AI optimization and acceleration so far has mostly been driven by novel hardware innovations. However, recently there have been new products developed by the AI start-up and research communities that operate at the software level in the machine learning technology stack, which reduce training length, inference time, energy consumption and memory usage of AI and machine learning (ML) models.Learning how to benchmark, deploy, and use these tools, as well as scrutinizing existing real-world use-cases is business-critical because these tools disrupt the AI market by: Enhancing the running of ML models on CPUs, therefore, reducing the costs of ownership of AI hardwareImproving the performance of standard AI accelerators (GPUs, FPGAs, DSPs etc.)Helping more data science and machine learning projects get into full productionIncreasing the competitiveness of products in the crowded edge device and IoT marketsThe Software for AI Optimization Summit will demonstrate how to deploy and integrate the latest novel software optimization tools into AI/ML workflows. Consequently, innovators and adopters in data science, machine learning, AI engineering and IoT solutions will uncover new techniques that enable the training and inference of AI and ML models to be faster, more cost-efficient, and still maintain high levels of performance and accuracy.Consequently, you will be ahead of the latest innovations across the ML technology stack enabling you to make informed decisions about strategic partnerships on who to partner with to optimize AI on the cloud and at the edge.

The AI Systems Summit, Research brings together researchers, engineers and architects at research/public institutions worldwide that will focus on the design and deployment of AI systems for scientific purposes. The event is a miniature virtual conference that is free to attend for registrants from research institutions & national laboratories.Attend to meet researchers working on the same challenges as you in an hour of dedicated networking time on each day. There will also be an opportunity to meet and hear from some of the leading AI Hardware technology providers . On the HPC side, the summit will cover topics such as the convergence of HPC & AI (University of Stuttgart), heterogeneous systems design (Lawrence Livermore National Lab), the application of AI hardware to non-AI HPC workloads (Simula Research Laboratory) and early impressions of the Cerebras CS-1 system and the advent of 3D monolithic SoCs (US National Energy Technology Lab).On the Edge device side, the summit will look at design & engineering challenges for robotic systems (NASA) and neuromorphic and brain inspired systems (UTSA and Oak Ridge National Lab).

After three successful events (two in the US, one in China) we’re delighted to be bringing the AI Hardware Summit to Munich. We’ve been honoured with previous keynote speakers such as Dr. John L. Hennessey, Chairman of Alphabet, Lip-Bu Tan, CEO of Cadence Systems and Naveen Rao, CVP & GM of Intel’s AIPG. Europe is struggling to keep up with developments in machine learning at the same pace as the US and China, where many of the top labs are located and investments are significantly larger. Due to its academic strength in AI, the potential for Europe to catch up in AI research is massive and could add enormous value to its combined economic output. Institutes like the ELLIS Society are working closely with top European academics to enable Europe to perform the best fundamental AI research, but there is still a large gap in GDP investment when compared to the US and China. Given Europe’s strength in automation across industrial, robotics and autonomous vehicles manufacturing, this summit will focus on inference systems in edge computing, connecting these industries through common challenges in systems architecture/engineering. The AI Hardware Summit Europe will bring together key players across the AI hardware ecosystem to promote innovation and adoption of systems & silicon for processing AI.

This past year, the cloud gaming industry has become a vibrant marketplace where influential partnerships are happening - such as the Microsoft alliance with Sony and Google with BT - amplifying the opportunity that cloud gaming represents for the gaming industry and highlighting its growth potential. But the industry faces some significant technological challenges, as ensuring the continuity and quality of the gaming experience is crucial, in order to build an audience and a trustworthy reputation for cloud gaming platforms and products.The Cloud Gaming Congress Webinar Series will congregate the critical mass of the emerging cloud gaming industry, to define game streaming business models, overcome technological hurdles in distribution and create partnerships between content creators & platforms, and gaming companies & tech providers.The series will navigate between technical and business-related topics, critiquing current service offerings, evaluating consumer behaviors, and extracting key insights for smart business decisions and stronger business models. Telco & network experts will discuss leveraging 5G to optimize the cloud gaming experience and assess strategies to overcome challenges such as latency and packet loss. Finally, we will discuss the future direction of the gaming industry, analyzing the relationship between console/PC and cloud gaming going forward.We will be exploring the expansion of content creation through cloud gaming as well as how the cloud will revolutionize the gaming industry by allowing game publishers & developers to close the gap with their customers, and by enabling a new wave of data insights for gaming companies.“What will be interesting is when you develop cloud to run natively onto servers and you take advantage of thoseresources whether it’s computer storage or into network creating a persistent world that can host millions of gamers at the same time continuously, that transformation is gonna be enabled over the next decade and it’s gonna change the world of gaming and cloud gaming is really just the beginning...” Marcus Kennedy, GM, Gaming Division, Client Computing Group, IntelRegister to attend the second webinar in the series: 

Take advantage of over 9 hours of dedicated networking time to meet key industry leaders, whilst also exploring:  Luminary Keynotes: Unique perspectives from industry luminaries in hardware (John Hennessy), AI (TBC) and investment (Lip-Bu Tan).Innovations and Optimizations of Silicon & Systems for AI Training & Inference: Presentations and product launches from C-level executives from AI chip start ups, semiconductor companies and systems OEMs.Training & Inference at Hyperscale & AI Accelerators in the Data Center Hardware Environment: Deployment & maintenance of AI infrastructure in data centers, hardware requirements for training & inference at scale.Inference in Client (Edge) Computing: Applications for AI accelerators in cameras, consumer electronics, autonomous vehicles etc.Beyond Compute: AI’s Impact on Memory, Storage & Networking: Innovations in HBM, on-chip memory and NVM, I/O bottlenecks, data transfer & high-speed interconnects.The Impact of Future ML Models on Hardware Design: Machine Learning co-design, robustness & reprogrammability, model standardization & interoperability.AI Chip Design & Commercialization: Design, testing & manufacturing, form factors & routes-to-market.Financial & Industrial Analysis: Market growth and maturity, VC investment trends & dynamics, the commoditization of the inference market, benchmarking & metrics.“Two overarching efforts are indispensable in this AI chip development frenzy: objectively evaluating and comparing different chips (benchmarking), and reliably projecting the growth paths of AI chips (road mapping).” White Paper on AI Chip Technologies: Tsing Hua University & Beijing Innovation Center for Future Chips, December 2018.

人工智能硬件峰会是第一个也是唯一一个专门探讨开发用于神经网络与计算机视觉的硬件加速器这一生态系统的行业峰会。参加峰会，与200多位来自人工智能芯片领域的新创企业、半导体公司的资深技术领导、系统供应商/原始设备制造商、数据中心、终端用户、金融服务机构、投资人及基金公司汇聚一堂，共同构建新兴人工智能芯片市场的总体综合路线图。利用独特的交流机会，与主要行业领袖进行1-2-1单独接触，就以下继分领域进行深度交流： 高性能芯片架构的创新与优化：GPU（图形处理器）、 ASIC（专用集成电路）& TPU、FPGA（…替代性方案：复合数字/模拟计算、神经形态学和神经忆阻系统（neuromemristive systems）与量子计算。软件的作用：辅助软件、软件框架、生态系统和应用。生成用户友好的原代码，并将TensorFlow&Caffe等开源机器学习框架链接起来。行业发展：创业与企业研发、风险资本与企业风险投资的投资趋势、战略性并购与合作。  各项预测预言，人工智能(芯片组)市场有望从2018年的70.6亿美元增长到2025年的592.6亿美元， 期间年复合增长率达到35.5%(Business Wire，2018)。因此，芯片市场将迎来停滞不前。我们正处于计算范式的转变期，毋庸置疑的是，目前正在开发的处理单元的结合将加快人工智能的研究与开发，而人工智能的性能将远远超过现有平台。。因此，对如何将这一趋势转化为机遇，如何规划未来，需要我们分享我们的认识与见解。您现在可以查看议程并在微信上付款 

The cloud has rapidly grown from being a market disrupter to an essential IT strategy for enterprise.  77% of businesses have at least one application or portion of their enterprise IT infrastructure in the cloud.  Migrating to the cloud is tough. Whether you are fully cloud-native, in a hybrid cloud environment or just taking the first steps, The Cloud Migration Summit will debunk migration myths and help IT Strategists and I&O leaders properly foster a corporate cloud culture and avoid migration pitfalls that might ramp up expenses bureaucracy and hamper digital transformation efforts. Discover how the cloud front runners and industry leaders accelerated their digital transformations through a series of dedicated case studies, panel discussions and round tables.83% of IT workloads will be in the cloud by 2020. Don't get left behind the curve.