Shulou(Shulou.com)11/24 Report--

Intel injects AI power into advanced technology, providing open, scalable and reliable solutions to help customers win in the AI era

In the era of AI, this new era of global growth driven by chips is closely related to everyone. This is the "core economy", in which the AI-driven system is full of autonomy and initiative, it will be integrated into people's daily life, helping to complete a variety of mental and physical tasks.

At Intel's on Technology Innovation Conference, Intel demonstrated a rich set of technologies that "make AI ubiquitous", making AI more accessible in a variety of workloads, from the client and edge to the network and cloud. Some highlights include easy access to AI solutions in the cloud, the price-to-performance advantages of Intel data center AI accelerators, tens of millions of new Intel PC that support AI will ship in 2024, and tools to support reliable deployment of AI at the edge.

Accelerating AI innovation requires the development of a wide range of solutions with openness and reliability in mind. Intel's AI hardware and software portfolio-from CPU, GPU and Accelerators to the oneAPI programming model, OpenVINO developer kits and libraries that support the AI ecosystem-can provide customers with competitive, high-performance, open-standard solutions that support rapid scale deployment of AI.

More information: 2023 Intel on Technology Innovation Conference (press kit)

The fifth generation Intel Xeon processor

The fifth generation Intel Xeon processor

Intel developer cloud platform is fully online

Intel announced the full launch of the Intel ®developer Cloud platform, which provides developers with an easier way to test and deploy AI and scientific computing applications and solutions on the latest Intel CPU, GPU and AI accelerators. Developers can also use cutting-edge tools to achieve advanced AI and performance. The details are as follows:

The Intel developer cloud platform is based on advanced central processing unit (CPU), graphics processing unit (GPU) built specifically for AI, Intel ®Gaudi ®2 accelerator for deep learning, and open software and tools. The cloud platform development environment also authorizes developers to use the latest Intel hardware platforms, such as the fifth generation Intel ®Xeon ®extensible processors (codenamed Emerald Rapids) and the Intel ®data center GPU series Max 1100 and 1550, of which the fifth generation Intel ®Xeon ®extensible processors will be used in the Intel ®development cloud platform in the coming weeks and will be released on December 14.

Developers can use the Intel developer Cloud platform to build, test, and optimize AI and scientific computing applications. They can also run AI training, model optimization, and reasoning workloads from small to large to achieve high performance and efficiency. Intel developer Cloud platform is based on oneAPI, an open multi-architecture, multi-vendor programming model, which provides developers with hardware choices and gets rid of proprietary programming models to support accelerated computing, code reuse and meet portability requirements.

More information: Intel developer cloud platform

Intel Gaudi2 Accelerator

Strong performance and innovative practice of data Center products

Intel announced new AI performance upgrades and strong industry momentum for its data center and AI portfolio, including Intel Gaudi2 and Gaudi3 accelerators, fourth-generation Intel Xeon scalable processors, fifth-generation Intel Xeon scalable processors, and next-generation Xeon processors codenamed Sierra Forest and Granite Rapids, respectively. The details are as follows:

● Intel announced that a large AI supercomputer will be built entirely on Intel Xeon processors and 4000 Intel Gaudi2 accelerators, with Stability AI as its main customer.

● Dell Technologies is working with Intel to develop AI solutions to meet the diverse needs of customers in the AI development process. PowerEdge systems based on Intel Xeon scalable processors and Gaudi accelerators will support AI workloads ranging from large-scale training to basic reasoning.

● Aliyun reports that fourth-generation Xeon is a viable solution for real-time large language model (LLM) reasoning on its smart product model service platform (DashScope), with average model response times up to 3x 1 thanks to Xeon processors and other software optimizations with Intel Advanced Matrix extension (Intel AMX) accelerator built-in.

The ● Granite Rapids contains a performance core (P-cores) that provides leading AI performance, and its AI performance is expected to increase by 2-3 times compared to the fourth-generation highly scalable processors.

More information: Intel unveils the next generation of extremely strong processors with both excellent performance and efficient architecture

Glass substrate test unit

Intel Core Ultra processor

Intel Core Ultra processor drives the new AI experience

Intel will enter the AI PC era with the upcoming Intel Core Ultra processor codenamed Meteor Lake, which is equipped with Intel's first integrated neural network processor (NPU) to deliver energy-efficient AI acceleration and native reasoning experiences on PC. Intel confirmed that the new Core Ultra processor will be released on December 14. The details are as follows:

● Intel Core Ultra processors provide connection-independent, low-latency AI computing with greater data privacy.

● Core Ultra integrates NPU into a client chip for the first time. NPU is designed to achieve low power consumption and high-quality computing and bring a whole new PC experience. It is ideal for migrating workloads with high-quality and energy-efficient computing requirements from CPU, or workloads that typically run in the cloud due to lack of effective client computing power.

● Core Ultra is a turning point in Intel's client processor roadmap: it is the first processor designed with a client core using Foveros encapsulation technology. In addition to significant improvements in performance and power consumption in NPU and Intel 4 process processes, the processor brings standalone graphics-level performance by integrating Intel sharp graphics cards.

● Core Ultra's separate module architecture achieves a balance of performance and power consumption in AI-driven tasks:

○ GPU has performance parallelism and throughput, making it ideal for AI computing in media, 3D applications, and rendering pipelines.

○ NPU is a dedicated, low-power AI engine for continuous AI and AI offloads.

○ CPU can respond quickly and is ideal for lightweight, single-inference, low-latency AI tasks.

● Intel highlighted its partnership with Acer to introduce AI into its upcoming Core Ultra system, demonstrating how the new Acer Parallax software feature uses NPU to add a 3D look to user images.

Overview of Intel AI portfolio in the overall AI workload

Support for AI at the Ed

Due to the growing demand for system automation and data analysis through AI, there are huge opportunities for edge computing. OpenVINO is Intel's AI reasoning and deployment runtime kit and is the first choice for developers on client and edge platforms. Through the OpenVINO developer toolkit, Intel is making access to edge AI easier. In the past year alone, downloads by developers of OpenVINO toolkits have increased by 90 per cent year-on-year. The details are as follows:

Version 2023.1 of ● OpenVINO, supported by oneAPI, makes it easier to deploy generative AI to real-world scenarios and enables developers to write once and deploy in a wide range of devices and AI applications.

The latest version of ● can be downloaded from OpenVINO.ai, allowing Intel to further realize its vision of deploying any model on any hardware.

● OpenVINO version 2023.1 enables developers to optimize standard PyTorch, TensorFlow, or ONNX models and provide full support for the upcoming Core Ultra processor. It also provides more model compression technology, improved GPU support and memory consumption for dynamic input, as well as higher portability and better performance across the entire computing continuum (across the cloud, client, and edge).

In ● 's keynote speech on the first day of Intel's on Technology Innovation Conference, Intel demonstrated Fit:match, an AI solution that revolutionizes today's retail and health experiences. Fit:match 's 3D concierge experience uses Intel RealSense with lidar sensors, Intel Core processors and OpenVINO depth cameras. The solution focuses on reliability and privacy and can scan and match thousands of products to ensure that they are more suitable for customers, thereby increasing the purchase conversion rate and reducing the return rate.

Forward-looking statement (Forward-Looking Statements)

This release contains forward-looking statements, including with respect to Intel's business plans and strategy, process and product roadmaps, and current and future technologies, as well as the anticipated benefits therefrom. Such statements involve many risks and uncertainties that could cause our actual results to differ materially from those expressed or implied, including: changes in demand for our products; changes in product mix; the complexity and fixed cost nature of our manufacturing operations; the high level of competition and rapid technological change in our industry; the significant upfront investments in rotated and our business, products, technologies, and manufacturing capabilities; vulnerability to new product development and manufacturing-related risks, including product defects or errata, particularly as we develop next generation products and implement next generation process technologies Risks associated with a highly complex global supply chain, including from disruptions, delays, trade tensions, or shortages; sales-related risks, including customer concentration and the use of distributors and other third parties; potential security vulnerabilities in our products; cybersecurity and privacy risks; investment and transaction risk; intellectual property risks and risks associated with litigation and regulatory proceedings; evolving regulatory and legal requirements across many jurisdictions; geopolitical and international trade conditions; our debt obligations; risks of large scale global operations; macroeconomic conditions; impacts of the COVID 19 or similar such pandemic And other risks and uncertainties described in our earnings release dated July 27, 2023, our most recent Annual Report on Form 10 Murk and our other filings with the Securities and Exchange Commission. All information in this press release reflects Intel management views as of the date hereof unless an earlier date is specified. Intel does not undertake, and expressly disclaims any duty, to update such statements, whether as a result of new information, new developments, or otherwise, except to the extent that disclosure may be required by law.

Intel does not control or audit third-party data. You can consult other sources to assess accuracy.

2 based on the architecture forecast for the fourth generation Intel Xeon processors as of August 21, 2023. The results may be different.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.