Chip War 2.0: The Global Battle for Semiconductor Supremacy
芯片战争2.0：全球半导体霸权之战
In this solo lecture, Chris Miller, author of the acclaimed book "Chip War", delves into the intricate world of semiconductors, exploring how microchips have become the cornerstone of global power and economic influence. Miller will provide an analysis of the geopolitical ramifications of semiconductor production, focusing on how the fierce competition between global powers for dominance in this critical industry will further develop.
广受好评的《芯片战争》一书的作者克里斯·米勒深入研究了半导体的复杂世界，探讨了微芯片如何成为全球权力和经济影响力的基石。米勒对半导体生产的地缘政治影响进行分析，重点关注全球大国之间争夺这一关键行业主导地位的激烈竞争，会如何进一步发展。
Moreover, he will also spotlight the strategic role of South Korea and East Asia in the semiconductor landscape. Miller will examine how South Korea's advanced semiconductor capabilities and the region's pivotal position in global supply chains influence the broader geopolitical tensions between major players like the United States and China.
此外，他重点介绍韩国和东亚在半导体领域的战略作用。米勒研究韩国先进的半导体能力和该地区在全球供应链中的关键地位，如何影响美国和中国等主要参与者之间更广泛的地缘政治紧张局势。
以下是外国网友的评论：
godzillamothra5983
This semiconductor race is a marathon, and Confucius said a journey of thousand miles begin with a single step.
这场半导体比赛是一场马拉松，孔子说千里之行，始于足下。
haochengzhai7156
那是老子说的。
stachowi
is it though? not sure we can make things any smaller... i think we need nuclear-powered massive data centers....
不确定我们能不能把东西做得更小…我认为我们需要核动力的大型数据中心....
manmohanmehta5697
What a lecture with brilliance oozing out. The issue that singles out is the supremacy over chip manufacturing. Huge supply chain issues and world, is now divided into camps facing each other for supremacy in AI . This is obviously an issue of greater concern. Who knows the consequences of this geopolitical issue.
突出的问题是芯片制造的霸权。庞大的供应链，现在分裂成两个阵营，在人工智能领域争夺霸权。这显然是一个更值得关注的问题。谁知道这个地缘政治问题会有什么后果？
mariajiao4855
I agree. This guy is a opportunist, he basically said what the hawks like to hear. Not much insights.
这家伙是个机会主义者，他基本上说的都是鹰派爱听的话。没有太多的见解。
mariajiao4855
I don't understand why this is called Chip War 2.0. It's just some collection of news topics, not much insights.
我不明白为什么这被称为芯片战争2.0，只是一些新闻话题的集合，没有太多的见解。
闲云野鹤-g5w
I think 1.0 is with Japan
我认为1.0是此前和日本打的
NeoFromMatrix123-b1v
I really enjoyed his book 'Chip War' because of the way it looks at the topic with a historical perspective (i.e. chronologically).
我真的很喜欢他的书《芯片战争》，因为它从历史的角度（即按时间顺序）看待这个话题。
ravichanana3148
Very good lecture! After Japan and South Korea, China is next in to manufacturing chips. India has lost to China. Some design is there at the Semiconductor Laboratory at Mohali. India can still complement with stress on education and research.
继日本和韩国之后，中国是下一个制造芯片的国家。印度输给了中国。在莫哈里的半导体实验室有设计一些产品，印度仍然可以通过加强教育和研究来追赶。
deltrinos
The guy is just a walking news feed and press review, with no actual interest, and not even a drop of intelligence coming from his own mind. The US has triggered the collapse of the semiconductor industry by pushing China toward self-sufficiency. Already, we can feel the pressure building in the mature semiconductor sector, and soon, the pack will catch up to the leaders. For the optimists, China is now only 18 to 36 months behind the leaders. In one or two years, we’ll see a complete reconfiguration of the industry, and in the end, the US has simply sabotaged the very field it was still capable of dominating.
美国推动中国实现自给自足，引发了半导体行业的崩溃。
我们已经可以感受到成熟半导体行业的压力，很快，这群人就会赶上领先者。
对于乐观主义者来说，中国现在只落后于这些国家18到36个月。
在一两年内，我们将看到该行业的彻底重组，
最终，美国毁了自己仍然有能力统治的领域。
linwang3896
A hollow talk!
空谈！
ZahidHussain-zy8ds
Thank you for highlighting an important but so far undiscussed aspect of semiconductor research.
It is not important who wins the race or fails to catch up. It is the use of the technology behind the product, evil or good with its possitive or negative outcome, which matters.
Anything else is not imporrant at all.
感谢您强调了半导体研究中一个重要但尚未讨论的方面。
谁赢了比赛或没能赶上并不重要。重要的是使用产品背后的技术，是善是恶，其结果是积极的还是消极的，其他的都不重要了。
BobQuigley
Define good Define evil. Provide examples. Is it acceptable to use it for good in your own nation and have evil purposes against your enemies dujour?
如何定义善恶，提供例子。你对内为善，在仇敌身上为恶，这也可以接受吗？
xngpan8130
He is not Chinese, why is he talking from the Chinese's point of view. He should just talk from his point of view.
他不是中国人，为什么要站在中国人的角度说话。他应该站在自己的角度说话。
NorCalMoDo
Celebrate while you can.
趁还能庆祝的时候好好庆祝吧。
cnhugovideo
The advancement in China chip industry the last two years has totally trashed his book Chip War.
过去两年中国芯片产业的发展啪啪打脸他的《芯片战争》一书。
genuinennessbefitting4734
What advancement in China have you noticed? We see no advancement in China from the viewpoint of a tsmc engineer
你注意到中国的哪些进步？从台积电工程师的角度来看，我们没看到中国的进步
SalesManual
hehehehe....
china....impossible....
呵呵，中国……不可能……
JigilJigil
You mean the advancements in their dreams.
你是说他们在梦中的进步。
kalipotmeng
Chris Miller is a russologist, specialist in cold war. Somehow jumped on the train of chip technology. I doubt he has ever programmed something useful in his life. He's obviously wrong by saying the future of AI resides in finding more data to train on and that people are finding more and more data sources. The current largest LLM practically exhausted all the data available. New developmental leaps reside in finding different architecture.
克里斯·米勒是俄罗斯学家，冷战专家，不知怎么跳上了芯片技术的列车。我怀疑他一生中没编写过什么有用的程序。
他说人工智能的未来在于寻找更多的数据来进行训练，而人们正在寻找越来越多的数据源，这显然是错误的。
目前最大的LLM几乎耗尽了所有可用的数据。新的发展飞跃在于寻找不同的架构。
urimtefiki226
Wisdom first, not money.
As long as you think only about money, you are in the level of subhuman.
首先是智慧，而不是金钱。
只要你只想到钱，你就处于次等人的水平。
NorCalMoDo
Microsoft and meta think money
微软和meta考虑的是钱
qingzhou9983
This guy is a Chinese expert. But he clearly does not understand China looks at an issue with decades horizon, like the International Space Station, missiles, advanced fighter etc.
Decades!
China just announced the 28 nm DUV machine after 6 years. Most of chips needed are 14 nm and above!
这家伙是中国专家？中国是用几十年的眼光看待问题的，比如国际空间站、导弹、先进战斗机等。6年后，中国刚刚宣布推出28纳米DUV机。大多数需要的芯片是14纳米及以上！
JigilJigil
During the last 4 decades China has spent over 100 billions of dollars to make its own domestic airliner, and it couldn't, after miserable failures, they asked US to make them one, Comac, an airliner, that all its ciritcal parts, and7 0% of its components come from US and Europe. China made its space station, with decades of free accessible data and design from NASA and ESA, etc... when it comes to space stations literally any country that can afford it can build it, the case of China's station is even worse, built on mostly the technologies that was transfered from Russia, and based on the outdated Soviet technologies, and it's not it, the best part is that they couldn't even make the docking system by themselves, again Russia tech transfer, basically they asked russian to made them one, also they couldn't make the crew vehicle, it's essentially the outdated Russian Soyuz vehicle with Chinese letters on it.
在过去的40年里，中国花了1000多亿美元造国产客机，但在经历了悲惨的失败后，他们要求美国为他们制造一架中国商飞，一架所有关键部件和70%的部件都来自美国和欧洲的客机。
中国建造了自己的空间站，有几十年的免费数据和来自NASA和ESA等的设计…说到空间站，任何有能力的国家都可以建造，中国空间站的情况更糟，主要是建立在从俄罗斯转移的技术上，基于过时的苏联技术，他们甚至不能自己制造对接系统，基本上是他们要求俄罗斯人造了一个。他们也不会造载人飞船，本质上是过时的俄罗斯联盟号飞船，上面印上中文。
qingzhou9983
You forget 40 years ago, PRC was the poorest country on earth, worse than worse countries in Africa.
So you just need to wait another 40 years.
Be Patient!
40年前，中国是地球上最贫穷的国家，比非洲国家还糟糕。
所以你只需要再等40年。要有耐心!
AZ-hj8ym
You don't need cutting edge chips to train AI. You can also do it with chips with the same computational power. China can do it without Nvidia.
不需要尖端的芯片来训练人工智能。也可以用同样计算能力的芯片来做，没有英伟达，中国也能做到。
JJGlyph
That would insanely bloat the energy demands and is not feasible because they have not figured out how to train AI across multiple data centers like Microsoft has. It's a scaling issue from both a computational and energy perspective.
此举会使能源需求疯狂膨胀，而且是不可行的，因为他们还没有像微软那样弄清楚如何跨多个数据中心训练人工智能。从计算和能量的角度来看，这是一个缩放问题。
qingzhou9983
There is no doubt they will figure that out like Microsoft.
For energy, PRC has the most Green Energy and is adding it at a speed greater than anyone else many times.
So far, China will be mostly run on Green Energy with Beth low cost.
毫无疑问，他们会像微软一样解决这个问题。
在能源方面，中国拥有最多的绿色能源，并且正在以比其他国家快很多倍的速度增加。到目前为止，中国将主要使用低成本的绿色能源。
qingzhou9983
You just forget China is A develo country, not developed countries. And China is also the Factory for the World. So huge part of its carbon emissions is for people in other countries.
你忘了中国是发展中国家，不是发达国家。中国也是世界工厂。中国的碳排放有很大一部分是为其他国家的人排放的。
longdog33
This talk is total BS. US restrictions on chips to China is not about national security and AI per se. It’s about the economy. If it was about national security and AI, then why not let Huawei have access to advanced chips for its phones? Huawei was the first target of Americans restrictions and still is. Chris miller is an apologist for American empire.
完全是胡扯。美国对中国芯片的限制与国家安全和人工智能本身无关，与经济有关。如果是为了国家安全和人工智能，那为什么不让华为获得先进的手机芯片呢？克里斯·米勒是美帝的辩护者。
tedcatzynski5444
He is lying. China imports fewer chips is 2024 than it did in 2014.
In 2014 it imported $1.7b worth of chips per month.
In 2024 it imported $1.5b worth of chips per month.
Imports peaked in late 2021.
他在说谎。
2024年，中国进口的芯片将少于2014年。
2014年，中国每月进口价值17亿美元的芯片。
2024年，中国每月进口价值15亿美元的芯片。
进口在2021年末达到峰值。