area51novel #511 - Unable to jump in lithography machine iteration
#511 - Unable to jump in lithography machine iteration
For the people of the *Bright Sword* world, the shocks from Dongda in recent years have been too numerous, leaving them somewhat overwhelmed and even a little aesthetically fatigued.
The advent of the aircraft carrier, in reality, only caused a stir for a day before returning to calm.
After all, compared to sending people into space, the impact of the aircraft carrier's arrival on ordinary people was far less significant.
People even deliberately ignored it.
Because whether or not it existed, everyone was already clear that Dongda was not to be trifled with.
So, regarding Dongda's first aircraft carrier, they were only mildly surprised before moving on.
They couldn't even feel the passion of the Dongda people for this aircraft carrier.
Of course, Dongda's aircraft carrier was no longer a secret to the *Bright Sword* world, as Dongda had never intended to develop it in secrecy from the start, even pulling in John Bull to contribute some of his technology.
Therefore, the progress and such were actually well-known to all major powers in the world.
However, after its launch, to the bigwigs of the *Bright Sword* world, Dongda's progress seemed a bit fast. Completing an aircraft carrier in just over three years was almost comparable to the rogue Eagle during World War II!
Keep in mind that this wasn't a simple escort carrier converted from a merchant ship; it was a true war machine designed from scratch!
However, Ren Zhong didn't care about these things. With the final piece of the military puzzle in place, this new aircraft carrier designed from the ground up essentially utilized almost all of the aircraft carrier data that Ren Zhong had painstakingly transported from the main world.
The team now had a practical process, completely familiarizing themselves with a brand-new, modern aircraft carrier.
It could be said that apart from the insurmountable era differences in carrier-based aircraft and avionics, in terms of power, the number of carrier-based aircraft, deck design, catapults, and arresting gear, the current Yandu aircraft carrier was not inferior to any modern aircraft carrier.
This essentially gave Dongda a head start of more than half a century in entering the aircraft carrier era.
With this start, the future Dongda navy would have no bottlenecks. From large submarines, large destroyers, amphibious assault ships, to aircraft carriers, everything would be available, and in terms of quality, it had already begun to completely surpass any current opponent.
Therefore, after careful consideration and clearing up several technological development paths in his hands, Ren Zhong resolutely refocused his strategic vision on the crucial field of the electronics industry.
Although current lithography technology has made breakthrough progress, entering a new 2-micron era and birthing new production equipment for the 286 chip, in Ren Zhong's profound vision, having read the history of technological development in the main world, this is only the budding stage of the electronics industry's vigorous development.
The 286 computer is but a fleeting moment in the long history of computer development. It marks the personal computer's journey from initial budding to preliminary application, but it is far from reaching the peak of technology. The lithography machines currently developed by Dongda, even the latest models, are, from Ren Zhong's perspective in the main world, merely prehistoric relics. Those ancient devices are almost impossible to find in the main world, only occasionally glimpsed in dusty historical data.
The development of lithography machines that truly led the electronics industry into the modern era was far more complex and glorious than imagined.
The birth of the first generation of modern GLINE lithography machines truly marked the first leap in the history of semiconductor manufacturing technology. They used a g-Line light source with a wavelength of 436nm, a technological innovation that made it possible to produce chips with 0.8 to 0.35-micron process technology. The corresponding equipment was the first generation of modern contact and proximity lithography machines, opening up new horizons for the manufacturing of 486 and later new CPUs.
In the early CPU manufacturing process, the 1-micron process could only support the production of 386-level CPUs at most, and the maturity of the 0.8-micron process marked the arrival of the 486 era. To further leap to the 586 Pentium-level CPU, the process had to evolve to 0.35 microns, which was undoubtedly a huge challenge in the field of semiconductor manufacturing.
However, in terms of the current development of Dongda's CPU lithography technology in the *Bright Sword* world, they are still far from touching the threshold of modern lithography machines.
The 2-micron process, in Ren Zhong's eyes, is nothing more than a relic of the prehistoric era. In the future, there are still four generations of different light source lithography technologies waiting to be conquered, each step full of unknowns and challenges.
The second-generation lithography machine uses i-Line as the light source, with the wavelength shortened to 365nm, and technological advances have made it possible to produce chips with 0.8 to 0.25-micron process technology. This process level, in the main world, corresponds to the glorious era of the Pentium III CPU. The Pentium III, as a classic product of Intel Corporation, not only achieved significant improvements in performance but also set a new benchmark in semiconductor manufacturing processes.
Subsequently, the third-generation lithography machine adopted a KrF light source, with the wavelength further shortened to 248nm, and the process node was upgraded to the 180 to 130nm level. This technological innovation provided strong support for the production of the first and second generations of Pentium 4. The 180nm process technology first-generation Pentium 4 Willamette, and the subsequent second-generation Pentium 4 processor Northwood, which used 130nm process technology a year later, were both products of this technological advancement. They not only improved the performance of the CPU but also promoted the development of the entire semiconductor manufacturing industry.
The fourth-generation lithography machine is an extremely important milestone in the development of lithography technology. The introduction of the ArF (DUV) light source shortened the wavelength to 193nm, and through technological innovation, the actual wavelength utilization rate was increased to 134nm. This technology is the famous immersion lithography technology, which further improved the ArF lithography level: by filling water between the projection lens and the wafer, due to the refractive index of water being close to that of glass (at a wavelength of 193nm, the refractive index of air = 1, water = 1.44, and glass is about 1.5), the light emitted from the projection lens enters the water medium, the angle of refraction is small, and the light can be normally refracted from the objective lens. The actual equivalent wavelength of the ArF light source plus immersion technology is 193nm / 1.44 = 134nm.
This technological breakthrough full of ingenious ideas enabled the widespread modernization process after 130nm, and the most advanced process can even be improved to the 7nm level (of course, under such extremely limited processes, the yield rate is far less than that of EUV lithography machines). This generation of lithography machines is the most widely used and representative generation in the main world. From the third generation of Pentium 4 onwards, most CPU, GPU, and memory particle chips are produced by this generation of lithography machines. In Ren Zhong's main world, this is also the strongest process technology that Dongda can master.
However, the challenges didn't stop there. The fifth-generation lithography machine, using EUV as its light source, shortens the wavelength to 13.5 nm and employs extreme ultraviolet light technology. This technological innovation allows process nodes to reach the 14nm to 3nm level, making it one of the most advanced products on the market today. In the main world, it is hailed as the strongest lithography machine, widely used in the production of the latest CPUs and GPUs. The emergence of this technology has not only driven another leap in the semiconductor manufacturing industry but also laid a solid foundation for future technological development.
Faced with such a daunting task, Ren Zhong knew that breaking through step by step would likely lead to detours for Dongda in the Bright Sword world.
Although the g-Line light source is more in line with the technological background of the Bright Sword world, the i-Line light source feels somewhat like redundant research. Therefore, in the subsequent evolution of lithography machines, Ren Zhong decided to directly challenge the second-generation lithography machine with the i-line light source from a technical perspective. This technical route can ultimately impact the 0.25-micron limit, producing modern CPUs like the Pentium III. At this level, it can basically meet the multimedia performance requirements that Ren Zhong envisions for the network era.
However, this would directly increase the technical difficulty.
Modern chip manufacturing processes, not to mention wafer production, include multiple complex processes such as initial oxidation, photoresist coating, exposure, development, etching, and ion implantation.
These process flows require a wide variety of equipment, including oxidation furnaces, photoresist coating and developing machines, lithography machines, thin film deposition equipment, etching machines, ion implanters, polishing equipment, cleaning equipment, and testing equipment. In fact, even the final packaging and testing is not simple.
This is one of the reasons why Ren Zhong cannot directly jump to DUV, because at this stage, leaping too far is simply impossible!
Even if Ren Zhong can get the principle data of these devices, without bringing prototype products to the Bright Sword world, it would be impossible to research such complex things in a short period. You need to know that a DUV contains more than 100,000 parts! This is far more complicated than building an atomic bomb!
This is just the hardware aspect; the software system aspect also requires time to accumulate and cultivate true top-notch computer scientists!
Ren Zhong can bring top-notch minicomputers to the Bright Sword world to create special R\u0026D zones like the Fifth and Ninth Districts of Dongda, deploying a truly modern development environment that is comparable to the main world. However, the cultivation of experts cannot be achieved overnight.
Currently, the Fifth and Ninth Districts, in addition to normal R\u0026D for research and verification of new products,
Ren Zhong has also recruited a large group of geniuses in mathematics and physics, who are undergoing in-depth training as computer system experts. Ren Zhong used top computer and industrial control textbooks from the main world, re-edited them into PDF documents, eliminated as many traces of the main world as possible, and provided real-time operating systems such as μClinux, μC/OS-II, eCos, FreeRTOS, mbed OS, RTX, Vxworks, QNX, NuttX, large distributed systems like Hongmeng Open Source System, as well as DOS, Linux open source server and desktop systems, and many other system source codes for these seed players to learn and verify.
These potential computer seeds are frantically throwing themselves into the process of learning, digesting, and introducing new knowledge.
The latest novels are first published on 69shu.com!
This R\u0026D team, composed of top scientists and engineers, is selected not only from schools but also from various industries, picking out talented people with comprehension. Therefore, the team members come from different fields, and they will eventually apply what they have learned to their own industries.
It's just that the chip capabilities of the Bright Sword world are not strong enough now. The boards and chips for testing these systems and ideas all have to be brought in by Ren Zhong from the main world. Fortunately, the weight of the chips is not large. After completing the circuit board design content of the main world in the Bright Sword world, Ren Zhong brings in mainly chips, instead of having to bring in complete board cards like at the beginning.
It is the supply from the main world that allows the experts in the Fifth and Ninth Districts to make rapid progress in computer technology, and the chip design progress even surpasses Moore's Law in the main world!
A new generation of chips is iterated out in less than a year, because the current chip design programs in the Bright Sword world are basically the same as in the main world. Ren Zhong spent a lot of money to buy a complete set of software for chip design from the main world!
For design software that can support the design of billions of integrated circuit chips, facing the small scene of hundreds of thousands of circuits like the 286/386 is even easier.
Not only is it rapidly evolving along the X86 technical route, but also in another technical route direction, the ARM and RISC-V open source architectures are also constantly evolving, because these two can directly buy the source design data and circuit diagrams. Compared to the X86 route, Ren Zhong can go faster on this technical route!
However, the ARM and RISC-V design data currently purchased are still too high-end for the processing technology of Bright Sword design!
Basically, even entry-level industrial control chips on the market, even low-end chips like the stm32 Cortex-M3, require a manufacturing process that reaches the 130nm level.
Therefore, Ren Zhong faces too many tasks in the chip field.
Moreover, after cross-generational research, most of the technologies and equipment involved are beyond the scope of the Bright Sword world, and there is no way for them to solve it themselves.
Not only is there the lithography chip process, but wafer production is also a new challenge.
Wafer manufacturing involves multiple steps, including wafer growth, cutting, cleaning, etching, chemical mechanical polishing, and photolithography. Each step requires strict control of conditions, and there is a strong demand for automated control. In the early 8080 chip manufacturing, Dongda researched the first-generation three-inch wafers, but in the further 8086 chips, it evolved to 4-inch wafers.
To further produce 386 and 486-level chips, basically 6-inch wafers are required, and 8-inch wafers are required for Pentium-level chips to achieve a good yield.
It's just that the technological progress of chip manufacturing basic materials must keep up, otherwise, even if there is a very good lithography machine, it will be useless.
However, now, whether it is lithography machines or the preparation research of chip basic material wafers, the R\u0026D costs required for each generation of iteration and evolution are increasing in orders of magnitude.
Ren Zhong needs to promote the development of this entire industry technology, and there are many interconnected and mutually restrictive conditions that need to be overcome in the middle.
For example, DUV, such a precise control device, requires a more powerful industrial control system, and this industrial control system requires the support of chips with stronger processing performance. The current 8086 chips are weaker than the STM32 F3 in the main world, so they cannot complete this industrial control system.
To solve this problem, we must first think of producing early lithography machines, such as 1-micron process lithography machines, to create chips that are stronger than 8086, such as 386 or 486 level, and then use the industrial control system produced by such chips to match the production of stronger lithography machines, such as 0.35-micron lithography machines!
This will have the ability to produce Pentium-level chips, and use the new Pentium chips to develop new equipment that can control the 0.13-micron process. After repeating this cycle several times, it will be possible to produce embedded systems similar to stm32 Cortex-M3 chips, so that the industrial control system can gradually evolve to the current level of the main world.
This is one link after another, and it cannot be achieved overnight. It will take at least ten or twenty years.
It can be said that it is too difficult to jump-level development.