The front and back of an intel xeon scalable processor

Server Processors: Choosing the Right Server CPU

Find the right server processor with the right performance, manageability, security, and AI acceleration features to drive your business success.

Server Processor Key Takeaways

  • Server processors execute calculations and tasks at a far more complex and robust level compared to PC processors.

  • Not all processor cores are created equal and knowing what to prioritize for a server CPU depends on your intended use case.

  • Technology partners are already familiar with most applications for server CPUs and can help guide you in choosing the right solution.

  • For any challenge, there is an ideal Intel® Xeon® processor solution backed by Intel’s history in the data center and world-class support.

author-image

作者

What Is a Server Processor?

Processors are the brain of the computer, executing calculations, tasks, and functions. Server processors differ from PC processors in that they are typically designed to handle heavier, more complex workloads, such as:

  • Email exchange, file sharing, and database transactions for an office or remote workers and customers all over the world.
  • Telecommunications services that process network traffic routing for millions of connected devices.
  • Connection and control of multiple devices on a factory line, from conveyor belts to robotic arms and cameras.
  • Extreme compute tasks such as using AI to map genomes or simulate global weather patterns.

Factors to Consider in Choosing a Server CPU

Choosing the right server processor depends on the intended application or use case. This process entails balancing anticipated needs against total platform cost with the goal of high utilization, which refers to all components of a server platform operating at or near capacity. In other words, aim to drive the most value out of your technology investments and avoid situations where your technology sits idle.

Cores, Clock Speed, Threads, and Motherboards

When browsing for a server CPU, you may come across terms such as cores, threads, and frequency. Here is some basic processor vocabulary to help guide you during the research phase of your journey:

  • Processing cores are the physical units of compute within a processor.
  • Threads are virtual lines of code that the processor core executes; most cores can process up to two threads.
  • Frequency, or clock speed, measures how fast a core can process a thread.
  • Cache refers to the processor’s dedicated, onboard memory that helps the processor execute workloads.
  • Thermal design power (TDP) describes the processor’s ability to dissipate operational heat, expressed in watts.
  • The motherboard, or server board, houses the processor in a socket and connects it to all other components in the server, including power supply, memory, storage, and add-in PCIe cards such as accelerators, network interface cards (NICs), or graphics cards/graphics processing units (GPUs).

Intel® processors offer a wide range of entry points for cores, threads, clock speeds, and TDP to provide the right solution match for any use case. In addition, advanced features can help improve workload processing and server efficiency for specialized use cases, helping to drive down cost of ownership while enabling innovation and the delivery of new services.

  • Processors with built-in AI acceleration can help lower the compute requirements to deploy AI for automation or analytics, especially as these use cases become mainstream and necessary to maintain a competitive advantage. Select Intel® processors offer Intel® Deep Learning Boost (Intel® DL Boost) to accelerate AI training and inference workloads.
  • Built-in security capabilities can help establish a root of trust during server start-up and provide fast data encryption with a low performance overhead. Intel® Software Guard Extensions (Intel® SGX) on select Intel® processors helps protect data in memory for additional platform hardening.
  • Integrated graphics processing units, or iGPUs, can be beneficial to servers that process graphics workloads or AI workloads, which depend on similar floating-point operations. Workstations may also use server-grade CPUs if they are built for specialized tasks such as computer-aided design (CAD) or rendering. Intel® processors feature graphics capabilities driven by Intel® Xe graphics architecture to help reduce the need for discrete GPUs.
  • Support for error correction code (ECC) memory helps improve server uptime by detecting and correcting single-bit data corruption in system memory. The CPU must support ECC, and your server must have ECC memory modules to take advantage of this capability. Select Intel® processors offer support for ECC, which is especially useful for embedded or industrial use conditions.

Server Processors for Cloud Instances

In a public cloud service provider (CSP) environment, the number of cores and threads and clock speed are secondary considerations because compute resources in the cloud are abstracted from the underlying hardware. However, there are advantages to knowing what CPUs power the cloud instances that your business uses. Look for hardware-enabled security capabilities in the CPU that can help protect your data in a multitenant environment by keeping it encrypted and isolated from other workloads.

What to Prioritize in a Server CPU

Buying the fastest, highest-core-count processor is not helpful if you only ever use a small percentage of its total compute power. This would be analogous to buying a race car just to commute to work. The cost to maintain a high-end platform for simple tasks may outweigh the benefit entirely.

Not all cores are built the same. Some are manufactured for heavy workloads with raw compute power, others feature integrated graphics to accelerate video and AI processing, and still others may feature integrated security and manageability features to extend the capabilities of the IT department. For these reasons, you should prioritize the features that can help you achieve your desired business outcomes.

Customer-Centric Design

Intel is continuously innovating to design processors that address specific customer needs across applications and target use cases.

  • Intel® Xeon® Scalable processors offer a viable entry point for any application and are backed by a proven history of three successful platform generations with a fourth on the way. Bronze, Silver, Gold, and Platinum SKUs offer customers increasing levels of performance to support growing data center needs, with key hardware-based AI acceleration and security capabilities. Within each product line, there are specialized SKUs, such as N-SKUs for networking or M-SKUs for media, that your technology provider can recommend for even more targeted performance.

While Intel® Xeon® Scalable processors offer a solution for virtually any application or need, additional flavors of Intel® Xeon® processors are available to address targeted use cases.

  • Intel® Xeon® D processors are systems-on-chip (SoCs) designed for AI at the edge, IoT, and telecommunications use cases. This processor can go where others can’t, with ruggedized, space-constrained, and outdoor applications in mind.
  • Intel® Xeon® E processors are entry-level platforms that deliver business-ready performance suitable for small and medium-sized businesses looking for an on-premises server.
  • Intel® Xeon® W processors are workhorse CPUs designed for workstation applications such as rendering for VFX or computer-aided design or AI-driven workflows in settings such as medical diagnostics.

Work with a Technology Partner

There are many different use cases, applications, and business goals that a server must fulfill. No matter what path you choose, the soundest advice is to not go it alone. Technology partners and experts are available to advise on the best configurations across any targeted use case, and most use cases are well mapped and understood. Talk to your OEM, system integrator, solution provider, or chipmakers like Intel for answers and guidance.

Why Intel: The Right Solution, Not Just the Right Processor

In addition to server processors, Intel offers a comprehensive portfolio of server hardware, including high-performance memory and nonvolatile storage with Intel® Optane™ technology, Intel® Ethernet products, programmable switches with Intel® Intelligent Fabric Processors, and Intel® FPGAs. Intel’s ecosystem of technology partners—across hardware, software, and services—delivers proven, reliable solutions to accelerate your time to market and time to value. Intel also supports the global community of developers with resources, training, and forums to connect and share ideas.

Standing by Our Products

Intel works through channel partners to provide serviceability of up to 15 years, in addition to warranties up to three years out of the box. For any issue or quality concern, Intel will work in a transparent manner to address customer needs.

Software and Optimization

Hardware doesn’t exist in a vacuum. Software can be tuned to run more efficiently on a processor to help maximize the value from your technology investments. Intel employs thousands of developers worldwide whose goal is to collaborate on finding the most efficient ways to run the most popular software solutions on Intel® architecture. The open source community also benefits from close collaboration and support from Intel to spur innovation and solve industry-wide challenges.

Getting Started

The best way to get started in selecting a server CPU is to work with your technology provider to find the right Intel® Xeon® processor-enabled solution for your specific needs.

Frequently Asked Questions

Server processors are designed to support the heavy-duty workloads that servers handle, from multioffice file sharing and email exchange to cloud service provider data centers with millions of customers. Compared to PC processors, server processors generally provide higher compute power, deliver hardware-accelerated AI and security capabilities, and support specialized technology such as ECC memory.