Considering ARM Servers: Is It Time to Make the Switch?

In recent years, ARM has been gaining significant attention.

It has not only dominated traditional fields like phones and embedded applications, but is also starting to emerge in PC desktop systems, and now even in high-performance computing, particularly among public cloud vendors claiming they are using ARM processors.

With this momentum, can we imagine that it won’t be long before we can use ARM servers?

Today, I will throw out some ideas and welcome experts to contribute in the comments section.

Although both x86 and ARM are processors, their design philosophies are fundamentally different.

First, in terms of naming. Intel and AMD processors are called x86 processors because the first processor in this series was identified by a number ending with the sequence “86”. In other words, the term x86 refers to the instruction set architecture series based on the Intel 8086.

On the other hand, ARM is named after the Advanced Risc Machine company. This company does not manufacture chips but licenses its chip designs to other companies.

Secondly, in terms of architecture. ARM processors are based on RISC (Reduced Instruction Set Computing); while x86 processors represented by Intel and AMD are based on CISC (Complex Instruction Set Computing).

The architecture names themselves indicate that CISC architecture is more complex, as it provides more instructions, is more adaptable, and can deliver better performance, but the downside is increased power consumption. RISC architecture, in contrast, is more about being small and efficient, with a simpler instruction set that doesn’t need to handle complex tasks, thereby reducing power consumption.

Additionally, x86 has aninternal program (microcode). This program can execute instructions, allowing manufacturers to improve it. All these factors make x86 very fast and efficient, but it occupies more physical space and consumes more power. In contrast, ARM processors do not have this microcode, and they have fewer processing stages (usually 3 to 8, whereas x86 has 16 to 32).

This explains why ARM-based devices predominantly appear in phones and embedded devices, while x86-based devices have the advantage of being able to run virtually any standard operating system on any computer (including Windows and Linux). Because the two represent fundamentally different design philosophies from the outset.

Currently, the main use of ARM-based processors in the enterprise market is still in Ethernet adapters or SmartNICs, used to offload network processing tasks from the CPU, but taking on the heavy responsibilities of enterprise business is still quite challenging.

Therefore, in the short term, x86 servers remain the best choice for most enterprise users.

In fact, regarding ARM, a reporter asked Michael Dell, Chairman and CEO of Dell Technologies last year, and he believes that producing an ARM-based device is not difficult, the key is whether the user needs it: