Daguo Technology started from the Shanzhai system

Chapter 356 Black Technology, Dawn X86-S Instruction Set

After all, on the surface, the processor frequency that has the greatest impact on the performance of the CPU processor is not much different from both sides, and the processor frequency of both parties is in the 3Ghz range.

As for overclocking, Sugon Technology seems to be slightly better, and can automatically overclock to a maximum of 3.8Ghz as needed.

But the problem is that Pentium 4 also has a high-end CPU processor called Pentium 4 670, and the frequency of that processor reaches 3.8Ghz!

In this case, the Sugon Ryzen processor, which needs to be overclocked to reach 3.8Ghz, does not seem to be dominant.

Of course, which one is stronger and which one is weaker depends not only on the frequency, but other parameters are also very important.

Among them, the sincerity of Sugon Technology's Ryzen processor in the first-level cache, second-level cache, and third-level cache is also quite satisfactory.

Because the CPU processors of Intel and AMD currently on the market take cost factors into account, they are quite stingy in the first-level cache, second-level cache, and third-level cache.

For example, the first-level cache of the Pentium 4 processor usually has only one 32 KB, while the Sugon Ryzen processor directly provides 2×64KB.

So the actual L4 cache size is 4 times that of the Pentium [-] processor!
Because 2×64 means that there are two 64KB L[-] caches, considering that the Ryzen processor is a dual-core processor.

This also means that each processor core has a 64KB L[-] cache, which can be said to be quite sincere.

Of course, in addition to this, the second-level cache is also full of sincerity.

The current second-level cache of the Pentium 4 processor is usually only 1MB or 2MB, and the Sugon Ryzen processor has a maximum of 6MB.

It can be said that the size of the secondary cache is more than three times that of the Pentium 4 processor's secondary cache.

As for the subsequent three-level cache, needless to say, the sincerity of the Sugon Ryzen processor is also quite full, which is generally more than three times that of other processors!
As for the use of the first-level cache, second-level cache, and third-level cache after they are enlarged, the larger the processor cache, the better it is for software, games, and other aspects.

Among them, especially for video rendering or 3D graphics, the super large cache can play a role in turning decay into magic.

In the previous life, AMD faced the menacing 12th-generation Core processors and 13th-generation Core processors.

In order to maintain their market position, they directly produced a version of the Ryzen 100X5800D processor with a large cache of up to 3MB. The final result is naturally a miracle.

相比于普通的锐龙5800X处理器，有着100MB超大缓存版本的锐龙5800X3D处理器性能直接原地提升15%！

This 15% performance improvement can be said to have reached the level of a new generation of toothpaste, which shows how important the large cache is to the improvement of the CPU processor.

Therefore, the large cache greatly improves the processor, which also shows the sincerity of Sugon Ryzen processor!
Of course, people at the scene paid more attention to the main frequency size and the number of cores, which significantly improve game performance, so they directly skipped the cache.

And when they saw other things, they also showed stunned expressions.

For example, the synchronous multi-threading technology, the Sugon virtualization technology, the Sugon X86-S instruction set, etc. all made them look confused.

It can be said that there are many professional terms in this Dawning Ryzen processor, and people look confused. Just by a name, they don't know how it actually works in the current market.

In this regard, Lin Chen naturally would not let people continue to be stunned. Lin Chen slowly said while people were stunned:

"First of all, the first thing to introduce is the Sugon X86-S instruction set, which is an efficient instruction set developed by us based on the X86-86 bit instruction set.

The main change is to delete the support for 16-bit and 32-bit modes, and greatly simplify the X86 instruction set.

Reduce the complexity and overhead of instruction execution, optimize the hardware mechanism, and discard many historical burdens...

This Sugon X86-S instruction set can greatly improve the performance and power consumption of the processor on the basis of maintaining X86 compatibility! "

Hearing Lin Chen's words, the people at the scene continued to be stunned.

However, if there are knowledgeable travelers from 2023 here, they can't help but make a sound of horror.

Because the Shuguang X86-S instruction set developed by Lin Chen is clearly the X2023-S instruction set that Intel was developing in 86 in the previous life!

Why did Intel develop an instruction set called X86-S in the previous life?

The reason is actually that the ARM processor's aggressive entry into the computer industry has put tremendous pressure on Intel, forcing Intel to reform!
If Intel does not reform and reverse the current situation where the energy efficiency ratio of the same process technology and the same number of transistors is far inferior to that of the AMR platform processor, then the X86 computer platform will definitely be eliminated by the AMR computer platform in the future!

In this case, even with decades of accumulated software and hardware ecosystem as a solid background, what can be done if the processor performance is not as good as others.

Because in the future, with more and more games and software on the AMR computer platform, the market share of the X86 computer platform will inevitably shrink. In the end, the one waiting for the Intel X86 processor will be its demise!
In this case, a voice to improve the bloated X86 instruction set and improve the comprehensive performance of the X86 instruction set naturally emerged.

Of course, taking into account the lesson that Intel confidently launched the new Itanium processor many years ago, the result was a disastrous sales failure due to incompatibility with the X86 instruction set.

In 2023, Intel Corporation finally launched the X86-S instruction set development plan after lengthy internal discussions.

The X86-S instruction set will directly abandon the historical burden of the past, reduce the complexity and overhead of instruction execution, and optimize the hardware mechanism.

It also abandons the 16-bit and 32-bit modes, and directly develops a more efficient X64-S instruction set based on the 86-bit mode.

Although the disadvantage after this is that the 32-bit system can no longer be installed, it does not affect the use at all.

Because which system on the market is not a 64-bit system now, the market share of the 32-bit system is not high, and abandoning the 32-bit mode does not affect anything.

As for 32-bit software and games, there is no need to worry at all, because there are many ways to achieve support for 32-bit software and games.

Therefore, the birth of the X86-S instruction set is actually Intel's Jedi counterattack, and it is the existence of their high hopes.

Under such circumstances, Lin Chen actually plagiarized the achievements of Intel in his previous life in 2023. Presumably, Intel in his previous life would be so angry if he knew this.

However, Intel in the previous life was so angry that he didn't know, but when the current Intel president Paul Odley heard that the X86-S instruction set was a pure 64-bit X86 instruction set, he felt it in his heart. Not good.

Paul Oudening does not understand technology. If he is asked to develop a CPU, he will definitely look dumbfounded, but this does not mean that Paul Oudening does not know anything about CPU processors.

PS: Welcome to the old book "Empire of Science and Technology Begins with Copycat System". The two books have different styles, different plots, and different emphases. Welcome to read.

　Welcome to the old book "Technology Empire Starts from Copycat System".

　　
　
(End of this chapter)