The CPU is a computer’s brain, comprising all of the hardware required to process input, store data, and generate output.
The CPU is always following computer program instructions that tell it which data to process and how to process it. We couldn’t execute applications on a computer without a CPU.
A basic calculator application, for example, may tell the CPU to add two integers, 2 and 2, and return the result.
Thanks to a control unit that understands how to read program instructions and an Arithmetic Logic Unit (ALU) that understands how to add numbers, the CPU can simply perform such instructions. The CPU can process considerably more complicated programs than a basic calculator when the control unit and ALU are coupled.
CPU Features Explained
Your central processing unit (CPU) is at the core of everything you do on your computer, it’s important to select one that’s right for you. To assist you in making your decision, read on to learn about the different features of a CPU.
Intel and AMD are the two companies that make the majority of computer processors. The i3, i5, i7, and i9 CPUs are part of Intel’s standard Core processor family, whereas AMD’s Ryzen processor family comprises the 3, 5, 7, and 9.
The Intel i3 and AMD Ryzen 3 processors offer budget performance, while the i9 and 9 chips deliver high-end workstation capability.
When selecting a CPU, it’s important to consider individual specifications as well as compatibility with other components in your design.
The clock speed of a CPU is the number of cycles completed per second. A 3.1GHz CPU, for example, can execute 3.1 billion operations per second.
The faster your computer runs, the greater the clock speed. The higher the clock speed, the more jobs the CPU can perform.
It’s worth noting that comparing CPU clock rates from various generations or manufacturers isn’t always a good idea.
An earlier quad-core CPU, such as the Intel Core i5-7500 3.4GHz from the 7th generation, will be surpassed by the Intel Core i5-10500 3.1GHz from the 10th generation.
Max Turbo Frequency
Select AMD and Intel CPUs can run at higher frequencies than usual for a brief performance boost when needed. Turbo Core and Turbo Boost are the names given to this technology by AMD and Intel, respectively.
The clock speed of a CPU rises as it ‘turbo-boosts.’ It must be given sufficient electricity to operate within its normal temperature range. The AMD Ryzen 5 3600, for example, has a base speed of 3.6GHz and a Max Boost Clock of up to 4.2GHz.
Again, this increase is just brief, and the CPU will return to its normal condition to avoid overheating and damage.
Overclocking is the process of raising your processor’s basic clock speed. To do so, be sure that both your CPU and motherboard are capable of overclocking.
Overclocking is supported by all AMD Ryzen CPUs, although Intel generally reserves this feature for its higher-end processors. Overclocking is typically possible with Intel Core processor versions that end in a ‘K’ or ‘X,’ such as the Intel Core i5-10600K.
If you want to overclock your CPU, you’ll need a good cooling system because an overclocked chip consumes more power and generates more heat.
Today’s CPUs have several cores, if not all of them. Within the CPU, a core is essentially a microprocessor.
A CPU can only operate on one ‘task’ at a time without considering hyper-threading or multi-threading. As we multitask on our computers, we don’t see the CPU splitting its time between multiple activities since current processors are so quick.
The faster a CPU can manage numerous processes, the better, which is crucial for multitasking or heavy workloads that might benefit from several cores.
A dual-core CPU will enough for most people, while a quad-core or greater processor is suggested for enthusiasts or power users.
Please keep in mind that for gamers, a faster clock speed may be more essential than having more cores, such as quad-core vs. 6-core processors. Many modern games are not intended to use more than a few of cores, although this is gradually changing.
A list of processor families is shown below, sorted by core count:
Threads And Multi-Threading
While the number of physical cores in a CPU is represented by the core count, the number of threads indicates the number of virtual cores that a processor may simulate. An Intel i5-10400 CPU, for example, has 6 physical cores and 12 threads.
To do this, Intel CPUs use ‘hyper-threading,’ whereas AMD CPUs use ‘simultaneous multi-threading,’ or SMT. Both technologies are essentially the same.
More cores and threads, as previously said, mean higher efficiency and multitasking since the CPU is better equipped to split its processing power among multiple activities.
The data needed to complete tasks is stored in three “L” caches on most CPUs. L1, L2, and L3 are the levels, with the capacity rising with each one. If the CPU can’t find the data it needs in the L1 cache, it’ll ‘seek’ it from the L2 cache, and so on.
These caches, beginning with the L1 cache, are the quickest places a processor can access data since they are integrated into the CPU itself.
While each core has its own L1 cache, the L2 cache can be shared or exclusive, and the L3 cache can be shared by all cores.
As it takes longer for a processor to look for essential data as cache size grows, L caches are significantly lower in capacity than RAM.
A typical L1 cache may hold up to 256 KB of data, whereas an L2 cache can hold anything from 256 KB to 8 MB.
It’s usually not required to consider L cache sizes when picking a CPU because manufacturers will have them tuned for each processor.
DDR4, or ‘Double Data Rate,’ is the most recent version of random access memory, which is supported by all current CPUs. Your computer’s RAM stores the data that your CPU requires to run your programs.
When choosing a CPU, there are a few memory specifications to consider:
The memory speed, which is measured in MHz, influences the pace at which data is transferred, with higher values indicating quicker data transfer. RAM with a speed of 3000 MHz operates at 3000 million cycles per second, or 3 billion cycles per second.
DDR4 memory rates of up to 3200 MHz are supported by the current AMD Ryzen CPU generation.
The i3 and i5 CPUs of Intel’s 10th generation enable memory speeds of up to 2666 MHz, while the i7 and i9 processors support speeds of up to 2933 MHz.
Ddr4 Memory Speeds:
Number Of Channels
A channel may be thought of as a communication path between the CPU and memory. The faster the data interchange, the more lanes there are.
Essentially, all current CPUs support at least dual-channel memory, with higher-end workstation CPUs like AMD Threadripper and Intel Xeon supporting more.
Install two RAM modules of the same kind into the two DIMM slots recommended by your motherboard documentation to take use of dual-channel memory.
Column Address Strobe (Cas) Latency
The CAS latency of DDR4 memory modules is generally at least 15, with lower values being desirable.
When two RAM modules have the same speed but differing CL latencies, the one with the lower CL will outperform the other.
Tdp (Thermal Design Power)
The amount of heat generated by a CPU under load is measured in watts by TDP, which stands for Thermal Design Power. TDP is not the same as power consumption, yet it is still useful for choosing the proper power supply and cooling system.
In addition to the other components in a system, such as the GPU, it’s critical to have a power supply that can manage the TDP a CPU is rated for. More powerful components use more energy and create more heat, necessitating the use of more sophisticated cooling systems.
The AMD Ryzen 3-3100 and Intel Core i3-10320, for example, both have a TDP of 65 watts, whereas the more powerful AMD Ryzen 7-3800X and Intel Core i7-10700K have TDPs of 105 and 95 watts, respectively.
Process of Fabrication, Micro-Architecture, and Generation
There are billions of transistors in a contemporary CPU. The semiconductor fabrication technique used to build a CPU is measured in nm (nanometers), and the manufacturing process is measured in nm (nanometers).
Please keep in mind that the measurement in nanometers, such as 7nm, does not indicate the size of the transistors and is more of a marketing phrase. Furthermore, Intel’s 10nm process transistor density is said to be comparable to AMD’s 7nm process.
More transistors can fit in the same space with smaller, more energy-efficient transistors, leading in more powerful, energy-efficient CPUs. This standard can give you an indication of how sophisticated a processor is, but it isn’t a reliable indicator of performance. Clock speed and core count are more essential specifications to consider.
The arrangement structure of electronic components on the processor chip is known as micro-architecture. Every contemporary architecture is meant to be more efficient and effective than the one before it.
The latest AMD 5000 series CPUs, for example, are based on the Zen 3 Architecture.
Similarly, Intel’s I series CPUs’ 10th iteration is built on the Comet Lake microarchitecture. Coffee Lake was the basis for the 8th generation CPUs.
It is usually preferable to use latest generation processors. However, earlier versions of CPUs may be offered at lower prices, resulting in a better performance-to-price ratio in some cases.
Type A Socket A CPU socket is a physical interface on a motherboard where a processor is installed in order to connect to other computer components. Both AMD and Intel CPUs have their own unique socket types.
The amount of pins utilized and their physical arrangement are determined by the socket type, so make sure your motherboard and CPU are compatible. Keep in mind that even if two processors have the same socket type, they will not work together on the same motherboard. This is due to the adoption of a different chipset (more on this below).
If you want to upgrade your CPU in the future, having a backward-compatible socket is advantageous. AMD’s AM4 socket, which supports older CPUs, offers the best backward compatibility, whereas Intel has a handful of distinct variants that are still in use today.
Current Intel Sockets:
Current AMB Sockets:
Older CPU versions are incompatible with the LGA 1200 socket. The new AMD AM4 socket, on the other hand, is backward compatible with previous AMD CPUs.
A motherboard’s chipset is in charge of data transmission between the various components of your computer. It influences how quickly your computer’s components interact with one another, whether you can overclock your CPU, how fast your RAM runs, and how many USB ports you may have.
PCIe® lanes are used to connect the components of a computer, and more lanes mean quicker data transfer. A PCIe lane may be thought of as the actual cables that link components on a motherboard.
A more cost-effective chipset, such as AMD’s B450, has 36 PCIe® 3 Lanes, but the more sophisticated and costly X570 has 44 PCIe® 4 Lanes.
Common Intel Chipsets:
Common Amd Chipsets:
Many CPUs include integrated graphics, which eliminates the need for a separate GPU to control your monitor. While an integrated graphics solution is a good choice for casual users, some users want to upgrade their system with a separate graphics card for better graphical performance.
When troubleshooting a discrete graphics card, integrated graphics might be useful since you can still see an image on your monitor even if your discrete GPU is malfunctioning.
Intel’s Intel HD Graphics, Intel UHD Graphics, and Intel Iris are available in certain CPUs, while AMD’s Radeon Vega graphics are available. It’s ideal to have a dedicated graphics card if you’re doing advanced photo/video editing or high-end gaming.
As processing technology improves, the two major CPU manufacturers (Intel and AMD) release new CPU series on a regular basis.
With each series, we see a newer, better architecture capable of delivering more performance.
In most cases, Intel and AMD are working on numerous CPU series at the same time.
Both manufacturers, for example, provide mainstream CPUs as well as server-oriented processors. Every few of years, fresh series of CPUs are released.
There are a variety of sockets to go with the various CPU series. The CPU socket is the location on the motherboard where the processor will be placed.
Some CPU series can share the same socket. The LGA 1151 socket is used by Intel’s Skylake and Kaby Lake CPUs, for example.
Some CPU series, on the other hand, employ new sockets when they are updated to a newer architecture. Intel’s Core processors, for example, have used a variety of sockets, including LGA 775, LGA 1156, LGA 1155, LGA 1150, and, most recently, LGA 1151.
It’s critical to match your CPU to a motherboard with the right socket. If you receive a motherboard with a socket that doesn’t work with your CPU, you’ll have to return one or the other since they won’t function together. So, the most important thing to remember regarding CPU series and socket types is that you should look for a processor/motherboard combo that not only has suitable sockets, but also has the most up-to-date series and architecture that you can afford.
Operating Frequency (GHz)
Clock rate, or CPU running frequency (measured in hertz), is an often misunderstood subject among first-time builders. Many first-time builders see the operating frequency as the be-all and end-all of assessing a processor’s worth. This, on the other hand, could not be further from the truth.
A processor’s operating frequency refers to how quickly it can finish a single work cycle. The greater the frequency, the faster a single work cycle may be completed.
A greater operating frequency, on the other hand, does not imply better performance. This is because CPUs can only handle a certain amount of instructions each clock cycle (Intructions Per Clock, or IPC). For example, a processor that can complete one million instructions per clock cycle and each clock cycle has a frequency of 4.0 GHz would still not perform as well as a CPU that can execute two million instructions per clock cycle and operates at 3.7 GHz.
If you can get the identical CPU to run at 5.0 GHz or more, the performance gap between it and the 3.7 GHz will narrow significantly. This is why performance aficionados are so fond of overclocking.
In the end, depending to your budget, you should seek for the CPU with the most cores, the greatest frequency, and the most instructions per clock cycle.
Hyperthreading essentially allows the CPU to work on two threads (sequences of instructions for the CPU to execute) at the same time. Hyperthreading, unlike core technology, does not allow the microprocessor to do several tasks at the same time. Instead, it lets the processor to operate on two separate tasks.
Consider working on a toy factory assembly line to have a better idea of what we are talking about. You must attach the head-piece to the body of an action figure on your production line. Every three minutes, the conveyor belt spits out a new headless action figure, and installing the head takes roughly ten seconds.
That leaves you with around 2 minutes and 50 seconds to do nothing but wait for the next headless action figure to descend the belt. If the toy manufacturer was clever, they would assign you to another line of production (which would be added behind you and you would stand in between the two assembly lines).
The conveyor belt on this manufacturing line would create a toy doll every three minutes, and your job would be to attach the doll’s arms. You could efficiently work both assembly lines without interruption if the second assembly line was staggered such that it produced a toy doll one minute and 30 seconds after each headless action figure was created.
In some ways, this is how hyperthreading works. It doesn’t let your CPU execute many things at once, but it does delegate jobs more effectively to the point that it can boost speed in some situations.
Hyperthreading isn’t widely used in gaming right now, but that might change with the next generation of games. So, if you’re looking for a new gaming CPU, Intel’s hyperthreading is something to think about, but it’s not a must-have feature right now.