Central Processing Units, Graphics Processing Units, and Computer Memory
A modern computer relies on several hardware components for the processing of data and execution of applications. Three key hardware components include the central processing unit (CPU), graphics processing unit (GPU), and computer memory, which is also known as random-access memory (RAM). While each of these hardware components performs different tasks, their performance is related. The CPU runs general commands and manages most of the computer operations; the GPU executes highly parallel operations and graphics rendering, while RAM temporarily stores the necessary data for the processors. With advancements in computers, these technologies have become faster and more efficient in terms of energy consumption and are increasingly used for performing such complex operations as AI computation, gaming, content creation, and other tasks.
Central Processing Unit
A central processing unit is the main general-purpose processor in a computer. It is tasked with retrieving instructions, understanding them, making the computations needed and coordinating the activities of the other computer hardware components. The functions performed by the CPU include conducting operating system operations, browsing through the Internet, running application programs, handling files, computations, and instructions execution. CPUs nowadays consist of many processing cores, which enable the CPU to perform more than one operation at a time. Information that is accessed often can also be found in cache memory close to the CPU much faster than from the main memory.
An example of a CPU used in modern desktop personal computers is the AMD Ryzen 9 9950X. This CPU is based on the AMD Zen 5 microarchitecture and consists of 16 CPU cores and thirty-two threads. The highest possible clock speed boost that can be achieved by this CPU is 5.7 GHz. Also, the total size of L3 cache in this CPU is 64 MB. Moreover, this CPU supports DDR5 memory and PCI Express 5.0 hardware. The thermal design power recommended by AMD for this CPU equals 170 watts (Advanced Micro Devices [AMD], 2024).
The Ryzen 9 9950X shows some of the strengths of contemporary desktop processors. The strength of having many cores and threads makes it suitable for use in multitasking while the high speed helps in handling applications that need high performance from each core. Another strength is that the processor is unlocked and thus enables users to change the operational aspects to increase the processor’s performance. Some of the weaknesses of the processor include its power and cooling needs. AMD recommends that the processor be cooled using liquid while it has a thermal design power of 170 watts. High-end processors can be costly to have in a system. Initially, AMD marketed the Ryzen 9 9950X as a premium processor whose cost keeps changing with time; currently it is approximately $549.
There are also other big companies which manufacture processors. Apart from AMD, there is Intel, a leading manufacturer of processors for desktop and laptop computers. There are firms such as Apple, Qualcomm, MediaTek, and Samsung which design processors that can be found in diverse types of mobile devices and computer systems. For instance, Intel Core Ultra 9 285K has twenty-four cores that include 8 performance cores and 16 efficient cores, its speed goes up to 5.7 GHz, and it supports DDR5-6400 memory (Intel, n.d.). The recommended retail price of this product is about $589–$599.
CPUs for desktops, laptops, and smartphones are made differently due to the different power, space, and cooling demands that each has. The desktop CPU can run at high efficiency due to the increased power allowance, as well as the ability to have bigger heatsinks and liquid cooling. They are usually found in sockets and hence can be replaced.
The CPUs that run on a laptop must strike a balance between performance and energy consumption. There is much less physical space in a laptop than in a PC, and laptops work on battery when they are not connected to any power source. It is because of this reason that aggressive energy management along with high-performance and efficient cores are used.
In the case of smartphone processors, the trend towards integration is taken even further. Instead of having a dedicated processor for each function of a phone, the smartphone would have a system on a chip (SoC). The system on a chip can integrate the CPU, GPU, neural processing units, memory controllers, graphics processors, and communications into one chip package. As an example, the Snapdragon platform uses its own Oryon CPU and Adreno GPU on a high-performance and low-power platform (Qualcomm, n.d.).
The processors in the iPhone represent another example of how desktop and smartphone processors differ. In the A18 Pro, for instance, a six-core CPU featuring two performance cores and four efficiency cores was introduced. According to Apple, this CPU could deliver a workload 15% faster than the preceding generation while consuming 20% less energy (Apple, 2024). Such innovations are especially relevant for smartphones, as there is no sense in having high performance in case of high energy consumption.
Future developments in CPU technology will be less about boosting clock speed and more about integrating various specialized processors. The present processors have combinations of performance cores, efficiency cores, integrated graphics, and neural processing units. For instance, in Intel Core Ultra 9 285K, there are CPU cores, integrated graphics, and an Intel AI Boost neural processing unit (Intel, n.d.). In the future, CPUs will become more AI-optimized, energy-efficient, have larger caches, better chip fabrication processes, and even further integrated graphics and memory.
Graphics Processing Unit
A graphics processing unit is a type of processor that can execute several calculations at the same time. While GPUs were invented for the purpose of creating computer graphics, their ability to process vast amounts of data in parallel helped them find applications in several other fields. The modern-day uses of GPUs include video gaming, 3D modeling, animation, scientific computing, video editing, cryptocurrency mining, and artificial intelligence. The major difference between a GPU and a CPU is that while a CPU has been customized to execute various kinds of sequential instructions, a GPU can perform similar mathematical calculations in parallel.
An illustration of this is the NVIDIA GeForce RTX 5090. The RTX 5090 features the Blackwell architecture from NVIDIA and has 21,760 CUDA cores, fifth generation Tensor Cores, and fourth generation ray-tracing cores. It has 32 GB of GDDR7 graphics memory on a 512-bit memory interface (NVIDIA, n.d.). NVIDIA released the RTX 5090 with a first price tag of $1,999.
RTX 5090 is an example of the evolution of GPUs from merely producing images on a screen. The former does the job of producing 3D images using its traditional graphics, while the latter takes care of creating a more correct representation of light, reflections, and shadows using ray-tracing hardware. Tensor cores take care of AI calculations while, through NVIDIA’s DLSS technology, AI is used not only to produce better images but also to create new images to allow the game to produce more frames per second (NVIDIA, 2025).
One of the benefits of using a strong GPU is that it can calculate copious amounts of identical data at the same time. It is useful for gaming, artificial intelligence, rendering, simulation, video editing, and any kind of parallel tasks. A strong GPU can also have its own video memory, which means that graphics tasks will not depend completely on the computer’s main memory. The downsides of using a powerful GPU are price, high electricity usage, heat emission, and enormous size. Expensive GPUs may need special cooling and powerful computer power supplies. Besides, there are many ordinary tasks like document preparation or visiting basic websites that do not require such computing abilities.
PC graphics processors come from several vendors, such as NVIDIA, AMD, and Intel. Apple and Qualcomm are among the vendors that design integrated GPUs for use with their processors and mobile platforms. Desktop graphics cards generally deliver the best overall performance in graphics tasks because they have more room to accommodate the necessary hardware, including memory dedicated to graphics processing.
However, GPUs that go into laptops have basically the same architecture, but they are usually designed to consume less electricity and generate less heat. Thus, a laptop variant of a GPU will behave differently from a desktop version of the same GPU despite having identical marketing labels. Designers of laptops must strike a balance between performance and such things as battery consumption, portability, cooling capabilities, and noise level.
The GPU in smartphones is embedded in the SoC instead of being a separate graphics card. The GPU needs to be able to give good graphics even with very stringent power and thermal limitations. In the Snapdragon series of devices from Qualcomm, there are Adreno GPUs embedded in them, and in the Apple A-series, there are Apple-designed GPUs. Apple said that the six-core GPU in the A18 Pro was up to 20% faster than the earlier one and capable of doing hardware-accelerated ray tracing twice as fast (Apple, 2024).
The development of GPUs is increasingly becoming linked to that of AI. Contemporary GPUs no longer have only shader units but are becoming enriched by matrix, tensor, neural, and ray-tracing accelerators. The new NVIDIA RTX 50 series highlights both GPU and AI functionality, and Qualcomm and Apple are also employing dedicated hardware for graphics, machine learning, image processing, and generative AI (Apple, 2024; NVIDIA, 2025; Qualcomm, n.d.). It can be expected that future GPUs will continue to enhance ray-tracing speed, AI acceleration, memory throughput, energy efficiency, and their importance for local AI.
Computer Memory
Computer memory, which is known as RAM, is fast, temporary memory that stores data that is currently being used by the CPU and other computer parts. Unlike permanent memory, RAM is volatile; thus, the data stored in it is lost in case of power loss. The function of RAM is to provide very quick access for the CPU to data and applications that are currently in use, instead of loading everything from SSD or another type of permanent storage.
An example is Crucial Pro DDR5-6400 desktop memory. DDR5 stands for Double Data Rate 5 and is the fifth generation of Double Data Rate Synchronous Dynamic Random-Access Memory (DDR SDRAM). As per Crucial, the latest version not only enhances performance but is denser than the DDR4 version; however, the DDR5 Pro overclocking memory offered by Crucial is available at 6,400 MT/s speeds (Crucial, 2025). Thirty-two gigabytes of memory are sufficient for gaming, multitasking, and various other tasks using 2 x 16-GB modules.
Computer memory manufacturers and brands are, among others, Micron, Samsung, SK Hynix, Kingston, Crucial, Corsair and other companies. It is necessary to separate DRAM chip makers from the assembly or marketing of the final RAM product brands. Crucial, for instance, is a brand of Micron that uses Micron memory technology.
Among the benefits of having adequate RAM are efficient multitasking, faster access to active data, and the ability to run bigger applications without having to rely on slower virtual memory. In addition, fast memory increases efficiency when the processing speed is bottlenecked by the memory bandwidth. However, RAM has its weaknesses as well. It is a type of volatile memory that is non-permanent, faster RAM can be more expensive, and buying more memory beyond the needs of your workload will give little benefit. Moreover, different generations of memory are not fully backward-compatible. Crucial states that DDR5 desktop memory chips have 288 pins and are not backward compatible with DDR4 and DDR3 memory slots (Crucial, n.d.).
There is an array of factors affecting the price of RAM, which include but are not limited to the capacity, speed, vendor, availability on the market, and the store selling the product. The more advanced the DDR5 modules are, the higher their price. Hence, one should treat price as something that changes from time to time rather than something constant. One should not choose RAM based only on the speed that is claimed by the vendor. For example, the Crucial Pro DDR5-6400 32 GB kit can cost approximately $390, although the price may vary depending on the retailer and market availability.
There are two differences between desktop and laptop memory. Firstly, they have distinctive designs. Secondly, there are differences in energy they need. While desktop computers are usually supplied with standard DIMM memory, laptops are typically equipped with SO-DIMMs. In the case of some ultrabooks, they do not feature any removable memory anymore. The RAM can be installed on the board right next to the processor.
Smartphones usually implement low-power memory like LPDDR instead of desktop DDR DIMMs. This is because smartphone memory must take up minimal physical space and consume minimal power as the battery life is a critical parameter. Replacement of memory in smartphones is also not envisaged for the users. The modern-day mobile CPUs also have extremely high requirements of memory bandwidth as multiple units like CPU, GPU, neural processors, camera processors, and others might access the same memory. According to Apple, the A18 Pro generation improved the memory bandwidth by 17%, which helped AI applications and its GPU (Apple, 2024).
Future memory technologies are expected to continue to emphasize more bandwidth, capacity, and power efficiency. The DDR5 generation has already offered better performance than prior generations, while dedicated technologies like low-power LPDDR memory and high-bandwidth memory serve computing conditions. JEDEC, an organization behind many semiconductor memory standards, is currently developing standards related to DDR SDRAM, high-bandwidth memory, low-power LPDDR, and other storage and memory technologies (JEDEC, n.d.). With the rise of AI and data-intensive technologies, it is crucial to be able to deliver vast amounts of data to processors fast enough.
How the CPU, GPU, and Memory Work Together
While the CPU, GPU, and memory all have distinct roles, the efficiency with which they run together will define the efficiency of the computer. First, the program needs to be loaded into RAM from the permanent storage. Then, the CPU takes the instructions and data from memory and implements the overall logic of the program. When the workload includes graphics rendering or some other tasks that are highly parallelizable, the CPU can delegate the tasks to the GPU. Then, the GPU computes the results.
Gaming is an example which clearly shows this relation. The CPU can process game logic, artificial intelligence of the characters, physics commands, and inputs from the user. The data in use during the gaming process is stored in RAM so that the CPU can access it easily. The GPU processes all information related to graphics including texturing, geometry, and lighting to provide the frame that appears on the screen.
The computer will suffer from a performance bottleneck if one part is much slower compared to other components. For instance, an incredibly powerful GPU will be unable to function at its full capacity if the CPU is slow in preparing the data. On the other hand, if the computer lacks enough RAM, then a fast CPU will face constraints since the computer will continuously have to transfer the data back and forth between RAM and the hard drive. The graphics workload might also suffer from a bottleneck due to insufficient graphics memory in the GPU.
Technological Advances and Future Development
The CPU, GPU, and memory technologies have improved considerably since current computing demands something more than just the operation of basic software on desktops. The CPUs have gone through a transformation from having a single core architecture to architectures having multiple cores and even combinations of cores. The GPUs, which used to be simple graphics processors, have become powerful parallel processing units for AI and scientific computations.
Among the most notable common trends for all three types of technology is specialization. Contemporary computers can have general CPUs, GPUs for parallel calculations, neural processing units, and media processing engines that decode and encode videos in them. This trend is illustrated by smartphones whose SoCs feature multiple processing systems on one small platform. The concept is now being followed by desktop and laptop processors.
Energy efficiency will also continue to be an important aim. The improvement of performance by simply using more electricity results in difficulties about the issues of battery life, cooling, noise, and operating costs. Progress in semiconductor fabrication technologies, processor designs, memory technology, and workload acceleration has made it possible to achieve improvements in performance without completely depending on increased electricity use.
Conclusion
The CPU, GPU, and computer memory are three basic components that play a significant role in defining how well a modern computer functions. The CPU acts as the main processing unit, the GPU acts as a specialized graphics and parallel computing unit, and RAM acts as fast temporary storage for programs and data. Examples of products in this area include the AMD Ryzen 9 9950X, the NVIDIA GeForce RTX 5090, and the Crucial DDR5 memory.
The disparities between desktops, laptops, and smartphones’ hardware further prove the fact that computing’s performance should not be assessed through its speed alone. Whereas desktops have the advantage of consuming more power and employing bigger cooling mechanisms to achieve maximum sustained performance, laptops must consider computing power as well as portability and battery life. Even more so, smartphones focus on efficiency and integration because their CPU, GPU, AI, and other hardware all go into a system on a chip (SoC). With the ever-evolving computing world and its direction toward AI, graphics, and data-driven applications, the development of CPUs, GPUs, and memory is bound to continue together.
 
References
Advanced Micro Devices. (2024). AMD Ryzen 9 9950X. AMD. https://www.amd.com/en/products/processors/desktops/ryzen/9000-series/amd-ryzen-9-9950x.html
Apple. (2024, September 9). Apple debuts iPhone 16 Pro and iPhone 16 Pro Max. Apple Newsroom. https://www.apple.com/newsroom/2024/09/apple-debuts-iphone-16-pro-and-iphone-16-pro-max/
Crucial. (n.d.). DDR5 RAM: Desktop and laptop memory upgrades. Crucial. https://www.crucial.com/catalog/memory/ddr5
Crucial. (2025, October 21). Crucial unleashes its most powerful gaming memory yet: DDR5 Pro OC 6400 CL32. Micron Technology. https://investors.micron.com/news/press-release/2025/Crucial-Unleashes-Its-Most-Powerful-Gaming-Memory-Yet-DDR5-Pro-OC-6400-CL32-10-21-2025/default.aspx
Intel. (n.d.). Intel Core Ultra 9 processor 285K specifications. Intel Corporation. https://www.intel.com/content/www/us/en/products/sku/241060/intel-core-ultra-9-processor-285k-36m-cache-up-to-5-70-ghz/specifications.html
JEDEC Solid State Technology Association. (n.d.). Technology focus areas. JEDEC. https://www.jedec.org/
NVIDIA. (n.d.). GeForce RTX 5090 graphics cards. NVIDIA Corporation. https://www.nvidia.com/en-us/geforce/graphics-cards/50-series/rtx-5090/
NVIDIA. (2025, January 6). NVIDIA Blackwell GeForce RTX 50 Series opens new world of AI computer graphics. NVIDIA Newsroom. https://nvidianews.nvidia.com/news/nvidia-blackwell-geforce-rtx-50-series-opens-new-world-of-ai-computer-graphics
Qualcomm. (n.d.). Snapdragon 8 Elite mobile platform. Qualcomm Technologies. https://www.qualcomm.com/smartphones/products/8-series/snapdragon-8-elite-mobile-platform