Unlike shared hosting and virtual servers, hosted dedicated servers give
you control over your choice of processor (CPU). Here’s some
information to help you compare processors when buying dedicated
servers.
How fast a processor works depends partly on the clock speed, which is the speed at which the processor executes instructions. The faster the clock, the more instructions the CPU can execute per second. It is the processor speed, measured in hertz (GHz).
But as processors progress, improvements in chip architecture mean that irrespective of the clock speed, a processor can execute more instructions simultaneously, thanks to things like multiple cores, multiple threading, subcycling, increased data throughput and reduced latency.
Most modern hosted servers offer multi-core, multithreaded processors, which means that one very good processor may perform more tasks, more efficiently and faster than two or more lesser or older processors, even if the latter have a faster clock speed.
Multi-core means that a server actually has more than one processor core working to complete the tasks demanded by its users. Although a 4-core processor won’t necessarily be 4x faster than a single-core processor, it will still be able to execute instructions even if one or more of the cores is being fully utilized, rather than stalling. Multithreading further improves performance by allowing multiple threads of code, or multiple parts of a process to be executed simultaneously.
Another important specification of a processor is the cache memory, memory that reduces the time needed to access data from the main memory of the server. The cache memory allows the processor to store and access frequently required data much more quickly. The higher the cache memory, the wider the variety of data that can be held in cache, speeding up the performance of the processor.
Beyond these workload multipliers, each generation of processor chipsets brings with it improvements in chip architecture designed to increase the speed and capacity of performance, by reducing latency and increasing throughput at various stages of the computing process.
According to Intel, their latest Xeon E5 chipset v3 “Haswell” version provides a 3x performance improvement over its predecessor, and a 12x improvement over its 4-year old counterpart, the company citing a variety of architectural improvements during its launch.
Intel Xeon E5 v3 dual-socket servers support up to 36 cores and are among the few processors that are compatible with the new DDR4 RAM version that uses vastly reduced power consumption and provides superior data transfer speed.
Processor features
The Central Processing Unit (CPU) in your server, also referred to simply as the processor, is what interprets and executes instructions, processing data and performing tasks like serving web pages, running database queries and executing other program and computing commands. The more processors in a server, the faster and more efficiently the server is capable of working, and the more instructions can be executed in a shorter space of time.How fast a processor works depends partly on the clock speed, which is the speed at which the processor executes instructions. The faster the clock, the more instructions the CPU can execute per second. It is the processor speed, measured in hertz (GHz).
But as processors progress, improvements in chip architecture mean that irrespective of the clock speed, a processor can execute more instructions simultaneously, thanks to things like multiple cores, multiple threading, subcycling, increased data throughput and reduced latency.
Most modern hosted servers offer multi-core, multithreaded processors, which means that one very good processor may perform more tasks, more efficiently and faster than two or more lesser or older processors, even if the latter have a faster clock speed.
Multi-core means that a server actually has more than one processor core working to complete the tasks demanded by its users. Although a 4-core processor won’t necessarily be 4x faster than a single-core processor, it will still be able to execute instructions even if one or more of the cores is being fully utilized, rather than stalling. Multithreading further improves performance by allowing multiple threads of code, or multiple parts of a process to be executed simultaneously.
Another important specification of a processor is the cache memory, memory that reduces the time needed to access data from the main memory of the server. The cache memory allows the processor to store and access frequently required data much more quickly. The higher the cache memory, the wider the variety of data that can be held in cache, speeding up the performance of the processor.
Beyond these workload multipliers, each generation of processor chipsets brings with it improvements in chip architecture designed to increase the speed and capacity of performance, by reducing latency and increasing throughput at various stages of the computing process.
According to Intel, their latest Xeon E5 chipset v3 “Haswell” version provides a 3x performance improvement over its predecessor, and a 12x improvement over its 4-year old counterpart, the company citing a variety of architectural improvements during its launch.
Intel Xeon E5 v3 dual-socket servers support up to 36 cores and are among the few processors that are compatible with the new DDR4 RAM version that uses vastly reduced power consumption and provides superior data transfer speed.