Many choices need to be considered when selecting a dedicated server hosting solution. One of the most basic questions to consider is what type of CPU best fits your needs and business objectives. The CPUs powering the servers you select can have an impact on your ability to address your expected workload.
This article is intended to help you decide between processors from Intel and those from AMD (Advanced Micro Devices). We will provide a comparison of the offerings available from the two chipmakers and attempt to identify key differences that may influence your decision when selecting dedicated servers.
Before attempting to choose a dedicated hosting platform, you need to understand your business requirements and the applications you intend to run. These factors may influence your selection to take advantage of the features provided by a specific chip manufacturer.
Both Intel and AMD have many different processor models on the market. The following discussion will highlight some of these chips but is not meant to be an exhaustive inventory of the products available from these manufacturers.
Preliminary Data Gathering
When starting the process of selecting a CPU for a new server, different CPU choices might make the most sense based on how you will be using the server. Factors to consider when choosing a CPU include the following.
- Obtaining hardware recommendations for the applications you intend to run on the server is a good place to start. Certain developers will suggest a specific CPU or clock speed as a minimum requirement for running their software. Take into account whether multiple core applications will be run from the same server. In the absence of specifications from the developer, try to see what type of hardware other customers are deploying to support the application.
- The number of concurrent users accessing the server will have a substantial effect on CPU load. As the number of users increases, so do CPU requirements. In some cases, it is more efficient to deploy multiple servers rather than attempting to service them all with a single, high-powered machine.
- Plan for future growth when selecting all aspects of a dedicated server, including its CPU. Estimating the number of concurrent users the server needs to handle in six months to a year can help determine the proper CPU. Failure to plan for growth may necessitate a CPU upgrade or deploying additional servers to address user requirements.
Overview of Intel and AMD CPUs
Let’s take a look at some of the most powerful CPUs currently offered by these two chip producers. We will concentrate on CPUs recommended by the manufacturers for use in data centers or servers as opposed to chips for desktop or gaming machines.
Intel
Intel produces multiple families of processors that are designed to power everything from personal laptops to mission-critical, cloud-based dedicated servers. Specifically, Intel’s Xeon and Xeon Scalable processors are the company’s most powerful server-focused CPUs. They are available in a large variety of configurations to meet the diverse business needs of their customers.
Intel Xeon Scalable processors are available in many models to address the unique needs of a computing environment. They offer varying power, functionality, and price points to handle virtually any type of dedicated server. Features common to all varieties of Xeon Scalable processors include:
- Built-in AI (artificial intelligence) acceleration supported by data science tools and multiple smart solutions;
- Intel SGX security solution which protects data and application code in real-time;
- Intel Optane Persistent Memory that provides up to 6TB of system-level memory capacity when combined with traditional DRAM.
Following are some examples of the processor classes available. As you move through the classes, the available feature set is expanded, allowing for the CPUs to be used in different usage scenarios.
Intel Xeon Bronze Processors:
- 8 cores;
- 9GHz speed;
- 11MB cache.
Intel Xeon Silver Processors:
- 8-16 cores;
- 11MB-20MB cache;
- 3GHz to 3.2GHz base speed with 3.2GHz to 4.0GHz max turbo speed.
Intel Xeon Gold Processors:
- 8-32 cores;
- 12MB-48MB cache;
- 0GHz to 3.6GHz base speed with 3.1GHz to 4.03GHz max turbo speed.
Intel Xeon Platinum Processors:
- 8-40 cores;
- 24MB-60MB cache;
- 1GHz to 3.9GHz base speed with 3.0GHz to 4.4GHz max turbo speed.
AMD
AMD’s 3rd Generation EPYC processors are the company’s current top-of-the-line offering for use in servers. These CPUs are widely supported by cloud vendors and have been used for many processor-intensive high-performance computing solutions. AMD supports all features across the board for all of its processors.
AMD EPYC 7002 Series Processors offer powerful performance powered by these base specifications:
- Up to 64 cores and 128 threads;
- Speeds up to 2.6GHz base, 3.4GHz boost, and 256MB cache;
- 4TB of DDR4 memory capacity across 8 memory channels;
- 128 PCIe 4.0 lanes;
- AMD Infinity Guard proprietary security solution.
AMD EPYC 7003 Series Processors are similarly powered with:
- Up to 64 cores and 128 threads;
- Speeds up to 2.45GHz base, 3.675GHz boost, and 256MB cache;
- 4TB of DDR4 memory capacity across 8 memory channels;
- 128 PCIe 4.0 lanes;
- AMD Infinity Guard proprietary security solution.
Processor Type of Specific Applications
Intel processors commonly provide enhanced performance for applications that only make use of a single core due to the TPC per core with Intel being faster over AMD with a single core versus a single core. Threaded applications that use multiple cores simultaneously can take advantage of the greater core density available in AMD CPUs. This fact highlights the need to understand the applications that will be using your dedicated server and to keep in mind hardware recommendations made by the developer.
Chip producers offer many models to allow systems to be configured to meet customers’ specifications. There is no need to overspend for unnecessary computing power or underspend and live with degraded performance when putting together a Dedicated Host solution.
Chip Supply, Demand, and Innovation
Executives at both Intel and AMD are concerned with the current chip shortage that is affecting many companies across numerous market sectors. The effects of the COVID-19 pandemic and supply-chain disruption have severely impacted the chip manufacturers and are not expected to be resolved until late 2022 or 2023.
CPU producers may be forced to concentrate on their most profitable lines while waiting for the shortage to pass. Dealing with this issue can also hinder innovation and delay the introduction of new products.
Intel and AMD are industry leaders in large part because of their ability to innovate and develop products that integrate with emerging standards. Both manufacturers are slated to support DDR5 and PCIe 5.0. It appears that Intel’s 12th generation CPUs will beat AMD to market and offer this support in late 2021. AMD’s offerings should start appearing in early 2022.
Making the Choice in CPU for Dedicated Hosts
The choice of CPU is important but is not the only factor that determines whether your chosen dedicated hosting solution provides the performance to meet your business objectives. In the end, a serviceable solution can be obtained using processors from either Intel or AMD. While the processor can make a difference, the real value comes from working with a cloud vendor that understands your needs and will tailor a solution to address them.
Locate a partner that offers the networking, memory, and storage capabilities to complement your server’s CPU to extract the most performance for your money. Your vendor may concentrate on a specific chip manufacturer when designing your system. Depending on the amount of processing power your business requires, the cost of the dedicated server’s CPU will vary. A collaboration with a cloud provider like Atlantic.Net can help you select the right Intel processor to handle your current workload and position you for future growth.
