Among all participants in the computing field, Intel is one of the oldest and largest companies. Whether Intel is as dominant as it was in the decade of the 21st century or as struggling as it was in the 1920s, it is difficult for people to love the status quo or a large company that has lost to small companies. On the contrary, Intel's competitor AMD has always been the weak, but today it is becoming a "giant".

If Intel were not a hot new innovation tycoon, it would not be a giant today. From time to time, Intel tries to change the situation in the CPU field to make it better. After years of development, do you know the best CPU processors Intel has ever had?

Intel CPU Processor

Intel 8086: Intel becomes the leader

Intel 8086 basically meets all the requirements of CPU: this is a huge commercial success, it represents a major technological progress, and it is the ancestor of all x86 processors. In fact, the x86 architecture is named after this chip.

Although Intel claims that the 8086 is the first 16 bit processor ever launched, this is only a very specific warning. The trend of 16 bit computing emerged in the 1960s, using multiple chips to form a complete processor capable of 16 bit operations. The 8086 is not even the first single chip processor with 16 bit processing power, because other CPUs have been defeated by General Instrument CP 1600 and Texas Instruments TMS 9900. In fact, the 8086 was hastily launched to make Intel equal to its competitors. After 18 months of development, it finally came out in 1978.

At first, the sales of 8086 were not good due to the pressure of 16 bit processors from competitors. In order to solve this problem, Intel decided to gamble and start large-scale advertising for its CPU. Intel, code named "Operation Smash", earmarked $2 million for advertising through seminars, articles and sales plans. The campaign has achieved great success. 8086 has been used in about 2500 designs, the most important of which can be said to be IBM personal computers.

The IBM personal computer (the original PC) launched in 1981 was equipped with Intel 8088, a cheaper variant of 8086, which soon occupied the entire home computer market. By 1984, IBM's personal computer revenue was twice that of Apple, and its device market share ranged from 50% to more than 60%. When IBM PS/2 came out, the 8086 itself was eventually used together with other Intel CPUs.

The great success of IBM personal computers and the promotion of Intel 8086 series CPUs are extremely important to the process of computing history. Because the 8086 is such a popular device, Intel certainly hopes to iterate on its architecture rather than create a new one. Although Intel has produced many different micro architectures since then, the most important x86 instruction set architecture (or ISA) has existed since then.

Another consequence is accidents. IBM asked Intel to find a partner who can produce more x86 processors in case Intel is short of production. The company Intel cooperated with was none other than AMD, which was only a small chip manufacturer at that time. Although Intel and AMD were partners at the beginning, AMD's ambition and Intel's unwillingness to give up led to a conflict between the two companies until today.

Celeron 300 A: the most economical CPU

In the 20 years after 8086, the modern PC ecosystem began to emerge. Fans built their own machines with ready-made components, just as we do today. By the end of the 1990s, it was obvious that if you wanted to assemble a PC, you needed Windows, which could only run on x86 hardware. Of course, Intel became a very dominant figure in personal computers because there were only two other companies with x8 6 licenses (AMD and VIA).

In 1993, Intel launched the first Pentium CPU and will launch this brand of CPU in the next few years. Each new Pentium is faster than the previous one, but these CPUs are not particularly eye-catching and certainly not as influential as the 8086. This is not to say that these early Pentium is poor, they just meet the standard expectations. All this was good until AMD launched its K6 CPU, which provided similar performance levels, and the price of Pentium CPU was lower. Intel had to respond to AMD and introduced a new CPU family: Celeron processor.

At first glance, Celeron CPU seems to have nothing better than cutting Pentium with a lower price tag. But overclocking these chips has turned them into mature Pentium. Mendocino based CPUs (not to be confused with AMD's Mendocino based APU) are particularly popular because they have L2 cache like high-end Pentium CPUs (although not so much).

In Mendocino movies, 300 A is the slowest, but it can overclock to an extreme degree. It can get its 450 MHz, a 50% overclocking. Intel's 450 MHz Pentium II was priced at about $700 at that time, while the Celeron 300A was priced at $180, which made Celeron very attractive to those who can handle the performance slightly lower due to less L2 cache.

Core 2 Dual Core E6300: Intel Counterattack

Although Intel had a very powerful chip empire in the late 1990s, it began to have problems since 2000. This year, Intel launched the Pentium 4 with NetBurst architecture. Through NetBurst, Intel decided to quickly improve the clock speed is the way forward; Intel even plans to reach 10 GHz by 2005. As for the company's server business, Intel launched Itanium, the world's first 64 bit x86 architecture. For Intel, she hopes everyone can use the server CPU.

Unfortunately for Intel, this strategy soon failed, because NetBurst obviously could not reach the clock speed that Intel thought. Itanium is also not doing well. Even though it is the only 64 bit CPU, its adoption is very slow. AMD seized this opportunity, and Intel began to lose market share of desktops and servers rapidly. Part of Intel's response was simply to bribe OEMs not to sell systems using AMD, but Intel also knew that it needed a competitive CPU, because the company could not always pay Dell, HP and other companies billions of dollars.

Intel finally launched its Core 2 series CPU in 2006, completely replacing all NetBurst based desktop and mobile CPUs, as well as the original Core CPU for laptops launched earlier this year. These new CPUs not only bring a completely improved architecture (the core architecture is almost similar to NetBurst), but also the first quad core x86 CPU. Core 2 not only makes Intel equal to AMD, but also makes Intel completely return to the leading position.

Although the high performance of high-end Core 2 CPU (such as Core 2 Premium X6800 and Core 2 Quad Core Q6600) is amazing, there is one CPU that really impresses everyone: Core 2 dual core processor E6300. The E6300 is a dual core and decent overall performance, but like the 300 A, it is a great overclocking.

Core 2 series and Core architecture have re established Intel's leading position in technology, which has never been seen since the 1990s. At the same time, AMD is hard to catch up, let alone remain competitive; It didn't even launch its own quad core CPU until 2007. However, Core 2 is just the beginning, and Intel does not want to slow down. At least not yet.

Intel CPU Processor

Core i5 - 2500 K: beyond AMD's reach

Unlike NetBurst, Core is not a dead end, which allows Intel to iterate and improve the architecture in each generation. At the same time, the company is also steadily building new manufacturing processes or nodes. This leads to the "tick stock" model, where "tick" represents process improvement and "stock" represents architecture improvement. The first batch of Core 2 CPUs is a stock (because they use the same 65nm process as NetBurst), and the later Core 2 CPUs are a tick, because they are made on the 45nm process.

By 2011, Intel had gone through two complete cycles and introduced better and better CPUs, just like clocks. At the same time, AMD is going through an extremely difficult period. Its Xinlong chip has finally brought a lineup of four cores (later six cores) to AMD, but these CPUs have few (if any) performance leaders, and AMD has returned to its old value oriented strategy. When Intel launched its second generation CPU in 2011, AMD was under great pressure.

The second generation Core CPU, code named Sandy Bridge, is a whole. In addition to increasing the frequency itself, it also significantly increases the number of instructions per clock (IPC). The end result is a 10-50% performance improvement over the first generation CPU. Sandy Bridge also has a pretty good integrated graphics card, and is the first CPU to introduce fast synchronization (a video coding accelerator).

At the same time, AMD is far behind; AMD needs to introduce a CPU that can compete with Sandy Bridge if it wants to be more than a budget substitute. Later in 2011, AMD finally launched a new FX series based on bulldozer architecture.

AMD performed badly. The flagship FX-8150 is sometimes comparable to the Core i5 - 2500 K, but it is generally slow, especially in the evaluation of single thread performance indicators; Sometimes it even loses to the old Phenom CPU.

Core i7 - 8700 K: Intel keeps pace with the times

Although Sandy Bridge is great, it foreshadows the dark age of PC users, who always expect that the next generation will be faster and cheaper than the previous generation. However, due to the withdrawal of AMD, Intel has no reason to provide better CPUs at a lower price. In the next six years, Intel only provided four cores on its mainstream platforms, and the price remained the same: i5200 dollars and i7300 dollars. Before 2017, all i3 level processors had no overclocking support, and most i5 and i7 were treated the same way before long.

At the beginning of 2017, when Intel's seventh generation of Kaby Lake came out, the situation became very frustrating. According to the tick stock model, Intel should launch a 10nm CPU with a structure similar to the 14nm 6th generation Skylake CPU from 2015. On the contrary, the seventh generation CPU is identical to the sixth generation CPU: the same 14nm process, the same Skylake architecture. To this end, Intel announced the end of the tick stock model and launched the process architecture optimization model, of which the seventh generation is the optimization model. People's dissatisfaction with Intel is understandable, because even the improvement of a generation is coming to an end.

It is ultimately up to AMD to change this situation and shake things. It must have done so. Only a few months after it launched the Reelung, the seventh generation CPU came out. Based on the new Zen architecture, the Ruilong 1000 CPU, with its outstanding single thread performance and extremely high multithreading performance, brought eight high-performance cores into the mainstream for the first time, and finally brought AMD back to the game. Intel's competitor, the seventh generation, does lead in single threaded applications and games, but it is not enough to make Zen a new bulldozer. For the first time in many years, Intel was forced to provide new products that were truly valuable.

Intel attaches great importance to Ryzen and launches a new generation of products as soon as possible. The seventh generation lasted only 9 months and was replaced by the eighth generation of Coffee Lake. This is another optimization of Skylake, but the clock speed is higher, and more importantly, there are more cores. Core i7 CPU now has 6 cores and 12 threads, Core i5 has 6 cores and 6 threads, and Core i3 has 4 cores and 4 threads (the same as the old i5). But one thing has not changed, that is, the price, which means that the value of the eighth generation is much higher than the previous Core CPU.

The Core i7 - 8700 K is equipped with 7700 K fast single thread performance and two additional cores, which is the best flagship product of Intel for many years. Compared with AMD's Reelung 7 1800 X, the 8700 K is only a little behind in the multithreading benchmark, and far ahead in other aspects.

However, Intel's prospects are bleak. In the eighth generation CPU, the process architecture optimization model has failed because the eighth generation CPU is the second consecutive optimization. When the 10nm Cannon Lake CPU finally came out in 2018, it was clear that Intel's latest technology was extremely old. How many more optimizations will Intel have to go through before finally launching new products?

Facts have proved that there are many.

Core i9 - 12900 K: a badly needed return

In 2018, 10nm is only suitable for barely running mobile chips. In 2019, Intel launched the mobile Ice Lake CPU, and the situation has improved, but these are just four cores and good integrated graphics cards that are far below the desktop level. In 2020, with the launch of the 11th generation Tiger Lake processor, the situation will improve again. This is the optimization of Ice Lake. It has better graphics, but it is still not a good desktop.

Intel urgently needs 10nm desktop CPU. Its 14 nanometer process is very old, which hinders the increase of the number of cores and clock speed. In contrast, AMD has made great progress on the Sharp Dragon 3000 Zen 2 CPU and the Sharp Dragon 5000 Zen 3 processor, one by one. Now AMD has even taken the crown of game performance from Intel. It needs a big way to make a comeback.

Finally, at the end of 2021, Intel launched its first 10 nanometer CPU for desktop computers: the 12th generation Alder Lake. These CPUs are completely different from the previous ones; Its hybrid architecture combines a powerful large performance kernel (or P kernel) with a smaller, more efficient and efficient kernel (or E kernel), providing incredible multithreading performance for top chips, and significantly improving the single threaded performance for all other components.

Intel's new flagship Core i9 - 12900 K uses a core configuration of 8 P cores and 8 E cores, which is excellent for both multithreaded and single threaded tasks. In Big Eye's assessment, it was found that 12900 K not only made Intel equal to AMD, but also firmly led in every indicator. The Ruilong 95950 X, launched as an expensive and high-quality flagship, suddenly looks like a budget substitute, but the 12900 K is much cheaper. Describing Alder Lake as a regression is a understatement.

The only disadvantage is that 12900 K (similar to Alder Lake) consumes a lot of power. One indication is that 10nm is not ready for prime time. Nevertheless, the resumption of competition has a very positive impact on almost everyone. The price of the Sharp Dragon 5000 CPU has dropped to match Intel and AMD's new model, which responds to budget buyers. The low-end Alder Lake CPU, such as the core i5-12400, costs only $100, which is cheaper than the 5600 X, but it is also significantly faster. Alder Lake proved once again that we need both Intel and AMD to compete, otherwise PC users will get a bad deal.

Intel's Uncertain Future

Alder Lake has a history of one year now. Intel is following Raptor Lake: optimization. This is a bit disappointing, but Intel will not go back to the past, because the 13th generation CPU provides more cores than the 12th generation CPU at the same price, which is similar to the situation of the 8th generation CPU. Raptor Lake is not super exciting, and its speed may not be enough to regain the leading position from AMD's Sharp Dragon 7000 series, but everyone agrees that it is a good deal to provide more cores at the same price.

But further, Intel's future is uncertain. The company has obviously made good progress. Its 7nm process (officially named Intel 4) will appear in Meteor Lake for the first time, but I have expressed some concern about Intel's strategy. Such a complex design integrates at least four different processes. I am very uncomfortable with the number of failure points of Meteor Lake. I hope Intel can use this design concept to execute its future CPUs well, because it can't afford any more delays.

However, even if Alder Lake succeeds, it is hard to see Intel return to its historical dominance. Earlier this year, AMD's market value exceeded that of Intel, which means that AMD is no longer a weak player, but a mature competitor. In this new era of competition between Intel and AMD, we will have to see what happens when the two companies compete on an equal footing. Intel is still downsizing and ceding its market share to AMD, but hopes that it can maintain equality rather than further disintegration. Theoretically, balance of power may be the best result for all.