With the introduction of the 4th Gen Intel® Xeon® Scalable processors (code-named Sapphire Rapids), the Intel® Xeon® CPU Max Series (code-named Sapphire Rapids HBM), and the Intel® Data Center GPU Max Series (code-named Ponte Vecchio), Intel today marked one of the most significant product launches in the company’s history. These products will give Intel’s customers a boost in data centre performance, efficiency, security, and new capabilities for AI
With the help of the 4th Gen Xeon, Intel is bringing unique systems and solutions at scale to its customers’ and partners’ toughest computing problems. Due to Intel’s distinctive strategy of offering acceleration specifically designed to accelerate particular workloads and highly optimised software geared for particular workloads, the business is able to give the ideal performance at the ideal power for the best possible total cost of ownership.
The 4th Gen Xeon processors provide clients a variety of capabilities for managing power and performance, making the most use of CPU resources to aid in their sustainability goals, and are Intel’s most environmentally friendly data center processors.
“The launch of 4th Gen Xeon Scalable processors and the Max Series product family is a pivotal moment in fueling Intel’s turnaround, reigniting our path to leadership in the data center and growing our footprint in new arenas”, “Intel’s 4th Gen Xeon and the Max Series product family deliver what customers truly want – leadership performance and reliability within a secure environment for their real-world requirements – driving faster time to value and powering their pace of innovation.”
Sandra Rivera, Executive VP and GM of the Data Center and AI Group, Intel
The 4th Gen Xeon series significantly improves on Intel’s purpose-built, workload-first strategy and approach, unlike any other data centre processor currently on the market or in customers’ hands.
With the Most Built-In Acceleration, Leading Performance and Sustainability Benefits
Over 100 million Xeons have been installed in the market as of late, powering everything from on-premises servers that run IT services, such as new as-a-service business models, to networking hardware that controls Internet traffic, wireless base station computing at the edge, cloud services, and other devices.
New 4th Gen Xeon processors deliver industry-leading performance with the most built-in accelerators of any CPU in the world to address customers’ most important computing challenges across AI, analytics, networking, security, storage, and HPC. They do this by building on decades of data centre, network, and intelligent edge innovation and leadership.
When using built-in accelerators, 4th Gen Intel Xeon customers can anticipate an increase in performance per watt of 2.9×1 on average for targeted workloads, up to 70 watts2 in power savings per CPU in optimised power mode with little performance loss, and a 52% to 66% lower TCO3 when compared to prior generations.
Because the built-in accelerators in the 4th Gen Xeon are so extensive, Intel is able to save power at the platform level, reducing the demand for extra discrete acceleration and assisting our customers in achieving their sustainability objectives. Additionally, for some workloads, the new Optimized Power Mode can provide up to 20% socket power savings with less than 5% performance impact11. The overall energy used by data centres is being further reduced by new developments in air and liquid cooling, while the 4th Gen Xeon was constructed using 90% or more renewable electricity at Intel sites with cutting-edge water reclamation facilities.
With integrated Intel® Advanced Matrix Extension (Intel® AMX) accelerators, 4th Gen Xeon processors in AI achieve up to 10×5,6 greater PyTorch real-time inference and training performance compared to prior generation. New levels of performance for inference and training across a broad range of AI tasks are unlocked by Intel’s 4th Gen Xeon. With up to a 20×12 speedup on big language models, the Xeon CPU Max Series extends on these strengths for natural language processing.
Developers can use their preferred AI tool while working more productively and developing AI faster thanks to Intel’s AI software package. The suite can scale out in the cloud and all the way out to the edge because it is portable from the workstation. More than 400 machine learning and deep learning AI models across the most popular AI use cases in every business sector have verified it.
A family of processors designed for high-performance, low-latency network and edge workloads is available with 4th Gen Xeon. These processors are an essential component of the infrastructure enabling a future that is increasingly software-defined across a range of sectors, from manufacturing and smart cities to telecommunications and retail. Built-in accelerators aid in increasing throughput and reducing latency for 5G core applications, while improvements in power management improve the platform’s responsiveness and economy. Additionally, the 4th Gen Xeon delivers up to twice the virtualized radio access network (vRAN) capacity without increasing power consumption as compared to earlier generations. In order to satisfy their essential performance, scalability, and energy efficiency objectives, communications service providers are now able to double the performance-per-watt.
Extremely Fast Computing
For the most demanding computational workloads in HPC and AI, the Intel Max Series product family and the 4th Gen Xeon processor deliver a scalable, balanced architecture that combines CPU and GPU with oneAPI’s open software ecosystem.
The Xeon CPU Max Series is the first and only x86-based processor with high bandwidth memory, enabling it to accelerate numerous HPC tasks without requiring code modifications. The highest-density Intel CPU, the Data Center GPU Max Series, will be offered in a variety of form factors to suit various customer requirements.
High bandwidth memory (HBM2e) of 64 gigabytes is available on the Xeon CPU Max Series package, greatly enhancing data throughput for HPC and AI workloads. The Xeon CPU Max Series offers up to 3.7 times10 more performance on a variety of real-world applications including energy and earth system modelling when compared to the top-end 3rd Gen Intel® Xeon® Scalable processors.
Furthermore, the Data Center GPU Max Series increases throughput for demanding tasks like physics, financial services, and life sciences by packing over 100 billion transistors into a 47-tile device. When running the LAMMPS molecular dynamics simulator, the combined platform, when combined with the Xeon CPU Max Series, achieves up to 12.8 times13 more performance than the previous generation.
Most Functional and Secure Xeon Platform to Date
With up to four Intel 7-built tiles combined on a single package and connected using Intel EMIB (embedded multi-die interconnect bridge) packaging technology, the 4th Gen Xeon represents the biggest platform transformation that Intel has ever delivered. It also boasts new features like increased memory bandwidth with DDR5, increased I/O bandwidth with PCIe5.0, and Compute Express Link (CXL) 1.1 interconn.
Security is the cornerstone of it all. Data security, regulatory compliance, and data sovereignty are all improved by Intel’s 4th Gen Xeon, which offers the most comprehensive confidential computing portfolio of any data centre silicon provider in the market. With Intel® Software Guard Extensions (Intel® SGX), which offers today’s smallest attack surface for confidential computing in private, public, and cloud-to-edge contexts, Intel continues to be the only silicon provider to offer application isolation for data centre computing. Additionally, Microsoft Azure, Alibaba Cloud, Google Cloud, and IBM Cloud will introduce Intel’s new virtual-machine (VM) isolation technology, Intel® Trust Domain Extensions (Intel® TDX), which is perfect for converting current workloads into a secure environment.
Finally, thanks to the 4th Gen Xeon’s modular architecture, Intel is able to provide a wide selection of processors across almost 50 targeted SKUs for customer use cases or applications, ranging from general-purpose SKUs to ones specifically designed for use in the cloud, databases, analytics, networking, storage, and single-socket edge use cases. The 4th Gen Xeon processor series is On Demand-capable and comes in a variety of core counts, frequencies, combinations of accelerators, power envelopes, and memory throughputs depending on the form factors and intended use cases being addressed by the processors.