will meet their requirements. We expect to see a lot of demand for this product, not just in automotive but also in other safety conscious markets.”

“Congratulations to Imagination Technologies for achieving ASIL-B certification for their latest automotive processor.  Throughout the assessment and audit process, Imagination demonstrated a culture that values safety, as well as suitably rigorous product design and development processes,” says Functional Safety Product Manager Liu Min SGS-TÜV Saar. 

The specific configuration certified was the IMG DXS-8-256 GPU. Imagination is planning to submit all Imagination DXS GPU configurations for independent certification over the coming months.

The post Imagination DXS GPU officially certified as ASIL-B compliant appeared first on Imagination.

The Imagination DXS GPU supports the graphics and compute requirements of an entire vehicle line-up with single core configurations starting at 0.25 TFLOPS and scaling up to 1.5 TFLOPS – 50% higher peak performance than Imagination’s previous generation automotive GPU IP. Imagination’s unique multi-core technology allows for two, three and four core configurations, while the low bandwidth bus between cores and support for isolation suits the use of IMG DXS in chiplets.  It adopts relevant features introduced into the PowerVR architecture in D-Series, such as Pipelined Data Masters and 2D Dual-Rate Texturing, to boost performance efficiency by 20% over previous generations of automotive GPU IP and ensure that it outperforms competitive cores on real world benchmarks.  

“Imagination is a leading provider of GPU IP into automotive,” says  Anshel Sag, Principal Analyst at Moor Insights & Strategy. “The demand for pixel processing in cars has grown enormously over the last decade, with digital cockpits, dashboard wide displays and rear entertainment screens offering smartphone-like visual experiences.  Imagination’s combination of great performance for 3D graphics, flexible multitasking and a robust approach to safety has seen its popularity explode.” 

DXS is optimised for the compute workloads needed for next generation vehicle intelligence, such as computer vision and data processing. By combining the raw performance of DXS with an additional FP16 pipeline and a set of new compute libraries and AI toolkits, users can achieve up to ten times higher performance for compute workloads compared to previous generations.   

“The GPU is no longer just used for graphics in cars,” says James Chapman, Chief Product Officer at Imagination. “The long-term nature of designing hardware in the era of the software-defined vehicles is making hardware designers choose flexible and programmable processors that can handle the AI workloads of today as well as adapt to whatever the future brings. Today, our automotive GPUs are as much in demand for their compute abilities as for their rendering.”   

The first Imagination processor to feature “Distributed Safety Mechanisms”   

All functions that a vehicle performs are graded based on potential risk according to the ISO 26262 Automotive Safety Integrity Level (ASIL). Automotive manufacturers and their suppliers must ensure that components involved in that function meet the necessary standards for fault detection and handling, with ASIL-A having the least stringent requirements and ASIL-D the most comprehensive.   

Its new “Distributed Safety Mechanisms” solution means DXS achieves ASIL-B functional safety at a fraction of the Performance, Power or Area (PPA) overhead of the two main existing methods: dual-core lockstep, which increases silicon area by 100%, and workload repetition, which halves processor performance for safety workloads.    

Distributed Safety Mechanisms has a near-zero impact on GPU performance and minimal area cost, estimated at just 10%.  It does this by taking advantage of the inherent parallelism of today’s processors and the fact that no thread is ever fully utilised. A patented mechanism combines these threads into pairs and injects safety tests in idle moments to identify faults within the timeframe set by the ASIL standard.  Further details on the complete Distributed Safety Mechanisms solution can be found on the Imagination website in the white paper, Innovations in Distributed Functional Safety.

New Compute Libraries and AI Toolkits   

Developed alongside the hardware is a new set of compute libraries (imgBLAS, imgNN, imgFFT) to help software developers achieve up to 80% GPU utilisation.  This will also accelerate common compute workloads such as computer vision and pre/post processing perception data for ADAS systems.   

The libraries combine with new reference toolkits (oneAPI and TVM) to create a practical software stack based on open standards with which software developers can easily port their compute applications onto Imagination IP based hardware, and from there maximise their performance.  

“Software portability and the ability to take advantage of heterogeneous hardware is essential to the future of the automotive industry. The UXL Foundation is looking to provide a clear and easy path for developers to do this through the ongoing open source development of the oneAPI standard,” says Rod Burns, Chair of the UXL Foundation Steering Committee. “Imagination is a long-standing supporter of our ecosystem and the compute software solutions they release today reinforce their commitment to building an optimal developer journey based on open standards.” 

“PerfXLab Technologies develops heterogeneous computing software stacks and infrastructure solutions for companies looking to accelerate AI”, says Zhang Xianyi, Chief Executive Office at PerfXLab. “We are using Imagination’s compute software solutions to run various AI applications, including our LLM inference engine, PerfXLM, on Imagination GPUs and have so far achieved performance gains of up to 100% compared to the CPU, with minimal time spent porting.” 

Software   

Imagination’s automotive GPUs support OpenGL® ES, Vulkan®, OpenGL® and OpenCL™. They run popular automotive operating systems such as QNX and Green Hills Software’s INTEGRITY RTOS, as well as Linux and Android. Hardware-based virtualisation enables Imagination GPUs to run up to eight operating systems simultaneously, with full memory isolation for fully secure GPU multitasking. Imagination partners with CoreAVI, the global leader in architecting and delivering safety critical graphics and compute software drivers and libraries, on the development of a safety critical driver, supporting OpenGL SC and Vulkan SC for Imagination GPUs.    

“As vehicles become more autonomous, greater emphasis is placed on their ability to detect and handle faults in an appropriate manner,” says Dan Joncas, Chief Marketing and Sales Officer & Deputy CEO, CoreAVI. “By working with us, Imagination is able to meet the complex requirements of the automotive industry with a complete safety package covering both hardware and software.”  

For more information on the Imagination DXS GPU visit our dedicated page.   

The post Imagination Announces Highest Performance Automotive GPU IP with FuSa Advancement appeared first on Imagination.

Imagination DXS GPU可支持一辆汽车的所有图形和计算要求，可从算力为0.25 TFLOPS的单核配置，扩展到1.5 TFLOPS的处理能力，这意味着该处理器的峰值性能比Imagination上一代汽车GPU IP提高了50%。Imagination独特的多核技术可实现两核、三核和四核配置，而内核之间的低带宽总线和对隔离的支持，也使IMG DXS可适用于多晶粒芯片（chiplet）形态。它采用了D系列PowerVR架构中引入的相关功能，例如Pipelined Data Masters和2D双速率纹理（2D Dual-Rate Texturing），使性能效率比前几代汽车GPU IP提高了20%，并确保其在实际基准测试中的表现优于竞品IP内核。  

“Moor Insights & Strategy的首席分析师Anshel Sag表示：“Imagination是汽车领域领先的GPU IP供应商。过去十年中，汽车中对像素处理的需求大幅增长，数字驾驶舱、全宽仪表盘显示屏和后排娱乐屏幕提供了类似智能手机的视觉体验。Imagination在3D图形性能、灵活的多任务处理和强大的安全性方面的出色表现使其受欢迎程度迅速攀升。”

DXS针对下一代汽车智能所需的计算工作负载进行了优化，例如计算机视觉和数据处理等。通过将DXS的原始性能与额外的FP16通道以及一套新的计算库和人工智能（AI）工具包相结合，用户可以实现比前几代产品高出十倍的计算工作负载性能。 

“GPU在汽车中的角色不再仅限于图形处理。”Imagination首席产品官James Chapman表示，“在软件定义汽车时代中，硬件设计需要具有长期属性，这促使硬件设计人员去选择灵活的、可编程的处理器；这些处理器不仅可以处理当前的人工智能工作负载，而且还能适应未来出现的新需求。如今，我们的汽车GPU不仅因其渲染能力而广受欢迎，同时也因其计算能力备受青睐。” 

Imagination第一款带有“分布式安全机制”的处理器

根据ISO 26262汽车安全完整性等级（ASIL）体系，汽车上运行的所有功能都要根据潜在风险进行等级认证。汽车制造商及其供应商必须确保支持该功能实现的组件都符合故障检测和处理的必要标准，其中ASIL-A等级的要求最低，ASIL-D等级的要求最全面。 

由于采用了全新的“分布式安全机制”解决方案，这意味着DXS实现ASIL-B功能安全在性能、功耗或面积（PPA）的开销仅为现有两种主流方法的一小部分：在执行安全任务时，主流的双核锁步方案需要硅片面积增加一倍；主流重复执行的方案会使处理器性能减半。

分布式安全机制对GPU性能的影响几乎为零，而面积代价极低，估计仅为10%。 它利用了当今处理器固有的并行性，以及任何线程都不会被完全利用的事实。获得专利的机制将这些线程组合成对，并在空闲时注入安全测试，从而在ASIL标准设定的时间范围内识别故障。有关完整的分布式安全机制解决方案的更多详细信息，请参阅Imagination网站上的分布式功能安全的创新与突破白皮书。 

全新计算库和人工智能工具包

与硬件同时开发的还有一套新的计算库（imgBLAS、imgNN、imgFFT），它们可帮助软件开发人员实现高达80%的GPU利用率。这也将加速常见计算工作负载的处理，如计算机视觉和ADAS系统感知数据的预处理和后处理。 

这些库与新的参考工具包（oneAPI和TVM）相结合，创建了一个基于开放标准的实用软件栈，软件开发人员可以利用该软件栈轻松地将其计算应用移植到基于Imagination IP的硬件上，并最大限度地提高其性能。 

“软件的可移植性和利用异构硬件的能力对汽车行业的未来至关重要。UXL基金会希望通过开源方式持续开发oneAPI标准，为开发人员实现这一目标提供清晰而便捷的途径。”UXL基金会指导委员会主席Rod Burns表示。“Imagination长期以来一直支持我们生态系统，他们今天发布的计算软件解决方案强化了他们基于开放标准，来为开发者构建最佳路径的承诺。”

“澎峰科技（PerfXLab Technologies）为希望加速人工智能应用的公司开发异构计算软件栈和基础设施解决方案。”澎峰科技首席执行官张先轶表示。”我们正在使用Imagination的计算软件解决方案，在Imagination GPU上运行各种人工智能应用，包括我们的LLM推理引擎PerfXLM。与CPU方案相比，我们到目前为止已实现的性能提升就高达100%，而且只需花费很少的时间来进行移植。”

软件

Imagination的汽车GPU支持OpenGL® ES、Vulkan®、OpenGL®和OpenCL™。它们可运行流行的汽车操作系统，如QNX和Green Hills软件公司的INTEGRITY RTOS，以及Linux和Android。基于硬件的虚拟化技术使Imagination GPU能够同时运行多达八个操作系统，并通过完全的内存隔离，实现完全安全的GPU多任务处理。Imagination与全球领先的安全关键型图形和计算软件驱动程序和库构建者和提供商CoreAVI合作开发安全关键型驱动程序，从而为Imagination GPU提供OpenGL SC和Vulkan SC支持。

“随着汽车的自主化程度越来越高，人们更加重视汽车能够以适当的方式去检测和处理故障的能力。”CoreAVI首席市场和销售官兼副首席执行官Dan Joncas表示。“通过与我们合作，Imagination能够通过全面覆盖了硬件和软件的完整安全性工具包，来满足汽车行业的复杂要求。”

有关Imagination DXS GPU的更多信息，请访问Imagination网站。 

The post Imagination推出性能最高且具有高等级功能安全性的汽车GPU IP  appeared first on Imagination.

Cybersecurity vulnerabilities are a reality of technology development. Responsible vendors have processes in place throughout the product lifecycle, from development to mass market shipping, to minimise the volume and impact of such vulnerabilities. 

The CVE Program is a collaborative initiative that allows authorised partners from around the world to identify, define and catalogue cybersecurity vulnerabilities within one publicly available database. This single point of information helps the global technology ecosystem communicate more easily around vulnerabilities, removing any potential confusion about which issue is under discussion. 

“Within Imagination, we commit to the highest standards of product security and foster a culture of transparency and collaboration,” says Tim Mamtora, Chief of Innovation and Engineering. “Our engineering teams work closely with customers, partners, security researchers and the wider technology community to handle vulnerabilities in a responsible manner, build trust and contribute to the advancement of global cybersecurity standards.” 

Becoming an authorised CNA demonstrates our commitment to cybersecurity and highlights the maturity of our vulnerability management practices.  For our customers and partners, it will streamline our vulnerability disclosure process, ease communication and reduce the time required to mitigate issues. 

For more information, you can view Imagination’s cybersecurity and vulnerability reporting policy here.

The post Imagination Becomes a CVE Numbering Authority  appeared first on Imagination.

Over the past 30 years Imagination has consistently pushed the boundaries of what is possible by working with partners to develop patent protected GPU, CPU, AI and Ray Tracing semiconductor IP. Today, Imagination’s IP is present in over 13 billion devices around the globe, in a wide range of industries including automotive, data centre and desktop, mobile, and consumer and IoT.

Imagination will use the financing to support its ambitious objectives for growth, which reflect the centrality of Imagination’s innovative solutions to Graphics, Compute and AI at the edge – markets that are expected to grow significantly in the coming years. As the complexity of semiconductor design and associated development costs continue to increase, the semiconductor IP market is expected to become an $11 billion revenue opportunity by 2026, a 15% CAGR since 2020. By 2030, the total artificial intelligence (AI) semiconductor market is expected to grow to more than $1 trillion, a 32% CAGR from 2025.

Imagination’s latest products have set new industry benchmarks, further reinforcing the company’s position of leadership in some of the fastest growing technology segments.  The recently unveiled APXM-6200 CPU, a performance-dense RISC-V application processor with artificial intelligence, is designed to support the compute and intuitive user experience needs for next generation consumer and industrial devices. And the recently announced IMG DXD, a scalable GPU IP with support for DirectX®, is tailor-made for desktop, laptop and cloud gaming graphics experiences and has already been licensed for use in the desktop market.

Ray Bingham, Executive Chair of Imagination Technologies, says:

“We are thrilled that Fortress has chosen to invest in Imagination.  This investment comes with a shared commitment to drive Imagination’s innovation in GPU, CPU and AI, and to help us accelerate growth, drive significant advancements, and deliver unparalleled IP value to customers.”

Eran Zur, Head of Intellectual Property at Fortress, says:

“We are deeply familiar with Imagination’s team, the company’s extraordinary track record of innovation, and its world-class patent portfolio. Imagination is exceptionally well-positioned to capture a greater share of the growth in technology, including in AI, given the burgeoning demand for ever-more efficient compute solutions across the global economy. We look forward to bringing Fortress’s expertise in intellectual property, technology and growth investing to help Imagination continue expanding its leadership in the industry, develop new capabilities, and maximize the value of its portfolio.”

Barclays and Jefferies acted as joint placement agents to the Company in connection with the transaction.

Contacts:

Imagination Technologies – press@imgtec.com

Fortress Investment Group LLC – media@fortress.com

About Imagination Technologies

Imagination is a UK-based company that creates silicon and software IP designed to give its customers an edge in competitive global technology markets. Its GPU, CPU, and AI technologies enable outstanding power, performance, and area (PPA), fast time-to-market, and lower total cost of ownership. Products based on Imagination IP are used by billions of people across the globe in their smartphones, cars, homes, and workplaces.

Visit https://www.imaginationtech.com/ to learn more.

About Fortress Investment Group

Fortress Investment Group LLC is a leading, highly diversified global investment manager. Founded in 1998, Fortress manages $49.0 billion of assets under management as of March 31, 2024, on behalf of approximately 2,000 institutional clients and private investors worldwide across a range of credit and real estate, private equity and permanent capital investment strategies.

The post Imagination Technologies announces new capital investment from Fortress Investment Group appeared first on Imagination.

“In the AI era, with vast opportunities to reshape thousands of industries, ESWIN Computing is building the future of intelligent computing based on RISC-V,” stated Bo Wang, the Vice Chairman of ESWIN Computing. “Computing power is the driving force of AI and we designed the EIC77 Series SoCs to meet the diverse computational requirements of various application scenarios. By combining powerful RISC-V processors from SiFive with a compute-focused GPU from Imagination and our independently developed high-performance NPU, ESWIN Computing have developed a series of outstanding products in the fields of edge computing, AI PC, AI acceleration, and other fields.” 

The EIC77 Series was first unveiled in early 2024, encompassing the single-die RISC-V edge computing SoCs EIC7700 and its higher computing power version EIC7700X, as well as the dual-die RISC-V AI PC ICs EIC7702 and its higher computing power version EIC7702X. EIC7700X will feature inside a development platform that is coming to the market in August 2024, SiFive’s HiFive Premier P550 Development Board.  Specifically, the EIC7700X features: 

• A quad-core SiFive Performance P550 core complex running at 1.4GHz+ (up to 1.8GHz) with a thirteen-stage, triple-issue, out-of-order pipeline based on the RISC-V RV64GBC ISA. 
• A high efficiency IMG A-Series GPU that provides the optimum balance of fill rate and compute in a compact silicon area.  Its 128-wide superscalar ALU (arithmetic logic unit) with dedicated AI pipelines deliver up to 0.25 TFLOPS, 1 TOPS and 8 Gpixels performance. 
• An independently developed NPU accelerator from ESWIN Computing that delivers 19.95 INT8 TOPS, 9.975 INT16 TOPS and 9.975 FTPS FP16 compute performance. 
• High utilization of computing units for DNN inference, up to 590fps ResNet50 and up to 2600fps MobileNetv2. 
• Up to 8K@25fps video encoding and 8K@50fps video decoding.   
• High memory capacity, up to 32GB 64-bit LPDDR4/4X/5 

“It is great to see our highest performance CPU cores shipping in premium products with partners like ESWIN, including the new SiFive HiFive Premier P550 development platform.  RISC-V is transforming the AI landscape with products that deliver superior compute density for modern workloads,” says Jack Kang, SiFive’s Senior Vice President of WW Business Development, Sales and Customer Experience (CX). “Tight integration with domain specific accelerators, such as a GPU and NPU, is key to maximising SoC performance. SiFive has long-standing partnerships with major IP vendors such as Imagination to make sure that our mutual customers get to market quickly with winning products.”  

“Combining a RISC-V CPU with a compute-ready GPU and NPU creates an incredibly flexible and capable AI platform,” says James Chapman, Chief Product Officer, Imagination. “Imagination gives system designers the best choice of RISC-V ready GPU IP and our engineering team’s invaluable experience of RISC-V system design minimises complexity for customers building a multi-vendor, heterogeneous SoC.” 

The EIC7700X is on show at the RISC-V Summit Europe in Munich from Tuesday 25 June to Thursday 27 June. Visit the ESWIN Computing booth to learn more. 

The post ESWIN Computing Pairs SiFive CPU, Imagination GPU and In House NPU in Latest RISC-V Edge Computing SoC appeared first on Imagination.