Have an appetite and time to delve into the battle of the CPU giants ?, If yes then let's go. Initially, smartphones are capable of operating with three different types of processor architecture: ARM, Intel, and MIPS. Certainly, ARM is at the forefront of today's most used smartphone processor architecture. In 2013, Intel stated that it was completely out of this market. MIPS has long been out of service. It seems that the arena has become completely empty for ARM architecture, but after its fruitful success in the smartphone market, do you think it will suffice with that?
 |
| The difference between Intel X86, ARM and clash of silicon chip giants |
Like the Octopus with Eight Arms and Three Hearts, ARM architecture is trying to break into the PC market, in fact it is able to end the life of X86 overnight - we know that the only advocates for X86 architecture so far are AMD and Intel - so getting rid of them won't be like that. Ease. We were very impressed with the reaction of the tech community when they confirmed that Apple abandoned Intel and the X86 and was able to develop its own processor chips for Macs that are based on ARM architecture. Why do we think that Apple did miracles? This question has not been answered, perhaps because it is Apple that created it and there are many fans and enthusiasts who support the American uncle and love him relentlessly - this is not the focus of our conversation - but ...
The war between Intel and ARM
The strange thing is that the ARM architecture has been working with personal computers and Windows systems for several years thanks to Qualcomm chips that we have seen in very thin computers such as the Surface Pro X and even with the ChromeOS system such as the Lenovo Mixx 630 laptop, as these computers were equipped with the Snapdragon 850 chip that is based on the ARM architecture .
And now we have just seen the M1 chip that Apple launched for Macs, which is also based on the ARM architecture. This time it's a real slap in the face of the x86 architecture and a serious threat to the competitors behind the x86, namely Intel.
A war is inevitably coming between Intel and ARM - AMD need not be mentioned now because the only whale to swallow this market up to this point with a market share of at least $ 61 billion is Intel. Intel’s concern here is whether laptop manufacturers try to rely on ARM architecture, just as Apple did and abandoned the X86 architecture. But what is the difference between ARM and X86 architecture and why do we think that there is a war coming very soon between the chipset giants.
And now we have just seen the M1 chip that Apple launched for Macs, which is also based on the ARM architecture. This time it's a real slap in the face of the x86 architecture and a serious threat to the competitors behind the x86, namely Intel.
A war is inevitably coming between Intel and ARM - AMD need not be mentioned now because the only whale to swallow this market up to this point with a market share of at least $ 61 billion is Intel. Intel’s concern here is whether laptop manufacturers try to rely on ARM architecture, just as Apple did and abandoned the X86 architecture. But what is the difference between ARM and X86 architecture and why do we think that there is a war coming very soon between the chipset giants.
A look at x86 and ARM architecture
Intel X86 VS ARM
Although the CPU / Central Processing Unit is the brain of a computer, it is not that clever. They only work if they are provided with very specific information that we symbolize by the term code or Instruction Sets. These instructions instruct the central processor to transfer data between Registers and RAM, or in order to perform a range of analyzes and computational operations such as multiplication and subtraction.
The types of central processor blocks, which we refer to as CPU blocks, differ in the types of instructions that they receive in order to work on them. Processor blocks become more complex and powerful in high-performance central processors, or in a more accurate sense in high-power central processors.
The types of central processor blocks, which we refer to as CPU blocks, differ in the types of instructions that they receive in order to work on them. Processor blocks become more complex and powerful in high-performance central processors, or in a more accurate sense in high-power central processors.
Applications running on smartphones are not written or programmed into CPU instruction windows - as this would be completely irrational with huge applications running across various devices and different chips, but instead are written in high-level programming languages like Java or + C + to work with specific code that is capable of operating and interacting either while using either the ARM processor architecture or the X86 processor architecture. These instructions are decoded in microcoding processes inside the central processor, which tend to use a specific area of the silicon chip and specific proportions of power.
To summarize all the complicated words into an understandable sentence: If we want a highly energy efficient CPU, the instruction windows of the CPU should be simple and uncomplicated, just as we get with the ARM architecture used in smartphones. If we want to get strong performance from complex instruction windows in the central processor, this is possible, but at the expense of consuming higher amounts of energy, such as the X86 architecture in which the processing chips of desktop and server devices are developed. This is the first and most important difference between ARM processor architecture and the X86 processor.
To summarize all the complicated words into an understandable sentence: If we want a highly energy efficient CPU, the instruction windows of the CPU should be simple and uncomplicated, just as we get with the ARM architecture used in smartphones. If we want to get strong performance from complex instruction windows in the central processor, this is possible, but at the expense of consuming higher amounts of energy, such as the X86 architecture in which the processing chips of desktop and server devices are developed. This is the first and most important difference between ARM processor architecture and the X86 processor.
The ARM architecture is based on what is known as RISC operations, which is an abbreviation for Reduced Instruction Set Computing, or as we call it Reduced Instruction Set Computing, while the X86 architecture is based on what is known as CISC, which is an acronym for Complex Instruction Set Computing and we call it Complex Instruction Set Computing. Help windows in ARM are atomic or atomic, with a very close association between the number of instructions and the exact processes. While CISC, in turn, offers many instructions that have the ability to perform multiple and more complex operations such as improved mathematics and traffic interaction - but this is at the cost of consuming higher amounts of energy to decode these operations.
This link between the instruction windows and the CPU hardware design is what makes the architecture and architecture of the CPU, and this means that different architectures of the CPU are designed according to different goals and purposes, either for multitasking and commands or for higher efficiency and energy consumption. By reducing the silicon wafer area to the lowest possible level. This is the second major difference between ARM and X86 architecture. The first is based on lower energy consumption, the number of reduced instruction windows, and the count of the minimum hardware components, and the second is based on the type of specific uses for each of the two architectures.
This link between the instruction windows and the CPU hardware design is what makes the architecture and architecture of the CPU, and this means that different architectures of the CPU are designed according to different goals and purposes, either for multitasking and commands or for higher efficiency and energy consumption. By reducing the silicon wafer area to the lowest possible level. This is the second major difference between ARM and X86 architecture. The first is based on lower energy consumption, the number of reduced instruction windows, and the count of the minimum hardware components, and the second is based on the type of specific uses for each of the two architectures.
Today, 64bit architectures have become the predominant architecture in both smartphones and computers, their purpose is to enhance the records and addresses of RAM used enough to use various types of lengthy data (0s + 1s) along with help windows and compatible hardware components for these operations. This also means that you will need an operating system that is compatible with 64-bit architecture, such as Android or Windows operating systems. In addition, the 64bit architecture aims to improve 3D display accuracy, encryption speed, and simplify addressing or processing over 4GB of RAM capacity.
Computers moved to 64-bit architecture long before smartphones, but in fact it was not Intel that helped reformulate the X86-64 architecture, which we also know as X64, but AMD was the first to announce the X64 architecture in 1999 when it worked on Improved X86 architecture, at this time Intel's IA64 Itanium architecture has taken a back seat.
Computers moved to 64-bit architecture long before smartphones, but in fact it was not Intel that helped reformulate the X86-64 architecture, which we also know as X64, but AMD was the first to announce the X64 architecture in 1999 when it worked on Improved X86 architecture, at this time Intel's IA64 Itanium architecture has taken a back seat.
ARM introduced the ARMv8 64-bit architecture in 2011 for the first time rather than simply expanding the 32bit instruction set. In fact, ARM builds a flawless 64bit architecture, meaning that the ARMv8 architecture uses two different instruction window states, AArch32 and AArch64, one for 32bit and the other for 64bit code. The strength and beauty of the ARM architecture design shines in that it is able to switch smoothly from one mode to another while executing the instructions, and this means that the 64bit instruction decoder is a completely new design that does not need to maintain compatibility with the 32bit era, and in this way the processor has full compatibility capacity With previous versions thanks to AArch32 windows.
The architectural differences we just discussed partly illustrate the current successes and at the same time the issues that the chip giants face. ARM method is ideal for low-power use that perfectly suits the requirements of thermal design targeting a value of 3.5W TDP for smartphones. On the other hand, Intel's Core I processor, with a 100-watt thermal design, is the most popular central processor used in high-performance desktop computers, servers and business stations, but has been struggling over the past years to reduce its size and reach thermal designs that do not exceed 5 watts.
Of course, we won't forget the role that silicon chip manufacturing has played in dramatically improving energy efficiency over the past decade.Smaller CPU transistors consume the least power, which was the reason for the tech community's indignation when they learned that Intel was stuck with a 14nm manufacturing node a year ago. 2014 up to the present - at this time (2014) smartphone chip transistors were developing at lightning speed, suddenly jumping from 20 nanometers to 14 nanometers and then to 10 nanometers and we are now dealing with processors with 7 nanometers transistors.
The architectural differences we just discussed partly illustrate the current successes and at the same time the issues that the chip giants face. ARM method is ideal for low-power use that perfectly suits the requirements of thermal design targeting a value of 3.5W TDP for smartphones. On the other hand, Intel's Core I processor, with a 100-watt thermal design, is the most popular central processor used in high-performance desktop computers, servers and business stations, but has been struggling over the past years to reduce its size and reach thermal designs that do not exceed 5 watts.
Of course, we won't forget the role that silicon chip manufacturing has played in dramatically improving energy efficiency over the past decade.Smaller CPU transistors consume the least power, which was the reason for the tech community's indignation when they learned that Intel was stuck with a 14nm manufacturing node a year ago. 2014 up to the present - at this time (2014) smartphone chip transistors were developing at lightning speed, suddenly jumping from 20 nanometers to 14 nanometers and then to 10 nanometers and we are now dealing with processors with 7 nanometers transistors.
All expectations indicate that the 5nm manufacturing accuracy will come in 2021, in fact we started seeing news about 3nm here and there from time to time, this continuous conflict and unprecedented development in the chip transistor was due to the field of fierce competition between both Samsung and TSMC.
However, what was characteristic of ARM architecture over the past years and its ability to maintain energy efficiency levels at their best for smartphones was thanks to what is known as heterogeneous compute, which we know as the term heterogeneous compute, the basic idea of which is the ability to establish an architecture It allows different parts of the CPU to work together in a typical environment that can access the processor chip to the best levels of energy efficiency and highest performance.
ARM made a point over the X86 when it announced the big-LITTLE core in 2011 in favor of both the large-core Cortex-A15 processor and the small-core Cortex-A7 processor, this idea was to handle heavy applications and heavy workloads in a smarter and smarter smooth way. At the same time, increased energy efficiency levels for managing background tasks and commands.
ARM made a point over the X86 when it announced the big-LITTLE core in 2011 in favor of both the large-core Cortex-A15 processor and the small-core Cortex-A7 processor, this idea was to handle heavy applications and heavy workloads in a smarter and smarter smooth way. At the same time, increased energy efficiency levels for managing background tasks and commands.
Certainly, it took ARM several years to reformulate this idea to appear at its best in 2017 when the ARMAv8.2 architecture and DynamicIQ technology were announced that allowed more than one CPU in one location and shared RAM resources in order to achieve a modus operandi. More efficient processing. DynamicIQ capability enables 6 + 2 CPU design, a design that we see up to now in mid-range phones.
Speaking of competitor Intel chips, they are without heterogeneous compute and cannot challenge ARM chips or even stand up to them when it comes to performance and energy efficiency levels. It took Intel several years to announce in 2020 the Lakefield chip with a 10nm transistor-sized manufacturing node that combined a high-performance Sunny Cove core with four high-efficiency Termont cores. However, it has not reached the efficiency of smart phone chips, as Lakefield chips operate at a rate of consumption of 7 watts and are very high for smartphones, instead they are ideal for laptops.
Today, the conflict between ARM and X86 is increasing in the world of laptops operating at rates less than 10 watts, and every day, Intel's share of X86 shrinks and the share of ARM increases. Apple's recent move to ARM instead of X86 is one of the most common examples that attest to this, and all this thanks to heterogeneous compute compute features that are supported by the ARM architecture, along with the engineering improvements that Apple has assigned to the new Mac computers.
Today, the conflict between ARM and X86 is increasing in the world of laptops operating at rates less than 10 watts, and every day, Intel's share of X86 shrinks and the share of ARM increases. Apple's recent move to ARM instead of X86 is one of the most common examples that attest to this, and all this thanks to heterogeneous compute compute features that are supported by the ARM architecture, along with the engineering improvements that Apple has assigned to the new Mac computers.
Another standard or fundamental difference between ARM and Intel is that Intel is responsible for manufacturing its own chips from start to finish and then selling them directly, while the other sells a license and offers a range of different products to partners such as Apple, Samsung and Qualcomm. Process chips are manufactured collaboratively through ARM CXC software, which allows partners to build a custom CPU and even make some adjustments to the instruction windows inside the chips. Certainly, building a CPU is a very expensive and complicated process, but doing it right is a fruitful success and ultimately yields huge profits.
Apple was able recently to show the world how superior ARM is and how it can stand up to and even outperform the X86. It has decided to abandon Intel and replace the Core I processor with its own M1 SoC chip, based on ARM. This chip has become in each of the Macbook Air / Macbook Pro / Mac mini computers that are small in size "13inch", and are characterized by great improvements in the level of performance. This means that ARM has already taken over the tasks of X86 and regains from them all the exclusive advantages that were unique to the X86 over many years in the more complex intensive computing, but it must be noted that the Apple project is based only on laptops and not on desktop computers.
Until this moment we are talking with you, the most powerful computer in the world, the Fugaku Supercomputer, which is built on the basis of ARM by the hands of Fujitsu engineers, is the first of its kind that works with ARMv8-A SVE architecture, not the X86. But Intel is still the leader in consumer performance. However, ARM competition is becoming more and more dangerous day by day in many sectors in which high performance factors and low energy consumption play a prominent role.
Applications and programs
It is assumed that the applications are not the source of the annoyance, but it is necessary to configure and program them to run for the specific CPU architecture. In the past, things were a steady and smooth hierarchy, as each CPU was configured to handle and naturally compatible with operating systems, such as Android with ARM architecture and Windows with X86 architecture. The applications did not need to be configured to run across different operating systems. But the growth in apps replaces that idea entirely.
If we look at some of the current platforms such as the Mac system, ChromeOS, and even Windows, they are now running through ARM architecture. This means that applications must be ready to compatible and run in X86 and ARM as well. Perhaps this is the main reason that application developers rely on emulating the code for their programs - that is, translating the bundled code to work with a central processor architecture that can run with another central processor architecture, instead of building and programming applications from scratch, in order to save time and effort. At the same time, however, this means that apps that run through emulators will be less efficient and performer than the original apps.
In recent times, the Windows system emulation on ARM architecture has been in good condition. On the other hand, applications with Chromebook systems run on the Intel X86 processor architecture very naturally. Apple has its own tool called Rosetta 2 for translating old apps and supporting it for Mac. However, no one can deny that the three systems suffer performance penalties when dealing with applications that are no longer specifically programmed to work with any of their systems.
Finally
ARM is in the lead and is gaining a large share when it comes to the energy efficiency required of devices, such as smartphones. At the same time, Intel has tried to enhance the energy efficiency of laptops, and this has become evident through the Lakefield architecture, which is a true competition for ARM architecture in the only playing field in which it plays. The future is still unclear whether the ARM architecture will be able to compete with the X86 in the computing-intensive sector or not, but clearly it will remain the favorite in the smartphone sector.
If we look at some of the current platforms such as the Mac system, ChromeOS, and even Windows, they are now running through ARM architecture. This means that applications must be ready to compatible and run in X86 and ARM as well. Perhaps this is the main reason that application developers rely on emulating the code for their programs - that is, translating the bundled code to work with a central processor architecture that can run with another central processor architecture, instead of building and programming applications from scratch, in order to save time and effort. At the same time, however, this means that apps that run through emulators will be less efficient and performer than the original apps.
In recent times, the Windows system emulation on ARM architecture has been in good condition. On the other hand, applications with Chromebook systems run on the Intel X86 processor architecture very naturally. Apple has its own tool called Rosetta 2 for translating old apps and supporting it for Mac. However, no one can deny that the three systems suffer performance penalties when dealing with applications that are no longer specifically programmed to work with any of their systems.
Finally
ARM is in the lead and is gaining a large share when it comes to the energy efficiency required of devices, such as smartphones. At the same time, Intel has tried to enhance the energy efficiency of laptops, and this has become evident through the Lakefield architecture, which is a true competition for ARM architecture in the only playing field in which it plays. The future is still unclear whether the ARM architecture will be able to compete with the X86 in the computing-intensive sector or not, but clearly it will remain the favorite in the smartphone sector.
Meanwhile, the X86 architecture continues to play a prominent role in the multitasking server and workstation segment. But the only thing that appears to be seen is that the future holds many surprises in terms of efficiency and performance factors for future engineering architectures.
Comments
Post a Comment