If markets were rational, why would we have speculative booms and busts?
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
If markets were rational, why would we have speculative booms and busts?
antiterror13
Brigadier
who said the market were rational? If the market were rational, there won't be any market at all .....
What I was saying people are willing to pay more for Tesla/TSMC shares than Toyota/Intel shares ..... that what matter, and whatever you/me think is unfortunately don't matter at all .....
We are straying away from the topic but this kind of valuation can be highly deceptive. People are buying the shares with the hopes of being able to sell them later for an even higher cost.
who said the market were rational? If the market were rational, there won't be any market at all .....
What I was saying people are willing to pay more for Tesla/TSMC shares than Toyota/Intel shares ..... that what matter, and whatever you/me think is unfortunately don't matter at all .....
Yep.
But what I've always found interesting (and somewhat depressing) is that it is rational to follow the herd, as per the performance studies.
Even if everyone knows a stock is irrationally valued, as long as you are towards the front of the trends, you'll make out well whilst the laggards are left with the losses.
Investors and fund managers who try to buck this get killed off. So built into any market is an inevitable (natural?) tendancy towards irrational highs and lows
Intel also focused too much on server and desktop CPUs. They optimized poorly for low power. This almost cost them the laptop market and eventually cost them the smartphone market.
Intel is in a bit of a bind here. AMD already tried that strategy of being a foundry and it flopped pretty bad. Also, AMD already had some technology to compete in the foundry business, as they sold access to their fabs for a long time as a way to earn extra crash and keep fab usage rate up. Intel's foundry business has thus far been pretty much a flop. Just a couple years they closed it, what makes you think they will do better this time when today they aren't even at the leading edge of process design anymore. Intel bought the major FPGA company, Altera, in an effort to keep their fab usage rate up as a result. Fabs keep getting more expensive and the x86 market is not large enough to keep them fed and justify the investment in process design anymore.
As Intel outsources some of its production to TSMC it will only mean their fabrication and process design unit will become less and less competitive. It won't have the economies of scale to be viable.
Intel could survive as a fabless chip design company but that was never their strength to begin with. Intel used to be a fabrication company first and a chip design company second.
It will be at TSMC since they will own the fabs. Once the contract with Intel expires they can use the fabs for whatever they want too.
Samsung's process is inferior but a lot of companies are using their fabs. The ones you mentioned and others like IBM. IBM's server chips for POWER and zSeries are manufactured at Samsung.
Intel was always going to have a tough time beating ARM in the mobile space. x86 on mobile would have no second source which means vendors would be locked in to Intel. A lot of people forget but when x86 started out IBM forced Intel to license their design so there were several other suppliers including AMD and others. As time went by Intel successively forced these players out of the market and bought these licenses out. VIA is one of the few other companies left standing with a license but even their license should expire soon.
Their only way out would be Intel selling their processors to a company like Apple which controls the operating system, hardware design, application store, and has a captive client base. But Apple has long since went their own way in designing chips so Intel no longer has any chance of that.
If Intel wanted to go after Qualcomm they should have made their own ARM compatible design. At one point Intel had the best ARM chip design in the market, XScale formerly known as StrongARM, which they got as a result of the purchase of DEC's chip design unit. This was used in PDAs like the Compaq iPAQ. Intel let it fester and die and now they are paying for it.
Yes, but relevant to the topic on hand, the valuation is not really based on the reality of semiconductor design and manufacturing. It's not like TSMC has ultra technology that no one can develop, just that the market conditions favour their current capabilities. If the market shifts to RISC-V or something, they might not enjoy the same competitive advantage.who said the market were rational? If the market were rational, there won't be any market at all .....
What I was saying people are willing to pay more for Tesla/TSMC shares than Toyota/Intel shares ..... that what matter, and whatever you/me think is unfortunately don't matter at all .....
I think the strategy for Intel in this case is to buy time.
You buy up TSMC's capacity and look to build up better capability and servicing of customers for the fabs.
Furthermore, any success on this front will allow for a greater separation of the business units to exist independently.
There would also be some hope that the US Geopolitical situation might force some onshoring.
IMO, I don't think you are right about some of the business factors.
It's not the lack of second source that was the issue (IMO). Qualcomm basically has a monopoly on the non-Apple ARM market in the US, if not the whole Western world. Not even Samsung is interested in selling their own Exynos based phones in North America.
As I mentioned, Qualcomm was actively blocking development of Intel-based phones.
Second, from the software side, please correct me if I'm wrong because I am not a software dev, but as far as working with Android, Intel had to keep up to date with making sure everything was working with the x86 on its own, but the OS by default is working with ARM CPUs.
Finally, I did not mention, but by 2012 Qualcomm had integrated the modem onto the SoC (Snapdragon S4), by basically not selling a separate modem to non-Qualcomm users (exception being Apple of course), this basically killed all those "second sources" you mentioned, TI OMAP, even nVidia Tegra. nVidia tried to buy a modem designer, Icera, but failed to bring a product to market.
Intel tried to alleviate this by buying the Infineon modem business, but they struggled to develop a power efficient all-global bands 4G modem solution (I don't know the technical reasons for this), but it was too late anyway.
Those competition you mentioned were largely irrelevant by the end of the 486 era/transition to P5-Pentium. Those companies like AMD and Cyrix all struggled to develop FPU performance on par with Intel. Furthermore, they were hamstrung by their inability to design some of the underlying architecture such as front side bus (again, I only have some rough knowledge of the technical aspects). Finally, just like Qualcomm did to them, Intel paid off the big OEMs, Dell, HP, Compaq, IBM, Gateway, etc. to never use AMD or Cyrix CPUs.
The x86 competition from this era is littered with slain corpses, Transmeta, NextGen (became AMD K6), IDT/Centaur (bought by VIA). It was not licensing that killed these companies, they were simply technically and financially inferior to Intel. So, again, it was there was no competition, these guys could not compete. (I actually owned a 233 Mhz AMD K6 computer)
As I pointed out above, XScale would have been irrelevant. It probably would be another also-ran without the modem. OMAP was the original Android CPU and didn't make it. XScale was sold to Marvell and became Armada... don't even think it got to market in handsets.
Benefiting from factors such as the substantial increase in the amount of orders, China Micro's (AMEC) 2021 net profit is expected to increase by 93.01%-109.26%
China Micro Company announced that it expects a net profit of 950 million yuan to 1.03 billion yuan in 2021, an increase of 93.01% to 109.26% year-on-year; the company will sign new orders for MOCVD equipment in 2021. A year-on-year increase of about 90.5%.
The company expects that the net profit attributable to the owners of the parent company in 2021 will be 950 million yuan to 1.03 billion yuan, which will increase by 457.808 million yuan to 537.8008 million yuan compared with the same period of the previous year (statutory disclosure data), a year-on-year increase of 93.01% to 109.26%.
The company expects that the net profit attributable to owners of the parent company after deducting non-recurring gains and losses in 2021 will be RMB 280 million to RMB 330 million, which will increase by RMB 256.6806 million to RMB 306.6806 million compared with the same period of the previous year (statutory disclosure data). Yuan, an increase of 1,100.72% to 1,315.13% year-on-year.
...
I thought it was a pretty good article.
You know what they say brother, people who understand economics, not that many. People who understand tech, very few.
That is why businesses really value people who do understand, and can put those two together. I think that guy at stratechry.com does a real good job of explaining it, because he understands both.
Often we find people who can understand one or the other, but not both.
If we want to find people who understand neither, all we need to do is look at the US government, the House of Congress and Corruption. That is where all those people work, who write all the laws.
That, that is a true story!
What mo-fos! You know what I mean!?
The author (Ben Thompson) has an undergraduate major in English; he is based in Taiwan. He did work for Apple, Microsoft before.
So he is not a subject matter expert in semiconductor, maybe in the market or the software, but not in the physical fabrication.
Even if the design changed to a different architecture it will still have to be fabricated somewhere. Due to the costs involved it will be much cheaper to use a foundry than to build your own fabrication facilities. I do not see how RISC-V will change the current dynamic towards foundries being more relevant. Also I think ARM pretty much has the smartphone market locked in by this point. RISC-V would have to come through in things like cheap IoT devices and the like if it is to have a chance I think.
But Intel already tried selling foundry services and failed. There is no evidence they will do better this time. I think this is just talk by Intel so they can get US and European funds to increase capacity.
You just contradicted yourself. Apple designs their own ARM processors. Also Qualcomm does have lots of competition besides Samsung with Exynos among Android CPU vendors. MediaTek for example is a major competitor which even has their own modem designs.
Yes that would be another issue with trying to put x86 on mobile.
MediaTek AFAIK has their own 5G modem. Even UNISOC has their own 5G modem. As did HiSilicon when they were still around.
Intel is not this all conquering company. They routinely fail at a lot of things. This is just yet another one of them.
They tried designing their own modem. When it took too long and cost too much they got rid of the division. This is typical Intel.
- Status