About CPU

Comprehensive CPU Processor Knowledge

Comprehensive Explanation of CPU Knowledge

 

In today's market, the CPU landscape is predominantly defined by two major camps: Intel and AMD. With its significant market share, Intel holds the mantle as the leader in the CPU industry. However, since the advent of AMD's Ryzen processors, performance metrics have increasingly favored the Intel platform. AMD has strategically pursued a path emphasizing cost-effectiveness. Consequently, the performance disparity between Intel and AMD platforms is steadily narrowing, each possessing its strengths. Therefore, both choices are commendable, whether for Intel or AMD processors.

 

 

Part I: Understanding CPU Model Naming Conventions

 

 

Within CPU model names, one can glean various parameter information such as the CPU's tier, generational iteration, overclocking support, and integrated graphics presence. Let's illustrate with several examples of Intel CPU models:

 

 

Note: For Intel CPUs, models without an "F" suffix typically include integrated graphics.

 

 

For instance, consider the model Intel Core i7-12700KF: "Intel" denotes the brand, "Core i7" signifies a high-end tier, "12" represents its generation (12th generation CPU), with higher numbers following "700" indicating superior performance. The "KF" suffix indicates support for overclocking and the absence of integrated graphics.

 

 

Similarly, take the model Intel Core i7-11700K: "Intel" denotes the brand, "Core i7" signifies a high-end tier, "11" represents its generation (11th generation CPU), with higher numbers following "700" correlating to enhanced performance. The "K" suffix indicates support for overclocking, while the absence of the "F" suffix signifies integrated graphics.

 

 

Next, the model Intel Core i5-9400F: "Intel" represents the brand, "Core i5" signifies a mainstream tier, "9" denotes its generation (9th generation CPU), with higher numbers following "400" indicating better performance. The "F" suffix signifies the absence of integrated graphics.

 

 

Lastly, the model Intel Core i3-8100: "Intel" signifies the brand, "Core i3" signifies a mid-to-low-end mainstream tier, "8" represents its generation (8th generation CPU), with higher numbers following "100" correlating to improved performance.

 

 

It's important to note that, for example, the model Intel Core i5-750: Any three-digit number should not be assumed to represent a 7th-generation product. Instead, three-digit numbers denote 1st-generation CPUs, while 7th-generation CPUs are distinguished by four-digit numbers, such as i5-7500 or i7-7700.

 

 

Now, let's consider a few examples of AMD models:

 

 

Note: Nearly all AMD Ryzen CPUs lack integrated graphics. Only models with a "G" suffix include integrated graphics, typically classified as APUs. Additionally, AMD's early naming conventions needed to be more consistent. Here, we primarily focus on the Ryzen series.

 

 

For instance, consider the model AMD Ryzen 7 5800X: "AMD" represents the brand, "Ryzen" mirrors Intel's Core i series, and "R7" signifies a high-end tier akin to Intel's i7 series. The model belongs to the 5000 series, signifying the fourth generation Ryzen, with higher numbers following "800" indicating better performance. The "X" suffix denotes support for XFR technology.

 

 

Similarly, the model AMD Ryzen 5 3600: "AMD" signifies the brand, "Ryzen" mirrors Intel's Core i series, and "R5" signifies a mainstream tier similar to Intel's i5 series. The model belongs to the 3000 series, meaning the third generation Ryzen, with higher numbers following "600", indicating superior performance.

 

 

Lastly, consider the model AMD Ryzen 3 3100: "AMD" signifies the brand, "Ryzen" mirrors Intel's Core i series, and "R3" signifies a mid-to-low-end tier akin to Intel's i3 series. The model belongs to the 3000 series, signifying the third generation Ryzen, with higher numbers following "100" indicating better performance.

 

 

Intel's Core i3, i5, i7, and i9 tiers correspond to mid-to-low, mainstream, high-end, and flagship tiers, respectively. These designations only apply within the same generation and cannot be used to compare tiers across different generations. This is because each generation sees varying degrees of performance improvements. For instance, the mid-tier i5 12600K from the 12th generation can outperform the flagship i9-11900K from the 11th generation. AMD follows a similar pattern. Thus, when comparing CPU performance, one must rely on something other than i3, i5, i7, or i9 designations. Behind each model lies a complex interplay of architecture and manufacturing processes, which we'll briefly cover later in this article. For novice users, the most straightforward method to assess CPU performance is to consult CPU comparison charts.

 

 

The performance of the latest i3 models is quite impressive. For example, the 10th generation Core i3 10105 can outperform the 7th generation Core i7 7700. The i3 tier represents the lowest level within the Core lineup, with entry-level tiers being represented by Pentium and Celeron processors. Examples include the 10th generation Pentium G6405 and the 10th generation Celeron G5920. Models with names beginning with "G" typically belong to the Pentium or Celeron series. AMD also offers entry-level CPUs under the Athlon series, such as the Athlon 3000G, similar to Intel's Pentium-tier CPUs.

 

 

Desktop CPU Model Suffix Meanings:

 

 

K: Intel CPU suffix indicates support for overclocking and integrated graphics, e.g., i5-12600K, i7-12700K.

 

 

F: Intel CPU suffix indicates no integrated graphics, e.g., i5-11400F, i7-11700F.

 

 

KF: Intel CPU suffix, indicates support for overclocking but no integrated graphics, e.g., i5-12600KF, i7-12700KF.

 

 

T: Intel CPU suffix, denotes low-power version, with lower power consumption and slightly reduced performance compared to standard models, e.g., i7-10700T.

 

 

X/XE: Intel CPU suffix, denotes flagship series, e.g., i9-10980XE.

 

 

KS: Intel CPU suffix, e.g., i9-9900K and i9-9900KS, with the latter having higher factory clock speeds than the former, essentially an official overclocked version, e.g., i9-9900KS.

 

 

G: AMD CPU suffix, denotes APU with integrated powerful graphics, e.g., R5 5600G, R7 5700G.

 

 

X: AMD CPU suffix, different from Intel's X suffix, indicates support for XFR (eXtended Frequency Range) technology, allowing for additional overclocking beyond boost frequencies, varying performance based on cooling solutions (air/water/liquid nitrogen).

 

 

XT: Equivalent to non-T version but with slight performance improvements, e.g., R9 3900XT, R7 3800XT, R5 3600XT.

 

 

Notebook Mobile CPU Model Suffix Meanings:

 

 

U: Low voltage, lower performance but lower power consumption, typically found in ultrabooks, e.g., i7 10510U, R7-5700U.

 

 

H: Standard voltage, higher performance, often found in gaming laptops, e.g., i5-11300H, R5-5600H.

 

 

Y: Ultra-low voltage, very weak performance with extremely low power consumption, usually found in thin and light laptops, e.g., i3-10110Y.

 

 

HK: Typically used for high-end overclockable Intel CPUs, e.g., i9-11980HK.

 

 

HX: Typically used for AMD's high-end enthusiast-level CPUs, e.g., R9-5980HX.

 

 

G: G1, G4, and G7, with the number indicating the strength of the integrated graphics, typically, smaller numbers denote integrated UHD graphics, while larger numbers or those equal to or greater than 4 indicate integrated Iris graphics. Intel mobile CPU suffix, e.g., i5-1155G7, i3-1115G4, i3-1005G1.

 

 

HS: Similar to H series but with slightly lower power consumption, often found in versatile thin and light laptops, e.g., R7 5800HS, R5 5600HS.

 

 

HQ: Standard voltage, quad-core CPUs, an older suffix, e.g., i7-7700HQ.

 

 

MQ: Standard voltage, quad-core CPUs, another older suffix, e.g., i7-4810MQ.

 

 

M: An older suffix indicating mobile CPUs, used to differentiate from desktop CPUs, e.g., i7-2620M.

 

 

 

Part II: Understanding CPU Architecture, Clock Speed, Cores, Threads, and Cache

 

 

 

1. CPU Architecture

 

CPU architecture refers to the design framework provided by CPU manufacturers for products within the same series. Generally, architectures differ between brands (Intel and AMD) or generations. Intel and AMD continuously introduce new generations of CPUs with improved and upgraded architectures. Typically, newer architectures offer better performance. We can liken architecture to the logistics within a company for moving goods. An older architecture might use flatbed trucks for transportation, whereas a newer architecture, like a forklift, improves efficiency significantly. Therefore, architectural improvements and upgrades have a substantial impact on CPU performance.

 

2. Manufacturing Process (Process Node)

 

The manufacturing process, or process node, refers to the precision of integrated circuit manufacturing during CPU production. A more advanced process node allows for smaller transistor sizes, enabling more transistors to be integrated into the same area of a wafer, thus enhancing performance while reducing power consumption and heat generation. It also facilitates further upgrades in architecture. For instance, terms like 28nm, 14nm, 10nm, and 7nm (nanometers) represent different levels of manufacturing precision, where smaller numbers indicate better precision.

 

3. Clock Speed (Frequency)

 

CPU clock speed denotes the operating frequency of the cores and represents the CPU's processing speed. Higher clock speeds equate to greater processing power. However, the comparison of clock speeds is limited to CPUs of the same generation. Due to architectural differences, newer architectures may exhibit increased performance even at the same clock speed. For example, Intel's 12th Gen Alder Lake architecture shows a 19% increase in IPC (Instructions Per Clock) performance compared to the 11th Gen Rocket Lake architecture at the same frequency.

 

CPU clock speed comprises base frequency (Base Clock) and boost frequency (Turbo Boost). The base frequency represents the standard operating frequency during light computer use. In contrast, the boost frequency represents the maximum operating frequency reached under heavy loads, such as gaming or running large applications. Turbo Boost intelligently adjusts frequency and voltage to enhance performance based on the current workload. It allows the CPU to operate at its maximum potential during heavy tasks while maximizing energy efficiency during lighter loads. Conversely, overclocking involves manually increasing the CPU's clock speed in BIOS settings to achieve higher performance levels. While overclocking can provide a performance boost of around 5%-10%, it carries the risk of damaging the CPU and requires compatible motherboards adequate cooling solutions, and isn't usually covered under warranty in case of damage.

 

 

4. Cores and Threads

 

Cores represent the processing units within a CPU. Manufacturers progressively increase the number of physical cores to improve multitasking performance, resulting in CPUs with multiple cores, such as quad-core, hexa-core, or octa-core CPUs. Threads, on the other hand, are a technology developed by Intel to simulate two logical cores for every physical core, allowing simultaneous execution of two threads. For example, a quad-core processor with eight threads effectively doubles multitasking performance. To simplify, think of cores as arms: more arms mean more tasks can be done simultaneously. Single-core, single-thread performance is akin to having one arm and one hand, while multi-core, multi-threaded processors simulate having two hands per arm, significantly boosting work efficiency. The more cores and threads a CPU has, the more tasks it can handle concurrently.

 

5. Cache

 

CPU cache is a crucial parameter situated between the memory and the CPU, acting as a faster but smaller storage unit. It alleviates the disparity between CPU processing speed and memory read-write speed. Thus, higher cache capacity is preferred. The principle of cache operation is that when the CPU needs to access data, it first searches the cache. If the data is found, it's immediately retrieved and sent to the CPU for processing, significantly reducing the CPU's memory access time. If the CPU doesn't find the data in the cache, it needs to retrieve it from the slower memory and simultaneously store it in the cache for future access, thereby avoiding memory calls. CPU cache is subdivided into levels: L1 cache, L2 cache, and L3 cache. In actual data retrieval, the most crucial cache is L1 cache due to its fastest speed, followed by L2 cache, and finally L3 cache, which is the slowest. However, L3 cache has the largest capacity. When accessing cache, the CPU first checks L1 cache, then L2 cache, and sometimes retrieves data from L3 cache if it's not found in L1 or L2 cache.

 

If we liken the CPU to a large restaurant kitchen, memory to a warehouse storing ingredients, and cache to a transfer hub between the kitchen and the warehouse, the closest cache to the CPU represents L1 cache, followed by L2 cache, and finally L3 cache (imagine three rooms in the transfer hub: the closest, the middle, and the farthest). When the kitchen needs certain ingredients to prepare a dish, it retrieves the necessary ingredients from the warehouse and temporarily stores them in the transfer hub to avoid fetching them from the farthest warehouse. The size of the cache is equivalent to the area of the transfer hub: the larger the area, the more ingredients it can store. However, if the hub's area is too small to store all the required ingredients, some ingredients must be fetched from the distant warehouse, impacting the overall cooking time. Hence, cache size affects CPU performance to a certain extent.

 

 

6. Integrated Graphics (iGPU)

 

Integrated graphics, formerly known as integrated GPUs, were initially chips integrated into the motherboard. However, modern CPUs from both AMD and Intel integrate graphics chips directly into the CPU. With integrated graphics, even without a dedicated GPU, computers can boot up and operate. However, CPUs without integrated graphics, such as certain Ryzen CPUs with no "G" suffix or Intel CPUs with "F" suffix, require a dedicated GPU for operation. Integrated graphics typically offer performance equivalent to entry-level dedicated GPUs, making them suitable for lightweight games like League of Legends. However, for playing 3D-intensive games, a dedicated GPU is necessary to avoid choppy performance.

 

7. TDP (Thermal Design Power)

 

Generally, lower CPU power consumption corresponds to less heat generation and better energy efficiency. TDP stands for "Thermal Design Power" and reflects the maximum heat dissipation of a processor under full load, measured in watts (W). TDP is not the actual power consumption but rather the maximum heat output. Understanding a CPU's TDP helps in selecting an appropriate cooling solution.

 

8. CPU Instruction Set

 

CPU instruction sets, stored internally, optimize and guide CPU operations. With these instruction sets, CPUs can work faster and more efficiently. Each system command assigned to the CPU relies on pre-set instructions to be completed. The accumulation of these pre-set instructions is collectively referred to as the CPU instruction set. CPUs rely on external instructions to "activate" memory instructions to manipulate and calculate data. Generally, the more pre-set instructions stored, the "smarter" the CPU. Advanced pre-set instructions result in a higher-level CPU.

 

9. CPU Packaging and Interfaces

 

Currently, there are three types of CPU packaging: LGA, PGA, and BGA.

LGA (Land Grid Array): Widely used by Intel for its desktop processors.

PGA (Pin Grid Array): Widely used by AMD for its desktop processors.

BGA (Ball Grid Array): Commonly used for mobile versions of processors, soldered onto the motherboard, making replacement challenging without professional tools.

Intel and AMD CPUs are incompatible due to different packaging methods. Even with the same packaging, different interfaces prevent compatibility. For example, the 12th Gen Intel Core i5 12600K uses LGA1700 packaging, requiring a compatible Intel 600 series motherboard. Previous-gen motherboards like the LGA1200 socket of the 500 series won't be compatible.

 

10. CPU Stepping

 

CPU stepping refers to the product numbering after improvements during the manufacturing process. For example, CPU stepping numbers like A0, B0, B1, C2, U0, etc., indicate the versioning of a CPU. The further along in the alphabet or the higher the number, the newer the product. Stepping numbers change with advancements in production technology, bug fixes from the previous version, or additional features, making different steppings normal for the same CPU model.

 

Part III: Common Questions About CPUs

 

 

1. Difference Between OEM and Retail Boxed CPUs

 

When assembling a desktop computer, we often encounter both OEM and retail boxed versions of the same CPU model. Here are the differences between OEM and retail boxed CPUs:

 

OEM CPUs: These typically come as standalone units without packaging or a bundled cooler. They don't enjoy Intel's official three-year warranty service but are covered by a one-year store warranty. However, starting from the 12th generation of Intel CPUs, OEM CPUs also support three-year store warranty service, though historical models remain covered for only one year.

 

 

Retail Boxed CPUs: These come in proper retail packaging and, in most cases, include a bundled cooler. While some models may not include a cooler, most do. Retail boxed CPUs are eligible for Intel's official three-year warranty service.

 

 

The main difference between OEM and retail boxed CPUs lies in their distribution channels. Retail boxed CPUs typically come from authorized channels, while OEM CPUs often originate from surplus stock from OEM brands, distributed through various channels. Therefore, in terms of quality, they are essentially the same.

 

 

Generally, CPUs are highly precise electronic products, and the likelihood of counterfeit products is extremely low. Thus, many users opt for OEM CPUs to reduce the overall cost of assembling a PC.

 

 

2. Should You Prioritize Clock Speed or Core Count When Buying a CPU?

 

 

The choice between clock speed and core count depends largely on individual needs. For gaming, which often requires straightforward calculations, higher clock speeds are generally favored, as most games are optimized for single-threaded performance. Thus, prioritizing a CPU with a higher clock speed is advantageous. However, for tasks such as multitasking, rendering, or other heavily threaded workloads, a higher core count becomes more important.

 

 

3. Is an i5 Always Better Than an i3, and Is an i7 Always Better Than an i5?

 

 

Generally, within the same generation, an Intel Core i7 processor offers superior performance compared to an i5 or i3. However, with the rapid advancement of technology, architectural improvements, and enhanced manufacturing processes, the performance gap between different CPU tiers within the same generation has narrowed significantly. For instance, the 12th generation Intel Core i5-12600K outperforms the previous flagship i9-11900K from the 11th generation. Therefore, simply categorizing CPUs into i3, i5, or i7 without considering specific models or generations is misleading. The performance disparity between older and newer generations of i3, i5, and i7 processors can be significant.

 

 

4. Does Higher CPU Clock Speed Always Equate to Better Performance?

 

 

Similar to core count, clock speed alone does not determine CPU performance. Other factors, such as architecture, core count, cache, and manufacturing process, also play significant roles. While higher clock speeds may offer better performance within the same generation, comparing CPUs from different generations based solely on clock speed is inadequate. For example, despite similar clock speeds, AMD's Zen 3 architecture achieved a 19% increase in IPC performance compared to the previous Zen 2 architecture.

 

 

5. Is AMD More Cost-Effective Than Intel?

 

 

Cost-effectiveness, or value for money, is determined by the balance between performance and price. AMD has historically employed competitive pricing strategies to gain market share. However, starting with the fourth generation Ryzen processors, AMD has gradually increased prices, deviating from its previous low-price strategy. Meanwhile, Intel has adjusted its pricing to focus on cost-effectiveness. Therefore, determining which brand offers better cost-effectiveness depends on factors such as market strategies and current market conditions.

 

 

6. In Case of a Limited Budget, Is it Better to Prioritize a Strong CPU or a Strong GPU?

 

 

When budget constraints are a concern, prioritizing a strong CPU or GPU depends on individual usage scenarios. In scenarios such as rendering, multitasking, or productivity tasks, where CPU performance is crucial, opting for a stronger CPU with integrated graphics or a lower-end GPU may be sufficient. Conversely, prioritizing a stronger GPU with a mid-range CPU might be preferable for tasks primarily reliant on GPU performance, such as gaming or GPU-intensive applications.

 

 

7. Is There a Significant Performance Difference Between i5 and i7 Processors for Gaming?

 

 

Within the same generation, the performance difference between Intel Core i5 and i7 processors in gaming scenarios is generally minimal. Most modern games do not heavily leverage additional CPU threads beyond a certain point. Therefore, having a higher core count, such as that provided by an i7 processor, may not significantly impact gaming performance. Instead, factors such as single-core performance and GPU capability play a more substantial role in gaming performance. Mid-range CPUs like the 10th or 11th generation Intel Core i5 or AMD Ryzen 5 series can adequately handle gaming requirements.

 

 

How to deal with the used computer parts like the CPU and other components? First, do not let them sit idle, consider contacting an ITAD company to understand their values first. You may be surprised by the return. You can check the examples on how to sell the most common computer parts: sell RAM, sell CPU, sell SSD, and sell GPU. And GPU might be the most value part.