"Fang Zhou -1" is worth cheering?
Silicon Valley
01-8-16 04:37:14
-------------------------------------------------- ------------------------------
Editor's note: On July 11th, this site used the "History of China's Information Industry" and celebrated the birth of my country's first 32-bit CPU chip "Fang Zhou -1". Today, we will issue an article again. Thinking from Silicon Valley Chinese engineers, I hope to perspective from another side and angle, perspective of the industrial connotation of Ark 1. From the words of this article, we don't even matter the author's hopes sent to the Chinese chip industry, of course, there is still a fierce words. However, we have more hope that more rational thinking will be taken out, and the wisdom of the group will provide advice to the development of China's information industry. Looking forward to your voice
When the first CPU "Fangzhou-1" is launched, we have seen the familiar propaganda model again. Due to the Chinese electronics industry for a long time for this CPU, from this unfamiliar propaganda model, we can clearly see the subsequent prominent expectations and heavy wings.
However, what did the core do?
According to reports, "C32" Fangzhou -1 ", 32-bit CPU chip uses a self-designed architecture, is a CPU chip for network applications. It has the characteristics of fast, low cost, high integration, and frequency To reach 166 MHz, 200 million instructions can be implemented per second, compared with foreign similar chips, "Fang Zhou -1" technical performance is quite, in the field of embedded CPUs reached the international advanced level. "
How did this "International Advanced Level"? The report also said: "Zhu Gaofeng, the president of the Chinese Academy of Engineering, is divided into Zheng, Deputy Director, 20-bit CPU Technical Appraisal Committee consistently consistently believed ... ... "The problem is that this" International Advanced Level "evaluation and the review of the review of high-tech fields are quite left. It is a bit thin, a little "Wang Po sells melon" taste (not involving individuals, please forgive). The "Fangzhou" chip belongs to the embedded CPU, and the performance assessment of the embedded CPU is a complete set of procedures for the universal test and evaluation of the industry (called Benchmark). If "Square-1" really wants to demonstrate that he belongs to "International Advanced Level", the most standard practice is to invite independent professional chip evaluation companies to run full assessment procedures. Use specific operation results and data, rather than doing "political and dangerous" propaganda.
However, "SMIC" did not do this, "SMIN" is more like completed a political task. If a company does not dare to take industry standards, there will be enough reasons to have a suspicion of its technical content. And since the "Fangzhou-1" has reached the "International Advanced Level", we may wish to see what is the international advanced level of embedded CPU.
Embeded MicroProcessor Forum in 2001 just has just driven. One of the most important topics of this year's Forum is the embedded CPU for network applications, and this is just a design goal mentioned in the report, and we will compare.
According to Eetimes Report [1], PMC Serra launched its online 1-GHzRM9000X2 Dual Processor, which used the annual forum. The chip uses a 0.13 micron process, and two two MIPS microprocessors can run in 1 GHz.
PMC Serra's chip is a design for fiber optic network OC192 10 GBITS / S data flow. SMIN's "Boat-1" also claims to be a network-oriented design, but in the frequency, the chip of PMC Serra is already six times that of "Fangzhou-1". The structure of its dual CPU makes the processing capabilities of the "Fang Boat" chip far difficult to look back. "Square" chip is used to make a low-end network terminal, and it is not intended to do high-end network data processing. This is "SMIN" so-called "design for network application"? This is a design that can be called "International Advanced Level"? Of course, PMC Serra is an old chip company with a reputation in the communication chip. Well, we don't have a PMC ratio, we and newly born companies "Start-UP" ratio. LEXRA, established four years, also introduced the OC192 design in this year's forum [2]. Lexra's chip can run 420 MHz, 2.5 times the "Fangzhou-1". What is more commendable is that LexRa successfully integrated 16 RISC CPUs on a chip, resulting in powerful computing power. In terms of computing ability, a Lexra's RISC CPU is already 2.5 times the "Fang Zhou -1", 2.5 is multiplied by 16. What is it?
If it is said that "SMIN" is a newly created company should put the standard, then why can the new founding company LexRa can take out the top design of the world? Dr. Li, who said that SMIN has 20 years of design experience, why the design of the shot and LEXRA is much better than that. How many years have you been in the general designer of LEXRA? 40 years? 50 years? There is no CPU in 40 years ago.
If it is because the domestic condition is poor, the foundation is weak, that is not. Domestic chip design companies use software and American chip companies, and foreign high-tech Start-Ups are all in the white hand, and they will not talk about what the foundation.
A possible explanation is that the entrepreneurs in SME are too conservative and locate themselves too low. It is also a strategy to position your own products at low ends in situations in the market and technology. But if this strategy does not get a timely adjustment, it will immediately make the company fall into a non-technical hand workshop. Dr. Li, who is "SMIN", said in an interview, "Fangzhou-1" is mainly used for network computers and set-top boxes. What kind of network computer, the set top box is a very bad computer, which is not too much than a "Dummy" terminal with a pure display. If the "SMIN" chip is ready to do this, I still don't claim to be "high-tech company."
Chinese chips should be inherited to catch up with acceleration, need to be sharp and rushing, rather than holding disabled and rising. It is necessary to have a deep understanding of the technology and the real part of the design, not a "political props" without technical content.
When Dr. Li's next time a copy is designed, cheering may be more valuable.
This article only represents the author's personal point of view!
This article is not allowed to reprint without permission! !
