Chinese were happy to be part of American system in term of science, tech development/applications, manufacturing, trading, finance, supply chain etc. until 2018 and especially since 2022, that the US forced China to change their lane to unfollowing the UsA. Now China is developing everything and doing everything to decouple from the US.
Starting from the Wolf Amendment in April 2011, and the stories went on, now it’s the history that goes on. One may be puzzled and asked why? Just can’t get along, simple.
@@leszekladzinski73really? The petro dollar agreement with Saudi Arabia is expiring on the 9th of June and Saudi will NOT renew it. Saudi will instead accept all currencies like the yuan. Cope
Kevin boy knows the original meaning of the term chinaman. Chinaman, as well as silkman and teaman, were the names for European aristocracies used by the commoner Europeans for thousands of years because they were the only Europeans who could afford china, silk, and tea. So, Chinaman is a term of admiration to Europeans. You must have heard of a type of furniture called a china cabinet. You may even own one and display your favorite china in it. Of course, you may not because they aren't cheap, and many Western families don't own one. Similarly, the meaning of Christianity used to be Godless because Romans had their God. The Asianman who crawled out of the virgin's pussy wasn't considered God in both Israel and Rome, of course, including Europe.
While US bans advance chip sale to China, China is also banning the sale of advance chip sale to US. When SMIC and Huawei are mass producing these photonic chips, it will be banned from selling to the west and is used exclusively for China. TSMC will also be banned from China.
The things about advance chips is that if you're not using its your companies and your country cannot compete in the global market meaning your economy will suffer.
That is not a good idea. I believe China won't do that. See the soil obtained from the moon by China's spacecraft is also shared with US scientists for research.
Photonic chips would be technolocal giant leap making the existing technology instantly obsolete. Its woukd be like a conscious server was born (like in the film “her”) or the atomic bomb only owned by the US in 1945. Panic would ensue, even worse than the panic of electric cars. It would be impossible to catch up because everything from severs to PCs to embedded war technologies would be instantly inferior in the West. The would not be whether to open or close the market but how to catch. Histotpric well known pattern of events will be like the atomic bomb.
Intel had a major opportunity to develop optoelectronic chips in the mid oughts. Indeed they (in partnership with UCSB) were the first to create a laser in silicon, which was a remarkable feat. But management being management decided to take the lazy, easy path and they shut the whole thing down because it might have cannibalized their other offerings (it's not even clear that it would have). This country has become a collection of anti-innovative large companies that rely on rent-seeking and government bailouts to stay in business. I hope China kicks the crap out of our flabby, lazy semiconductor industry, they brought it on themselves.
Shame when frankly Jensen said he still relies on Taiwan productivity. That's clearly what is missing in the $3 trillion pitch. only after federal subsidies, the human capital investment doesn't exist
Sounds exactly like the Kodak story. Kodak eliminated the digital camera tech that one of its engineers made, because it would destroy its film business lol
You are spot on, the system against the progress and innovation now in the west, why because there is no consequence doing bad things or being incompetent, not the case in China.
It's funny then how westerners always argue that Chinese companies can't be competitive and innovative because they're government owned and receive government funding. US companies receive the exact same benefits just in different cloth.
You always touch the very deep down of every topics. Chinese define weak or strong by the ability to solve problems, USA defines weak or strong by how much money the company worth, even worse, by how much the company worth on the stock market. It is very very different value system. Which one works? as daily, normal working class normal people, the answer is obvious.
Because the Central bankers clearly put these companies hostage. By banks I mean BIS, IMF, ECB by stirring the left ideology such as the letter S out of ESG. They're not bankers just plain cartels that China didn't get exposed to luckily..
Actually, I really feel that Westerners don't understand us Chinese people, I have to say it. When faced with threats and persecution, we become more united and diligent. At the same time, Chinese people are willing to sacrifice themselves for the country. I agree that China should not follow the American path. After all, the American path is all about profits for capitalists, and ordinary people are always exploited. In the past, I admired democracy in America, but as I've gotten older and seen more truths, it's also because I no longer believe in the democratic system in America. China is unwilling to engage in war and doesn't want to be at odds with the United States. But when we are pushed to the point where we have no choice, please forgive us for standing united, for our own happiness and for the glory of our country. This is the mindset of many Chinese people. I believe.
""... Chinese people are willing to sacrifice themselves... [America's] ordinary people are always exploited": oh no, back to nationalism and and textbook communist propaganda, aren't you? Then why have thousands of Chinese risked their life coming to US - Mexico border, asking for asylum? Why would they want to be "exploited" by the capitalists? When are they going to sacrifice themselves for China? After making 1st lots of $$ in the US?🤨🤣🤣
I was reading about light in chips back in college in the mid 1990s. The book was published in the 1970s, it was largely theoretical. No one really solved the technical issues of getting one to work. The chips we have use electrons. Using electrons require more energy, and it is slower than using light. Intel developed silicon photonic optical transceivers which are used in data centers, they are much faster than using the electron based transceivers. This is why they have fiber optic cables, fiber optics also use light. If Chinese scientists can solve all the technical issues of developing a photonic chip then it would be ground breaking.
Actually most has been solved in the last years, you just did not read up. This is the problem. We also had the first graphene based chips - just the manufacturing is not there.
@@dannylo5875 "I, Human: AI, Automation, and the Quest to Reclaim What Makes Us Unique" by Tomas Chamorro-Premuzic. Totally forgot it, but - I asked one of my AI agents to find it ;) All that stupid crap about how the world changes - we live in those changes now, suddenly, and it speeds up.
it is like everyone knows the theory and concept of E equal to m C square, and make it out is another story. And how many years this nuclear energy equation has been? And how many countries can make it out?
The U.S. is working against itself to save oil, coal, and the car industry. Our new chip factories will probably be obsolete around the same time China is releasing new photon chips. Legacy auto factories will be obsolete too, even with their new EV factories. These factories are slated to come online in two to three years and by that time China cars will be better and even more cheaper than they are today. Do these companies and politicians think people can afford high-priced vehicles forever especially with automation rapidly displacing workers? If we do not stop tariffs and understand the world will move on without us, we will hurt every future generation for years to come and may never recover all because we lack the vision from our own greed.
You can partly blame the problems due to ignorant brainwashed everyday Americans for demonizing everything China Russia Iran etc etc, and the politicians just consumed those sentiments into the US geopolitical agendas. Another part of those are the lobbies tentacles of the big corporations oil n gas, pharmaceutical, MICs, Healthcare, Autos etc etc Americans are so screwed...
The West should hack and steal information on a scale the Chinese do. Just in case of the small chance that Chinese chips will be better than the western ones😮
Wait, you think the American leadership doesn't know this? American government spending is 99% about enriching connected friends, and the best way to do that is worthless spending. Bridges, to nowhere, a dysfunctional military, Gazan piers that fall apart. People in the right places makes SO much money.
Western tech giants have outrageously underestimated China's might to pioneer semiconductor chips on the global stage. In fact, the sheer multitude of skilled Chinese STEM engineers really helps facilitate their robust economic growth in the foreseeable future.
No country with logic is going to trust CCP tech, it is a Trojan horse and the CCP have become pariahs. Replace the CCP with a government that the world will trust and the problem is solved. The west and US are not the bad guys.
It is foreseeable that the future world leader will be a country with the best and most scientists, professors, engineers and production line operators, plus a stable government that pursues a policy of peaceful development while maintains a military force that is sufficient to deter any foreign military interference. China is clearly the most competitive candidate.
I read long ago, China graduates more PHd and has more geniuses per capita than any other country. The shear size of it's population means it is destined to rule the planet, via technology. Anti-gravity, photonic chips, AI will become their tool of dominance, while they fly above us in motherships...
You hit the nail on the head. Many inhabitants of the DC swamp would not believe you when you tell them that China has 4 X the population of the US, and 8 X the number of Engineers. DC believes that the USD will always be the worlds reserve currency and that issuing world wide sanctions will always work.
Money is not meant to control people rather it is meant to be put to work producing more money for you. You cannot build wealth without putting money in its rightful place.
Thanks for the advice! I'm new to financial planning and wasn't sure where to start. Any tips on finding a reliable financial adviser or resource to guide beginners?
As a beginner, it's essential for you to have a mentor that is verified by finra and SEC to keep you accountable. I'm guided by a widely known financial consultant Stacey Macken
Its the other way round the banks and corporations write all the laws and rules to suit themselves,!!! What?? Do you honestly believe an imbecile like Lindsay Graham can write anything let alone laws??
Damn! You always come through with some off the wall observation that shakes my faith in us. Maybe our folks better get off the pot and find a new and better game. Thanks for the information and the scenery that I'll never see with my own eyes. Bill K.
The photonic chips are far from getting into the mass market. 3 yrs is pretty optimistic. But the huge investments by the US govt and Intel is a waste of money, even in the absence of photonic chips. 1) everything made in the US will continue to be 50% more expensive than what TSMC makes in Taiwan. Intel and TSMC's fabs in Ariz will have to lower their price to match Taiwan prices. That means near zero profit for TSMC in the US and perhaps losses for Intel. Without profits, there is no money for R/D for future generation devices. Intel is most at risk.
my family work in the photonics space, and light-based compute is very reasonable. I will do a quick explanation. we've used to communicate with copper wires and radio annata, which were replaced by fiber optics, why did it happen? the reason is that light has orders of magnitudes density w.r.t. the amount of information they can carry intuitively, a glass wire (a fiber) can carry simultaneously different colors of light (red, blue, ultra-violet) which are distinct communication channels, not interfering with each other. whereas a copper wire can only carry electricity, and it's pain to build distinct channels in the same path. that's how you get orders of mag. improvement when information is carried with light using the same property for computing is the next logical step that being said, science projects always start (grant wise) with some unrealistic expectations, that has nothing to do with countries. it could just be a failed promise in the future, but yeah we should understand it more aha
They are China haters to begin with so laowhy86 and SerpentZa just reinforce what they feel. What is the source of their hatred/hostility? Can't say for sure but if I had to guess there's a lot of racism mixed in there.
You’re simply amazing. How do you come with all of that informative contents? Not found anywhere else online I’ve told many of my associates about your work here and they’re quite impressed as well. Thank you
Is old news, is just that the US gov is suppressing it and keeping everyone in the dark. I wrote it on the comment section months ago but was shadow ban.
Nvidia won't be completely destroyed. They will always have the US / EU market. As for everyone else, yeah, they'll probably buy Chinese chips. Especially if the photonic chips can be mass produced for relatively cheap.
You're making these baseless claims from a 7 minute video which cites articles written in a language I guarantee you can't read. Sorry to break it to you but it's taken 50 years for traditional silicon to infiltrate every household, photonic chips aren't up ending the market in 5 😂. You're talking about an entire worlds worth of infrastructure that runs completely on silicon, you cant just interface everything with light because it sounds cool.
@codejunki567 You think because it took 50 or 100 years...this one will be the same 😂??? The world is now a global village and very easy to change anything. The best they can do is sanction but has it ever worked? Russia is now the 4th largest economy 😂. A country at war slapped with all the sanctions.
At present photonic chips are fantastic for anything that can be processed in batches, like photo recognition and analyzis, or anything involving gazillions of bits of information (I imagine meteorology and organic chemistry testing). It will basically work as ROMs and EPROMs which will interface with electronic computing. They will be black boxes that will perform thousands of times faster specific operations within a computational system. This will be a gigantic step forward, bypassing Moore's law and making all fab manufacturers become essentially junk.
Over 35 years ago The western world electronic institutes were experimenting with photonic chips, but somehow they slowed down and came to a halt..I had my own opinions at that time, of what would happen if the photons could not reflect the light from one of the photons, to the other photon would this render the chip unusable. Whereas the normal electric chips only have to transfer electrons alike the normal nerve cell lines in human beings, and copper wire..The other electric chip is the Field effect chip has to transfer voltage. Back in those early days the pioneers were at the front line of something new that they invented, but maybe they were a little bit stumped to come up with new ideas to improve the photonic chip.
The slow down in advancing upon photonic chips is all due to economics and ROI (return on investment). Management didn't see the immediate benefits in the economic returns as the technology didn't produce immediate results and so, funding was withdrawn as the idea was declared to be unworkable. It was far more profitable to continue with conventional production using copper tracks to transport electrons over to the switching gates. Hell - they will simply make chips tinier and more dense and keep pushing the envelope in what they know than play around with some fancy futuristic device that may or may not work or take far too long before one will see the financial benefits. This is the same sort of "intelligence" which befell Bell Labs when the first solid state transistor was invented leading to the inventors striking out on their own and becoming multi-billionaires - years later. Upper management never seem to be able to take the long road but it may also be due to their responsibility to their shareholders who are demanding immediate returns as well.
I like this guy, and he provides some interesting content on China. Like many people who don't understand much about technology, he gets very excited about prospective breakthroughs. Also, he ignores realistic timelines for implementation of new technology in practical business settings. We shouldn't look at the idea of photon based chips and believe that they will be implemented "in three years". Keep in mind that TSMC builds new fabs in Taiwan in 2 years, and the first TSMC fab in the US took 5 years to complete. Photon based chips will take years to develop and move towards volume production. Customers will have to design chips based on this new technology. This industry does NOT move fast. China is pushing the semiconductor industry forward because of its need to develop a domestic supply chain in response to US restrictions. They continue to develop domestic capacity for semiconductor manufacturing materials. Would you have them give up on this because of theoretical possibilities related to photon based chips? Personally, I'm looking forward to seeing how long it takes China to develop domestically produced EUV lithography machines. Now THERE'S a challenge!
what a joke, he did not speculate about the timeline, he is just stating the fact given by the Chinese lecturer. all that you said above is just your personal opinion which is BS as compared to what he said. even the 3 years' timeframe was given by a reputed Chinese professor, meanwhile, who are you? keyboard expert? there is no need to argue because 3 years is not very long, in fact, next year will be very interesting, lots of surprises, real soon now.
@@johntse8655 When it comes to timeframes for the deployment of yet to be developed technology, don't believe the "reputed professor". Believe the R&D director, R&D mini-line manager, and fab engineer. They will smile at you when you put forward your statements about how three years is not very long, and next year will be full of surprises. I am certainly not an expert in the semiconductor industry. I DO train a wide variety of employees from many different departments and locations in what most consider the most advanced and cutting edge semiconductor manufacturing company in the world. Training high level managers, deputy directors, etc. in communication skills in English involves getting them to speak out about their situations and the industry as a whole. I have heard much, and I have probed about the industry, enabling detailed communication from my trainees and satisfying my own curiosity. I also have a personal interest in the industry and the geopolitical situation that has driven the US to try to stop China from obtaining advanced semiconductor manufacturing machinery, a policy that I have always disagreed with. I remember in 2019 when the US government put pressure on TSMC through the Taiwan government to stop supplying Huawei with 5nm wafers. I thought that the US had no business interfering in the business between two companies, neither of which was American. The situation has gone far beyond this now, and the US is pressuring ASML, Tokyo Electric, Applied Materials, LAM Research and a whole slew of semiconductor machinery manufacturers to kill their earnings power by not selling to their number one market: China. Yes, China will develop a domestic supply chain. Yes, it will take a great deal of time. Yes, they will make some steps next year. Yes, China will be the next place that develops EUV lithography technology, which is now made by only ONE company globally, ASML, which is forbidden to sell the machines to China, as the US pressured the government of Holland to deny ASML an export license for China. Will China develop and deploy EUV lithography machines next year? Not a chance. As I said, this industry moves slowly, and developing, testing and deploying machinery on the road towards volume production is a slow process. TSMC will begin volume production of 2nm wafers at two different locations in 2025. They are well ahead of their competitors, Samsung and Intel, and they have won the 2nm battle. What is next? Lets' see.
Optical technology is very promising and will alway be. Comparing Digital GPUs to a Optical IC is a apple orange comparison. The optical processing are very fast for spectral analysis (requires light source, a len for the transformation and light sensor for output) the output is an analog level. Given the size of AI neural networks (1M's nodes) this would be difficult using optical in parallel unless integrated on to a single chip. Not sure how to train a optical neural network either.
"I don't know, I could be wrong and therefore I need to find out" is an attitude that westerners greatly lack. It's seriously, the lack of curiosity and lack of awareness of the possibility that they're wrong that is so common, and the result of so many issues that China can hit back on.
3 years is a long time to try and predict things in tech. I don't doubt that China will surpass the US soon, whether or not these chips pan out. There are just more Chinese people and they're more motivated. The US should accept that it will happen and figure out how to make the most of it. But it won't.
Hi Kevin, this is a report very worthy of attention for Western audience. Glad you have brought it to the fore. If you are familiar with Chinese language media, it was well discussed within. As a matter of fact, Huawei, the frontier of Chinese tech company, I believe, has patented photonic chip technology. As always, so glad to watch your great program!❤👍👍👍👍👍😍👏🙏🔥❤️
China are not the only country building photonic chip FABs, Also these new chips will not instantly replace all the purely electronic chips as they only so far do some functions and normally need integrated electronics. I must admit the rapid progress being made on photonics is exciting for the future.
3 years is about right from proven R&D results to final industrial application. It will always boil down to cost effectivesness and the availability of the materials and resources needed to move into mass industrial production and manufacturing. Good Luck to those who always remain as skeptics and miss the boat.
Fact: The Chinese don't have access to NVIDIA's best chips which is the reason why Xi Jinping recently mandated that all of NVIDIA Chips be done away with earlier this year. Photonic Chips have been under development for decades. My Alesis MMT-8 Sequencer (1988) has an "Optocoupler" that sends signals to the MIDI Output Port. I know this because I had to change it out and when I did, I put a socket on the board in case it failed again. Point being, this concept isn't new and the work is going on outside of China. The difference between China and the USA is that China has built more high speed rail than the rest of the world combined. The USA celebrates division, derision and outright ignorance; they practices leave the USA entirely dependent on international talent to make progress.
Kevin, gallium nitride (GaN) substrate is an order of magnitude more expensive than silicon substrate. Also, I consulted a semi-conductor expert (a seasoned researcher in the industry with Ph D from the most prestigious university in the US in this field). He pointed out that photonics is nothing new and doesn’t have the capabilities to replace electronics in the near future. There aren’t good solutions for photonic logic and memory devices. I’d appreciate it if you could post the paper from ChingHwa you cited.
Yes photonic is old, but this is not an argument. If Chinese claims are true (why not) then the light logic gates an memory are real : that is the only breakthrough needed.
It's highly likely that intel, Nvidia ....will become obsolete when china starts mass production of photonic chips. Why would anyone still use US chips that are much slower and more expensive
any having used the hybrid called hdmi via fiber optic cable will know china is on the right track . I predict rj45 hybrid fiber optic birth soon . Any fully networked fiber optic till their model saw the massive difference . If you use wifi 6 say meta 3 .
Yup, just like fiber optic compared to data cables, photons chips will be faster and won’t require expensive metals like gold to make and takes less energy. Nothing has been found to be faster than light.
Nvidia doesn't build there own chips, TSMC in Taiwan does! If China has a way to build faster and better chips, Nvidia could build their chips with that technology at a Chinese fab. TSMC would probably license the technology from China to build optical chips too. That's good business. China does like doing good business!
The presenter is a bit naive, failing to notice the vast amount of Corporate Welfare going on in the US Corporate Sector. Much of the US stock market makes no sense on merit. Even Warren Buffett is sceptical of most of it.
It's time for the science of quantum physics to come on board. The ideas are more than 100 years old but we've stubbornly stayed with mechanical physics. Chinese have improved the efficiency of photovoltaic ("solar") panels from 18% to 33%. They've been learning a lot about the dynamics of photons ("light"). It's not surprising that this knowledge will now spread to technologies of everything. In the USA in late 80's-90's we had the beginning of similar understanding as we manufactured improved photovoltaics &had excitement & inspiration behind it. But it was all quashed as the patents were bought up & disappeared & the technology we had was outsourced, PV factories closed.
Is there a way to be investing in this technology being based here in the states. I can’t believe what’s happening with Nvidia, and seeing this sealed the deal. I wouldn’t trust shorting nvidia as so many will blindly throw money at it until it’s too late. Seems like this could be a better alternative.
It makes sense bc Si chips can't be smaller than 1nm while in theory with some other materials it would be possible to go down way to 1 angstrom (0.1nm)
The issue is not the speed and lower energy use of the light based chips. It is the connections from the chip to the outside and the means of communication between chips.
Very cool: ) Electrons are needed to generate photons in a controlled, predictable manner, .... just thinking thru the differences of electrons amd photons. Light bulbs make photons from electricity, for example. Wondering more about the tech. Very cool.
I would not agree that the new fabs going up would be a waste, whether in the US or elsewhere. There is a major need for redundancy in semiconductor production as seen by what happened during the pandemic. There will ALWAYS be a need for conventional chips for embedded applications like automotive, appliances, and the like, despite the advances in photonic and hybrid chips.That being said, it is clear that most of our semiconductor companies have had it way too easy for too long and we absolutely should be questioning how they spend the $50B giveaway our government has bestowed upon them as spoiled lazy brats.
China will be doing it all… 👇 How Close Is China to World Dominance in Legacy Semiconductors? 27-02-2024 | By Paul Whytock * Bread and Butter Technology Obviously, China would like to be a major player when it comes to high-end sophisticated semiconductor devices, but that doesn’t mean they are not interested in the bread-and-butter end of the market, particularly when it comes to legacy products. In fact, they are very interested in the legacy market, and there are some very good reasons why. Legacy devices make up a huge amount of global chip sales. Most chips manufactured today are not advanced chips but legacy chips, and around 71% of devices * China's Aggressive Expansion in the Semiconductor Industry In September 2023, Reuters reported that China was set to launch a new state-backed fund aimed at raising about €43bn to support its chip industry, and according to research analysts, the Rhodium Group, in less than ten years, China is expected to domestically add nearly as much 50-180nm wafer manufacturing capacity as the rest of the World. The views of industry analysts and observers vary, but generally speaking, it’s thought that 22 wafer fabs are being built in the country, and there is an overall plan to create a total of 30 new wafer fabrication plants. Many of these will concentrate on the production of legacy devices. As for market share, industry intelligence gatherers Trendforce believe China’s legacy chip manufacturing base could provide as much as 30% of the global demand for older devices. ElectroPages
USA will need the fabs to process chips for washing machines and dishwashers to be 100% made in USA....gradually the skills to built power driven chainsaws and electric jack hammers will come home to roost😂❤
@@othmanmajid6380 Americans talk about buying American they just don’t want to pay for it Most polls I see over 80% of Americans don’t want to pay up to 10% more for a “made in the USA” vs imported I’m thinking really 10% 🙄🤔🙄 👇 Americans want U.S. goods, but not willing to pay more: Reuters/Ipsos poll A Reuters/Ipsos poll released on Tuesday found 70 percent of Americans think it is “very important” or “somewhat important” to buy U.S.-made products. Despite that sentiment, 37 percent said they would refuse to pay more for U.S.-made goods versus imports. Twenty six percent said they would only pay up to 5 percent more to buy American, and 21 percent capped the premium at 10 percent. Reuters
Nvidias valuation is based on it being an AI stock. If Huawei AI chips and photonic chips are way better than Nvidias product then Nvidias valuation will drop significantly. If the US stock market hasn’t crashed by then then the arrival of the Huawei AI chip will crash the US stock markets.
Huawei is a private company that doesn’t go on stock market. They can’t and will not sell to the American market, they are not welcome there, they are well aware of that.
“First off, photonics chips will never replace electronics ICs, and I say this as someone who designs lasers for photonics chips. I’ve discussed the details reasons for this statement, and some applications for photonics chips in a previous answer[1] but the gist of is that transistor densities are several orders of magnitude higher than optical switch densities because optical components have to be larger than the wavelength of light to be able to confine light without much loss. Electronic ICs are too good and too cheap to be replaced for general purpose computing, logic, digital and and analog circuits. Moving data at high speed is one thing which electronics can’t do well due to fundamental physics, and this is where photonics chips have a niche advantage. Photonics ICs are more useful for moving data, doing simple repetitive matrix operations, and in sensing applications. Even among photonics chips, there are two types-silicon photonics chips, and III-V photonics chips. Silicon photonics chips have much better economies of scale than III-V chips because of greater wafer sizes and the use of mature silicon processes for most components, so they are much cheaper than III-V chips. Currently, only American companies have high-volume silicon photonics foundry and fab facilities. American companies and universities like Intel and UCSB pioneered silicon photonics technology in the late 2000s, and Intel has been making silicon photonics chips in high-volume manufacturing since nearly a decade. Other recent companies that can fab photonics chips are Tower Semiconductor and GlobalFoundries, which are also in the US. Showing something in a lab is one thing, but having a high-volume manufacturing process in place is quite another. These American companies have been doing this for far longer so their processes are higher yielding and more mature than anything in China, to the best of my knowledge. The news release about China’s photonics chips seems to be about III-V photonics chips, from what I gathered, and this technology isn’t all that new. American companies like Infinera have been perfecting this since decades.” Karan Mehta. A Ph.D in Electrical engineer and a laser research engineer at Intel.
The most important development is the fusion reactor. Photonics technology enables AI by making it viable to mass market, which is essential to enhance data flow and speed up data collection worldwide. G6 and Satellite networking helps distribute that information to and from data centers strategically placed in orbit around the earth. AI then implements various strategies to implement small scale prototypes of fusion reactors. Whoever achieve feasibility in this field will be able to sustain colonies for human research and development on the moon or in space where the resources are infinite. We need to think out of the box, because humanity has yet still too much to accomplish together. However, we need to stop thinking of competition for profit or hegemony and think of it like an olympic competition that seek advance the limits of our capabilities as human race, using both brain power in addition to body strength this time around.
So in 3yrs basically most of the existing traditional chip foundries will see their stock values plummet? Taiwan economy will be in tatters due to over-reliance on TSMC? Seems like a future to behold! 😊
Photon. AI. Possibly quantum. These go together like pasta & meat sauce. I've seen this for a decade. Looks like d Chinese saw it too, way before me, and ran with it!
No worries. Money is no object to Americans, they could just turn on the press 24/7, and the world would happily accept. Exorbitant privileges as said by Charles De Gaul
@@peanut0brain if the current Saudi crown prince 🤴 were to get too cheeky, he would soon be replaced. Pretty sure something has already been fermented in the Kingdom to keep the Saudis in line. This happened before in the 70’s when the Saudi king stopped selling oil to the US and he got assassinated ASAP by one of his cousins and replaced with a pliant king who knew his place when it comes to the relationship with the US. This will happen again, pretty sure of it. Unipolar hegemonic power can’t be kept being nice and accommodating to subordinates.
You know how the Americans works. Announced first, works later. The Chinese is the opposite. Work first, announced later. So, if the Chinese already said that, that's gonna happen. If the Americans said that, not sure if I still live at the time they really made it.
Your comment is a very succinct description of how things are. Manufacturing processes have probably already been mobilised and when these are robust and cost effective they will be scaled up ready for integration into products. And, only then will the Chinese probably confirm that they have actually achieved their goals. Also, under circumstances where the west has warmaking designs, these chips are more than likely being tested to enhance weapons technologies at the moment. Regards
hmm ... photonic chips could be interesting. For instance .. a single photonic "circuit" could do parallel processing simply by working simultaneously on different parts of the spectrum. /just random, and probably faulty, thinking ps: i'm not worried about China making all this progress and inroads. I welcome it. Anything that disempowers the belligerent, international dictating US Empire is a good thing.
It's worth noting that Huawei's chips are 20% faster in some instances, and 20% slower in others. This is compared to NVIDIA's A100 chips, which are from 2020. The H100 (from 2022) has three times the performance of the A100, while only being 1.5-2x more expensive. According to lambda labs it can be up to 8x faster than A100 for training LLM's using FlashAttention and a multi-GPU setup. Mind you that LLM's are the reason the hype for AI started, and it's a major reason so many GPU's are being sold. In other words: For the main usage of these GPU's they are 8x faster than the A100 (which is comparable to Huawei's GPU's according to Huawei). This performance is leaps and bounds ahead of any competitor, and is one of the main reasons why NVIDIA's stock has gone up so significantly in the past two years! The reason Chinese companies prefer Huawei over NVIDIA GPU's are I imagine two-fold: 1. Huawei is China based,. This has multiple benefits. For example it is not affected by geopolitical tensions to the same degree, it's favorably viewed by the Chinese government, etc... 2. The latest and greatest chips from NVIDIA are not allowed to be sold to China, because The USA banned the sale of the most advanced chips to China. So the H100 is not available (at least legally) in China, which is very beneficial for Huawei. *TLDR;* Yes Huawei GPU's are comparable to the A100 NVIDIA GPU's (at least according to Huawei), but the H100 is 3-8x faster than the A100 at only 1.5-2x the cost, making it significantly better.
If you understand the true history of the East and the West, you will know the root of the West's fear of China. This is related to the "competitiveness" of scientific and technological culture.
Hi Kevin , good content though sometimes too bearish about progress outside China. In the case of optical chip applications to neural networks, many US starts are striving to make prototypes. Competition drives innovation.
As long as Wall Street remains dominant, manufacturing will never earn more than finance. Based on the profit principle, no one will engage in manufacturing work. Tesla's greatness is not because it is an American company, but because it has placed its largest production base in China. China's stock market has not grown for more than ten years. The benefits of this are obvious. No one wants to get something for nothing. Only when resources are invested in real industries can a country become prosperous and strong.