"Today, we officially announce to everyone that the lithography machine bottleneck that has plagued the industry for over ten years has been officially broken by us, AmsL."
"As I've said before, since the mid-1990s, the development of lithography machines has essentially stagnated. Due to limitations in light source technology, humanity has been unable to produce lithography machines with a manufacturing process below 70 nanometers, thus unable to produce corresponding chips."
"The reason chips possess terrifying computing power is actually by first wrapping a layer of insulating material around a single silicon crystal, and then using extremely fine light beams to burn away the insulating layer underneath. It's like turning light into a pen, directly drawing semiconductors one by one on the single silicon crystal, ultimately forming a computer chip with terrifying computing power."
"In the beginning, the lines humans drew were drawn by hand, with precision in the millimeter range, so a single silicon chip could only have a few semiconductors. Later, it evolved to using precision-machined metal molds for pressing, but even then, the line width was only at the 0.1-millimeter level, at most capable of drawing a few hundred to a thousand semiconductors."
"It wasn't until someone discovered that light could be used to etch onto single silicon crystals that the first lithography machine was born. However, it could only etch lines around 0.01 millimeters, or about 10 micrometers, meaning only tens of thousands of semiconductors on a piece of single silicon."
"But with continuous breakthroughs by humans, constantly replacing light sources and lenses, the light beams became finer and finer, eventually leading to the current 70nm-level process lithography machines, capable of etching tens of millions of semiconductors on a single silicon crystal."
"But that was the absolute limit. By the mid-1990s, humans realized that whether it was lenses or light sources, humanity had already reached its peak. Unless a finer light source than lasers and ultraviolet light could be found, or light was replaced entirely with another method of etching, this would remain an insurmountable dead end."
"Yet, in over ten years, no company has been able to achieve this. The limitations in lithography machine development have consequently restricted chip manufacturing precision. Therefore, over the past decade, the hardware technology of computer chips has seen very little development."
"However, corresponding software technology, particularly chip architecture, has developed rapidly. This at least prevented computer chips from reaching a state of complete iteration. But it ultimately couldn't change the hardware limitations. Starting around 2003, the development of computer chips has entered a state of extremely slow progress."
"Take Intel, for example. Although they release new high-end computer chips every year, the performance improvement of the new models each year is only around 10%, a far cry from the previous annual performance increases of 100%."
"Intel's practice has been dubbed 'squeezing toothpaste' by users, and Intel subsequently earned the title of 'toothpaste manufacturer'."
"There's been a persistent notion that Intel actually holds about 200% of the potential computing power improvement, but they are afraid that if computer chip technology never breaks through, they might not be able to release new products with iterative computing power in the future."
"Thus, they've resorted to 'squeezing toothpaste,' ensuring that this 200% computing power can last them for decades. They believe that in twenty years, they will surely break through the bottleneck of computer chip technology!"
"As it turns out, Intel didn't need to wait twenty years. In about five or six years, AmsL, a newcomer in the lithography machine industry, delivered a huge surprise to the world by being the first to introduce lithography machines with a manufacturing process below 70 nanometers."
"Logically, the most excited people in the audience should be chip companies like Intel. However, the representatives from these companies wore very calm expressions because they had already received the news. In fact, AmsL had already produced its first prototype machine more than half a year ago, and this prototype is currently undergoing production trials at Intel's headquarters. This is no longer a surprise for Intel."
"Other chip manufacturing companies, such as TSMC, Samsung, Philips, and some companies from the US and Europe, are all shareholders of AmsL, so they naturally knew about this news long ago."
"The truly most excited person in the audience was Old Huang, the boss of Nvidia, who would later be called 'Dad' by countless miners. He was excitedly watching the latest technology announced by AmsL on stage, then he walked over to the representative from TSMC, the Vice President of TSMC, and whispered, 'Brother, you must give your little brother a hand this time!'"
"'What are you saying?' asked the Vice President, his expression changing slightly as he feigned ignorance."
"'Brother, let's not beat around the bush. Of course, I'm talking about the first batch of 50nm chips. You must give us priority. Our 8800GTX is counting on this batch of chips!' Old Huang said unceremoniously."
"'Your request is rather difficult. Don't you see what they're saying? If this year's delivered lithography machines exceed ten units, that would be great. Half of these ten machines are definitely going to Intel. The remaining five will be up for grabs by other companies. It would be a miracle if our company can get its hands on two of them!'."
"'The chips produced by these two machines will definitely be prioritized for others. So, you can forget about it!' the Vice President said unceremoniously."
"'I know. That's why I'm not asking for this year's chips. I want next year's. But you must give me a few hundred samples this year for development, so that our brand new 8800GTX can be launched around 2007!' Old Huang said quickly."
"'This... is a bit difficult!'"
"'For the sake of being countrymen, give your brother a hand!' Old Huang grabbed the Vice President's hand, and with his little finger, he drew several vigorous circles in the Vice President's palm. The Vice President's face immediately showed a look of understanding. 'Fine, I'm afraid of you. I'll try my best to secure a quota of 10,000 units for you next year!'"
"'Only 10,000 units? That's too few! They'll be sold out in less than a day!' Old Huang quickly said."
"'Let's not talk about it. Do you know how many companies worldwide are eyeing this product? I'm telling you, not just next year, but for the next three years, this product will absolutely be in short supply, and you won't be able to get your hands on it. My agreeing to give you 10,000 units next year is already showing you immense respect.'"
"'Don't worry, I understand what you mean. You can at least get 10,000 units from us, but AMD won't be able to get a single chip for at least the next three years. Intel won't give them a chance!' the Vice President said with a rather pleased expression, and Old Huang beside him also smiled slightly."
"If that's the case, then even 10,000 units would be excellent. My only competitor at the moment is AMD's graphics cards. The market competition is currently very fierce, and neither side can defeat the other."
"But if two years later, Nvidia releases high-end graphics cards using the latest generation of chips, while AMD has nothing, then Nvidia can completely suppress AMD and dominate the graphics card field."
"AMD won't be able to launch corresponding products for at least three years. But this one year is enough for me to completely defeat AMD."
"Earlier, I mentioned that some people in the audience were happy and some were worried. The expression of the AMD representative was the worst; they also realized that they were probably doomed."
"In the market, AMD has always been Intel's biggest competitor, and the two sides have at times fought back and forth. However, after the 1990s, AMD was completely at a disadvantage, even filing for bankruptcy several times."
"If it weren't for people lending them money at critical moments, AMD would have long since turned to dust."
"Of course, this is if the person lending them money wasn't Intel."
"Yes, that's right. At critical junctures, Intel lent money to AMD and helped this company survive. Such behavior seems very strange, as if Intel is a great benefactor. But in reality, Intel had no choice."
"After all, the antitrust laws at that time were very strict. Even Microsoft was nearly split in two by the antitrust law, and Intel itself was born out of the antitrust law and was terrified of it."
"The best way to eliminate the threat of the antitrust law was to ensure that the conditions for its establishment did not exist. That is to say, there had to be a competitor in the market, and one that occupied at least about 10% of the market share."
"But finding such a competitor was too difficult. Under Intel's formidable strength, all early competitors had died out. In the end, only AMD remained as an old opponent."
"Therefore, Intel had to find ways to keep AMD alive, even proactively ceding a portion of the market to AMD so that it could maintain a market share of around 10%. This way, the antitrust law would never take effect."
"Well, this approach can be described with a not-so-fitting idiom from ancient China: 'nurturing a rival for self-preservation'."
"Only, Intel never dreamed that their 'nurturing a rival for self-preservation' would almost lead to their rival's counter-attack."
"After entering the 2000s, with the stagnation of lithography machine technology, AMD was on the same level as Intel in terms of hardware. At this time, AMD introduced several architecture masters, and their new generation architecture allowed AMD's chip performance to be only slightly inferior to Intel's, but the price was very attractive, only about 2/3."
"As a result, Intel's market share once dropped to below 60%. This competitor was truly on the verge of counter-attacking."