TSMC | Lawrence van Rijn - Law Lord to be

Tag Archives: TSMC

September 24, 2024 · 22:59

Two sides of technology

There are always two sides on any technology. The question is whether they are aligned or not. The first story is found at (at https://www.edgemiddleeast.com/ai/tsmc-and-samsung-consider-100-billion-uae-chip-projects) where the Edge Middle East gives us ‘TSMC and Samsung Consider $100 Billion UAE Chip Projects’, it all comes across as straight forward. We are given “Semiconductor giants TSMC and Samsung are in early talks to establish massive chip-making facilities in the UAE, potentially marking a significant expansion in global production.” It seems to me that this is a straight forward option, especially for the UAE. We are also given “develop potential chip projects in the United Arab Emirates, with investments that could exceed $100 billion. The discussions, which are still in the early stages, were first reported by The Wall Street Journal on Sunday” and this article ends with “Should these plans move forward, they would mark a significant milestone in the UAE’s efforts to position itself as a global technology hub.” The second article was initially from the Financial Times (but they are behind a paywall), as such I I cannot give the link, but the headline reads ‘UAE president meets Joe Biden in push for more US AI technology’ where we are given “Sheikh Mohamed bin Zayed al-Nahyan seeks to formalise fledgling partnership between both countries” as well as “The United Arab Emirates’ leader met US President Joe Biden in Washington on Monday to advance artificial intelligence co-operation as the Gulf nation tries to secure easier access to US-made technology” and “The UAE is one of the US’s most important allies in the Middle East, but relations have been strained at times in recent years. Talks for a formal security pact with Washington have stalled, and Abu Dhabi was infuriated by what it saw as a lukewarm US response to attacks on the UAE’s capital by Houthi rebels from Yemen in 2022.” This is a dangerous time for America. The trivialisation of the Houthi terrorists will cost America dearly, it has before and it will cost America more than they imagined. You see, as I personally see it. There is a bigger fish. The option that China will play nice with Taiwan when there is a larger part of the $100,000,000,000 could give China the edge they need. And in this setting China will have several bonus options that would fall away from American. That alone would entice China to play nice with Taiwan to a whole larger degree. Is it viable? I honestly cannot say as the media is massively anti-China. Ask Huawei is you doubt my view on this issue.

How could this happen?
There are several options, but if I were a betting man China would offer Taiwan independence UNDER China. Would Taiwan accept this? I don’t know, but if China would enable a diplomatic solution via the United Arab Emirates it could happen. China is more interested in the collapse of America sooner and will hand an independence ‘option’ to Taiwan. And the setting with Sheikh Mohamed bin Zayed al-Nahyan gives China a larger option to manouver. It is my believe that the Biden administration is driven to not make my speculation happen and for that it needs a slice of the UAE AI business and America will offer whatever it has to to make their entrance a done deal. On the positive side if Microsoft gets involved there is every chance that their affinity to mediocrity will blow up in their faces and the American stance becomes a whole lot weaker. This is not ‘fear mongering’, this is merely the view I have on Microsoft and the blunders they have made in the recent past. The UAE embraces perfection, as such Amazon (AWS) or Google would be a much better fit. But this is not about bashing Microsoft (it is fun though). The AI investments that could be coming the way of the UAE, there is a larger field. We hear all about ‘AI’ and the developers (Amazon, Apple, Google, et al) but most forget that Huawei has its own system. The FusionMind AI platform. I don’t know how good it is. Whatever the media tells us, once Huawei gets to demonstrate their system. No matter what others think, if the UAE considers it good enough, the American race for revenues goes in the wrong direction (for America that is). Don’t ask me how good or how bad the Huawei system is, because I have never seen it, but I know about it and the media is doing its best to ignore Huawei, but I am not convinced that this is a good move to make. The IT people (like me) want to assist people with solutions that WORK. I do not think it is a good idea to ignore the Huawei system. And I believe that neither Saudi Arabia or the United Arab Emirates are ignoring the Huawei technology side of it all. For me the larger setting isn’t merely what works, but it is the dim witted view of accusing Huawei whilst not offering ANY clear evidence. That is the larger stage and if Huawei, or the Chinese government can convince Sheikh Mohamed bin Zayed al-Nahyan to allow Huawei to present their case, American will have additional worries to deal with. I personally think that Google AI with Mandiant would be personally the better option. That is merely because I have have limited exposure to AWS and no exposure to Amazon security solutions. So my view is slightly biased. In all this, Google needs to convince the UAE that they have what the UAE needs. After that Saudi Arabia should be shown these solutions too (likely they have already seen them).

When we see these sides, one side is the technology, the other side is the software and when we optionally see these chip solutions the bigger winner becomes whomever sets the premise of their software to the hardware provided. I personally hope for Google (I am biased here), but the end game is nowhere near concluded at present. I reckon the Biden administration is hoping for a memorandum of intent, but that is something we might see on Wednesday. So keep looking.

It is almost Wednesday here and Vancouver is following in 18 hours. So anything is possible. Have a great day.

The abacus paradox

The stage of what might never be is out there. There is a larger stage of what is about to happen and what might never happen. And I for one have no idea where to stand. You see, it is about what we read and what we do not know, or do we? To get there I need to do an old setting, a simple calculation, like you had in primary school (year 1) the stage is 34-15. Now you do it like there is no tomorrow, but then it was different. In that setting we had four minus five is not possible, so we borrow 10 then it becomes fourteen minus five making nine. Three minus one is two, that minus one is one, one and nine make nineteen, that is how we learned the simplest calculation (in 1969). It is not the math, it is the setting of borrowing one from the left making fourteen instead of four minus five. It is the borrowing part. You see, a setting hit me when I was confronted with ‘Micron to build $7 bln plant in Japan to expand DRAM production – report’ (at https://www.reuters.com/technology/micron-build-7-bln-plant-japans-hiroshima-report-2021-10-20/). In the article we see “U.S. memory chip maker Micron Technology Inc (MU.O) will build a new factory at its Japanese production site in Hiroshima at a cost of 800 billion yen ($7.0 billion), the Nikkan Kogyo newspaper reported on Wednesday.” Under most situations I merely shrug my shoulders. What do I care? Yet when my mind saw this, it raced back to several of these announcements. I believe it was 2018, it could have been 2017. Yet the stage then was that the resources needed for DRAM were limited. A finite equation on resources and the pressure was on. I personally believe that shortage was never dealt with, so it is nice to see a plant being built ‘too address shortage of DRAM’, but if there is still a shortage for manufacturing the problem merely shifts, it does not go away. TechRadar gave us in April “the company wants to optimise the use of its limited supply of silicon”, as one source gives us in May “Demand is much higher for all types of silicone products than one year ago. Demand started to increase in 4Q 2020 as global economies opened. Even when parts of the world were hit with second and third waves of Covid, economies did not shut down as they did in 1Q and 2Q 2020. As of today, demand is the same or slightly higher than prior to the pandemic”, this is not enough. You see it is one of the most common elements in the world, so we should not run out, but the mining and acquiring of it comes with health risks and there lies the rub. Forbes also gives us “China has called chip independence a top national priority in its latest five-year plan, while U.S. President Joe Biden has vowed to build a secure American supply chain by reviving domestic manufacturing. Even the European Union is mulling measures to make its own chips. But success is anything but assured.” There we also see more. The article (at https://www.bloomberg.com/graphics/2021-chip-production-why-hard-to-make-semiconductors/) gives us a lot to think about and if it isn’t the $7,000,000,000 they expect to spend, it is nothing against getting the right people, as such it could be seen as a severn billion dollar long-shot. Yet the stage of getting the Silica Dioxide has not been reached yet. The article also shows us “Building an entry-level factory that produces 50,000 wafers per month costs about $15 billion. Most of this is spent on specialised equipment—a market that exceeded $60 billion in sales for the first time in 2020.” So exactly how does one save on a plant that is already 8 billion short in month one? The numbers make no sense. Now, it is perfectly possible that they are two different plants, but the stage of 7 billion is weird all whilst several plants are being build and there is still a need for Silica Dioxide for the wafers.

Now, I could be very wrong here, and that might be the case, but consider that 2020 gave us ‘TSMC to build advanced chip plant in southern Taiwan’ (May 2021), ‘Bosch opens Dresden chip plant as Europe aims for independence from Asia’ (Jun 2021), and let that not misdirect you, there are hundreds of plants, a massive amount of them in China, the stage becomes not merely who creates the chip, but who will have the plant to mine and create the required Silica Dioxide. It could be another stage for Saudi Arabia as sand is something they have in abundance, but to get one into the other takes time and takes the right secure machines as well as a few other skills. When this implodes (a not unlikely setting) the one who HAS the Silica Dioxide has an advantage. China claimed to be independent at present, most others cannot make that claim. China produces 64% of all silicon on the planet and the US pissed them off. Almost 10% is in Russia that leaves a lot less for the rest of the planet, so when I see the stage of two plants being build, I merely wonder where they get their source materials. The US at present creates less than 5% of all Silicon and the UK even less (number unknown to me). So when we see the larger stage and we look for Mineral Commodity Summaries we might learn that there are a few more kinks in those cables. Yes, I am willing to accept that they MIGHT solve those. Yet when was the last time you made a long-shot using billions on the word ‘MIGHT’?

In all this I might be the one wrong here, but when you see three of these plants being made within a year and you realise the cost involved and there is a stage where the required resources for manufacturing are in many places, yet the bulk is in two places that certain politicians cannot reach. How does that investment sound to you?

I wouldn’t bet my stove on that risk, let alone a house (or apartment) but they are willing to bet the bank, I wonder what optional short term benefits they get before these tactical deciders seemingly run off with a large bonus to a beach far, far away.

P.S. WordPress is still not fixing the colour issue.

Tag Archives: TSMC

Two sides of technology

The abacus paradox

Recent Posts

Archives