Intel has announced the arrival of its 3rd generation Intel Xeon Scalable processors. The company also unveiled additions to its portfolio of hardware and software AI solutions, enabling customers to accelerate the development and use of AI and analytics in data centers, networks and intelligent end-to-end environments.
Intel’s new 3rd generation Xeon Scalable processors have been introduced as the first general purpose server processors with built-in bfloat16 support. They enable wider deployment of artificial intelligence inference and training for applications that include image classification, recommendation systems, speech recognition and language modeling.
“Today, the ability to rapidly deploy AI and data analysis is critical for enterprises. We continue to work to improve integrated AI acceleration and software optimizations within the processor that powers data centers and edge solutions around the world, as well as to provide an unmatched silicon foundation to unleash the power of data,” said Lisa Spelman, Intel vice president and general manager, Xeon and Memory Group.
As Intel indicates, AI and data analysis are opening up new opportunities for companies in a wide range of sectors, including finance, healthcare, manufacturing, telecommunications and transportation. IDC predicts that by 2021, 75 percent of enterprise business applications will use AI. And by 2025, IDC estimates that about one-quarter of all data generated will be created in real time, with the various devices of the Internet of Things accounting for 95% of this volume growth.
Intel’s new data platforms, coupled with a thriving ecosystem of partners using Intel AI technologies, are optimized to enable enterprises to monetize their data through the deployment of AI and analytics services.
Intel continues to invest in integrated AI acceleration in the new 3rd generation Intel Xeon Scalable processors by integrating bfloat16 support in Intel DL Boost processor technology. Bfloat16 is a compact digital format that uses half the bits of the current FP32 format, but achieves comparable model accuracy with minimal (if any) software modifications.
The addition of bfloat16 support speeds up both AI training and inference performance in the processor. Intel optimized distributions for major Deep Learning frameworks (including TensorFlow and Pytorch) support bfloat16 and are available in the Intel AI Analytics Toolkit. Intel also provides bfloat16 optimizations in its OpenVINO toolkit and in the ONNX runtime environment to facilitate inference deployments.
The 3rd generation Intel Xeon Scalable processors (codename “Cooper Lake”) are evolving Intel’s 4 and 8-socket processor offerings. The processor is designed for Deep Learning, Virtual Machine (VM) density, in-memory databases, mission-critical applications and intensive analytical workloads. Customers looking to modernize an aging infrastructure can expect 1.9 times better performance on average on the most popular workloads and 2.2 times more VMs compared to 4-socket platform equivalents that are 5 years old.
As part of the 3rd generation Intel Xeon Scalable Platform, Intel also announced the 200 Series Intel Optane Persistent Memory, which offers customers up to 4.5 TB of memory per socket to manage data-intensive workloads such as in-memory databases, dense virtualization, analytics and advanced computing.
For systems that store data in all-flash arrays, Intel has announced the availability of its next generation high-capacity 3D NAND SSDs, the Intel D7-P5500 and P5600 NAND 3D SSDs. These 3D NAND SSDs are built with Intel’s latest three-tier NAND 3D cell technology (TLC) and a brand new low-latency PCIe controller to meet the intense I/O needs of AI and analysis workloads, as well as advanced features to improve efficiency and data security.
Intel has unveiled its upcoming Intel® Stratix® 10 NX FPGAs, its first AI-optimized FPGAs for high-speed, low-latency AI acceleration. These FPGAs will provide customers with customizable, reconfigurable and scalable AI acceleration for computationally demanding applications such as natural language processing and fraud detection. The Intel Stratix 10 NX FPGAs include high-bandwidth embedded memory (HBM), high-performance networking capabilities, and new AI-optimized arithmetic blocks called AI Tensor Blocks, which contain dense arrays of lower-precision multipliers typically used for AI model arithmetic.
As Intel expands its portfolio of advanced AI products to meet the diverse needs of its customers, it is also leading the way in simplifying heterogeneous programming for developers with its portfolio of oneAPI cross-architecture tools to accelerate performance and increase productivity. With these advanced tools, developers can accelerate AI workloads on Intel CPUs, GPUs and FPGAs, and make their code scalable for current and future generations of Intel processors and accelerators.
Intel has enhanced its portfolio of Intel Select solutions to accelerate the deployment of the most urgent IT needs, underscoring the value of delivering pre-verified solutions in today’s changing business climate. Three new and five enhanced Intel Select solutions were announced today. They focus on analytics, artificial intelligence and hyperconverged infrastructures. The enhanced Intel Select solution for genomic analysis is being used worldwide to find a vaccine for COVID-19 and the new Intel Select solution for VMware Horizon VDI on vSAN is being used to enhance distance learning.
3rd generation Intel Xeon Scalable processors and Intel Optane 200 series persistent memories can be shipped to customers today. In May, Facebook announced that 3rd generation Intel Xeon Scalable processors are the basis for its new Open Compute Platform (OCP) servers, and other major communication service providers, including Alibaba, Baidu and Tencent, announced their adoption of next-generation processors. General availability of OEM systems is expected in the second half of 2020. The Intel D7-P5500 and P5600 3D NAND SSDs are available today. And the Intel Stratix 10 NX FPGA is expected to be available in the second half of 2020.
Translated from Intel présente une plate-forme d’analyse et d’intelligence artificielle avec des solutions de processeur, de mémoire, de stockage et de FPGA
