The DRAM market exhibits a healthier and more balanced supply/demand relationship compared with the NAND Flash market because of its oligopolistic structure, according to TrendForce's latest investigations. The percentage distribution of DRAM supply bits by application currently shows that PC DRAM accounts for 13%, server DRAM 34%, mobile DRAM 40%, graphics DRAM 5%, and consumer DRAM (or specialty DRAM) 8%. Looking ahead to 1Q21, the DRAM market by then will have gone through an inventory adjustment period of slightly more than two quarters. Memory buyers will also be more willing to stock up because they want to reduce the risk of future price hikes. Therefore, DRAM prices on the whole will be constrained from falling further. The overall ASP of DRAM products is now forecasted to stay generally flat or slightly up for 1Q21.

Semiconductor manufacturing is a risky business. Not only is it heavily capital-intensive, which means that even some state-backed would-be players can fail in pooling together the required resources for an industry break-in; but the entire nature of the manufacturing process is a delicate balance of materials, nearly-endless fabrication, cleanup, and QA testing. Wafer manufacturing can take months between the initial fabrication stages through to the final packaging process; and this means that power outages or material contamination can jeopardize an outrageous number of in-fabrication semiconductors.

Recent news as covered by DigiTimes place one of Micron's fabrication plants in Taiwan as being hit with a 1-hour long power outage, which can potentially affect 10% of the entire predictable DRAM supply for the coming months (a power outage affects every step of the manufacturing process). Considering the increased demand for DRAM components due to the COVID-19 pandemic and associated demand for DRAM-inside products such as PCs, DIY DRAM, laptops, and tablets, industry players are now expecting a price hike for DRAM throughout 2021 until this sudden supply constraint is dealt with. As we know, DRAM manufacturers and resellers are a fickle bunch when it comes to increasing prices in even the slightest, dream-like hint of reduced supply. It remains to be seen how much of this 10% DRAM supply is actually salvageable, but projecting from past experience, a price hike seems to be all but guaranteed.

They say that it is hard to keep up with Moore's Law, however, for the folks over at Taiwan Semiconductor Manufacturing Company (TSMC), that doesn't seem to represent any kind of a problem. Today, to confirm that TSMC is one of the last warriors for the life of Moore's Law, we have information that the company has completed building its manufacturing facility for the next-generation 3 nm semiconductor node. Located in Southern Taiwan Science Park near Tainan, TSMC is expecting to start high-volume manufacturing of the 3 nm node in that Fab in the second half of 2022. As always, one of the first customers expected is Apple.

Estimated to cost an amazing 19.5 billion US Dollars, the Fab is expected to have an output of 55,000 300 mm (12-inch) wafers per month. Given that the regular facilities of TSMC exceed the capacity of over 100K wafers per month, this new facility is expected to increase the capacity over time and possibly reach the 100K level. The new 3 nm node is going to use the FinFET technology and will deliver a 15% performance gain over the previous 5 nm node, with 30% decreased power use and up to 70% density increase. Of course, all of those factors will depend on a specific design.

Silicon manufacturing is starting to get harder and harder every day, with new challenges appearing daily. It requires massive investment and massive knowledge to keep a silicon manufacturing company afloat. No company can survive that alone, so some collaborations are emerging. Today, thanks to the sources of Nikkei Asia, we have information that Taiwanese Semiconductor Manufacturing Company (TSMC) is collaborating with Google to push the production of 3D chip manufacturing process, that is said to overcome some of the silicon manufacturing difficulties. The sources also say that AMD is involved in the process as well, making Google and AMD the first customers of the advanced 3D chip design. The two companies are preparing designs for the new way of creating silicon and will help TSMC test and certify the process.

TSMC will deploy the 3D silicon manufacturing technology at its chip packaging plant in Miaoli, which is supposed to do mass production in 2022. With Google and AMD being the first customers of new 3D technology, it is exciting to see what new products will look like and how they will perform. The 3D approach is said to bring huge computing power increase, however, it is a waiting game now to see how it will look like.

IC Insights' November Update to the 2020 McClean Report, released later this month, includes a discussion of the forecasted top-25 semiconductor suppliers in 2020. This research bulletin covers the expected top-15 2020 semiconductor suppliers (Figure 1).

The November Update also includes a detailed five-year forecast through 2024 of the IC market by product type (including dollar volume, unit shipments, and average selling price) and a forecast of the major semiconductor industry capital spenders for 2020. A five-year outlook for total semiconductor industry capital spending is also provided.

Intel continues to "shed fat" on its business portfolio. After last year's sale of its smartphone modem chip business to Apple, the company is now parting ways with its power management circuitry division - Enpirion - and offloading it to Richtek, a division of Taiwanese MediaTek. The sale price of $85 million is a drop in the bucket for Intel's overall bottom line, so it's not a way for the company to cash in some desperately needed money - all accounts of Intel's troubles in the wake of its semiconductor manufacturing issues and AMD's market resurgence pale in comparison to Intel's revenues.

This actually looks like a company that's actually streamlining its R&D expenditures and focusing on execution for the markets Intel sees as most important for today and for tomorrow. Intel's Enpirion focuses on building power management chips for FPGA circuits, SoCs, CPUs, and ASICs, and will now serve to bolster MediaTek's SoC business while allowing the Taiwanese company to expand and diversify its business portfolio, even as Intel focuses on their core competencies.

Apple has recently announced its transition to Apple Silicon, meaning that every processor inside its products will be custom designed by the company. However, that seems to be becoming a bit of a problem. The sole supplier of chips for Apple has been Taiwan Semiconductor Manufacturing Company (TSMC), which Apple collaborated with for the past few years. The sheer capacity of TSMC is enough to satisfy the demand from several companies and thus it allows some of them to book its capacity. With Apple demanding more and more capacity than ever before, it is becoming quite hard to keep up with it. That is why Apple is, according to some analysts for Business Korea, looking for a foundry beyond TSMC's to manufacture its chips.

According to the source, Apple is looking at the direction of Samsung Electronics and its silicon manufacturing facilities. Samsung has recently started the production of its 5 nm silicon manufacturing node. We have reported that the first SoCs are set to arrive soon. However, it may be possible that Apple's M1 lineup of SoCs will be a part of that first wave. Apple is reportedly going to tap both TSMC and Samsung to qualify enough supply for the huge demand of the products based on the latest 5 nm technology.

Taiwanese corporation United Micro Electronics (UMC) has pled guilty on charges of trade theft. The charges, originally pressed in November 2018 by US authorities, placed UMC and China's Fujian Jinhua in hot waters under suspicion of stealing trade secrets from US-based Micron technologies, one of the world's foremost players in memory semiconductor technologies. UMC's guilty plea serves as a way for the company to avoid heavier penalties, and includes a provision for the company's assistance in investigating Fujian Jinhua's actions in regards to this IP theft.

The whole story revolves around UMC's hiring of three Micron employees from Micron's subsidiary in Taiwan, Micron Memory Taiwan (MMT), back around September 2015. At least two of these employees migrated Micron trade secrets to UMC, which then inked a deal with china's Fujian Jinhua for the development of 32nm DRAM and "32Snm" DRAM technologies that Fujian Jinhua could then deploy for the manufacture of memory products - a deal which had Fujian Jinhua paying $300 million for equipment purchase plus $400 million for technology development to UMC. This all fell in line with the Chinese government's Made in China 2025 plan, which aims to bring the country to semiconductor independence from the western world. UMC says that the company itself didn't partake in the underhanded IP delivery to Fujian Jinhua, claiming instead that rogue employees did so of their own volition. The company further states that it only pleads guilty because according to the US Trade Secrets Act, the company still bears legal responsibilities for employee acts, whether or not top management is involved.

Apple has established itself as a master of silicon integrated circuit design and has proven over the years that its processors deliver the best results, generation after generation. If we take a look at the performance numbers of the latest A14 Bionic, you can conclude that its performance is now rivaling some of the x86_64 chips. So you would wonder, what is inside this SoC that makes it so fast? That is exactly what ICmasters, a semiconductor reverse engineering and IP services company, has questioned and decided to find out. For starters, we know that Apple manufactures the new SoCs on TSMC's N5 5 nm node. The Taiwanese company promises to pack 171.3 million transistors per square millimeter, so how does it compare to an actual product?

ICmasters have used electron microscopy to see what the chip is made out of and to measure the transistor density. According to this source, Apple has a chip with a die size of 88 mm², which packs 11.8 billion N5 transistors. The density metric, however, doesn't correspond to that of TSMC. Instead of 171.3 million transistors per mm², the ICmasters measured 134.09 million transistors per mm². This is quite a difference, however, it is worth noting that each design will have it different due to different logic and cache layout.

IC Insights recently released its September Update to the 2020 McClean Report that presented the second of a two-part analysis on the global IC foundry industry and included a look at the pure-play foundry market by region.

China was responsible for essentially all of the total pure-play foundry market increase in 2018. In 2019, the U.S./China trade war slowed China's economic growth but its foundry marketshare still increased by two percentage points to 21%. Moreover, despite the Covid-19 shutdown of China's economy earlier this year, China's share of the pure-play foundry market is forecast to be 22% in 2020, 17 percentage points greater than it registered in 2010 (Figure 1).

Thanks to the latest reportings from DigiTimes, we have obtained information that Apple has been working on expanding its data centers. According to Taiwan's server upstream supply chain, Apple has increased orders for its data centers, which have been growing starting from the second quarter of 2020. For the full year 2020, the orders are expected to be doubled, and the growth in purchasing is going to continue well into 2021. This purchasing decision from Apple comes after strong demand for its services like iCloud, Apple Music, Apple TV, App Store, etc., all of which have millions of users. Being that Apple is expanding its data centers, that must represent a sign that Apple's user base is continuing to grow strongly.

TSMC had been working super hard in the past few years and has been investing in lots of new technologies to drive the innovation forward. At TSMC's Technology Symposium held this week was, the company has presented various things like the update on its 12 nm node, as well as future plans for node development. One of the most interesting announcements made this week was TSMC's state and ownership of Extreme Ultra-Violet (EUV) machines. ASML, the maker of these EUV machines used to etch the pattern on silicon, has been the supplier of the Taiwanese company. TSMC has announced that they own an amazing 50% of all EUV machine installations.

What is more important is the capacity that the company achieves with it. It is reported that TSMC achieves 60% of all EUV wafer capacity in the world, which is a massive achievement of what TSMC can do with the equipment. The company right now has only two nodes on EUV in high-volume manufacturing, the 7 nm+ node and 5 nm node (which is going HVM in Q4), however, that is more than any of its competitors. All of the future nodes are to be manufactured using the EUV machines and the smaller nodes require it. As far as the competitors go, only Samsung is currently making EUV silicon on the 7 nm LPP node. Intel is yet to release some products on a 7 nm node of its own, which is the first EUV node from the company.

The last cyclical upturn in DRAM contract prices began at the start of 2020 and was led by server DRAM, according to TrendForce's latest investigations. In 2Q20, the emergence of the COVID-19 pandemic shocked the global economy, but OEMs maintained or even stepped up procurement of components because they feared disruptions in the supply chain. As a result, DRAM suppliers' bit shipments surpassed expectations for the quarter, in turn widening the overall increase in DRAM ASP and raising the global DRAM revenue by 15.4% QoQ in 2Q20 to US$17.1 billion.

Nevertheless, TrendForce indicates that server OEMs are now carrying a rather high level of DRAM inventory after aggressively stocking up for two consecutive quarters. At the same time, customers of enterprise servers are holding back on procurement because the economic outlook is getting bleaker and more uncertain. Since server DRAM has the unique role of leading cyclical changes, this category is going to be first to experience price drop in the next downturn and thereby pull prices down for other types of DRAM products. As such, TrendForce forecasts at best a flattening of product shipments and decrease in DRAM prices in 3Q20, with DRAM suppliers suffering a decline in profitability.

A Nikkei investigative report uncovered that two Chinese semiconductor fabrication firms, namely Quanxin Integrated Circuit Manufacturing (QXIC), and Hongxin Semiconductor Manufacturing Co (HSMC), have poached over 100 veteran semiconductor engineers from TSMC since last year. Both firms are recipients of government funding under China's ambitious plan of complete electronics hardware industry independence by 2025. Both firms were floated as recently as 2017, and began hiring specialist engineers and executives with connections across the semiconductor industry, from TSMC. The two began development of a 14 nm-class FinFET node that would support manufacturing of a wide variety of electronics components, including SoCs, ASICs, transceivers, and storage products.

Nikkei estimates that in a span of a year, Taiwan lost more than 3,000 semiconductor engineers to various start-ups in the mainland, including large semiconductor fabs. Sources in TSMC tell the Japanese publication that the company is "very concerned" about the flight of talent toward China, although it didn't believe that there is any immediate danger to the company's output or technological edge. The source advocated a national-level strategy by various Asian governments to retain talent, not through coercion, but by offering better incentives and pay than the Chinese firms flush with public investment.

With its own 7 nm-class silicon fabrication node nowhere in sight for its processors, at least not until 2022-23, Intel is seeking out third-party semiconductor foundries to support its ambitious discrete GPU and scalar compute processor lineup under the Xe brand. A Taiwanese newspaper article interpreted by Chiakokhua provides a fascinating insight to the the new precious resource in the high-technology industry - allocation.

TSMC is one of these foundries, and will give Intel access to a refined 7 nm-class node, either the N7P or N7+, for some of its Xe scalar compute processors. The company could also seek out nodelets such as the N6. Trouble is, Intel will be locking horns with the likes of AMD for precious foundry allocation. NVIDIA too has secured a certain allocation of TSMC 7 nm for some of its upcoming "Ampere" GPUs. Sources tell China Times that TSMC will commence mass-production of Intel silicon as early as 2021, on either N7P, N7+, or N6. Business from Intel is timely for TSMC as it is losing orders from HiSilicon (Huawei) in wake of the prevailing geopolitical climate.

Taiwan Semiconductor Manufacturing Company, called TSMC shorty, has just become the world's biggest semiconductor company. The news broke after TSMC's stock reached a peak heights of $66.40 price per share, and market capitalization of 313 billion US dollars. That means that the Taiwanese company officially passed Intel, NVIDIA, and Samsung in terms of market capitalization, which is no small feat. And the news isn't that surprising. TSMC has been rather busy with orders from customers, just waiting for new spots so they can grab a piece of its production pipeline.

TrendForce, a market intelligence provider, estimates that TSMC has an amazing 51.9% of global semiconductor foundry share alone. That is no small feat but TSMC worked hard over the years to make it happen. With constant investments into R&D, TSMC has managed to make itself not only competitive with other foundries, but rather an industry leader. With 5 nm already going in high-volume manufacturing (HVM) in Q4 of this year, the company is demonstrating that it is the market leader with the latest node developments. Smaller nodes like 3 nm are already in development and TSMC doesn't plan to stop.

TSMC, one of the largest semiconductor manufacturing foundries, has officially confirmed that it will stop all orders from Chinese company Huawei Technologies. The Taiwanese silicon manufacturer has decided to comply with US regulations and will officially stop processing orders for Huawei on September 14th of this year. Precisely, the company was receiving orders from HiSilicon, a subsidiary of Huawei Technologies that focuses on creating custom silicon. Under the new regulation by the US, all non-US companies must apply for a license to ship any American-made technology to Huawei. Being that many American companies like KLA Corporation, Lam Research, and Applied Materials ship their tools to many manufacturing facilities, it would be quite difficult for Huawei to manufacture its silicon anywhere. That is why Huawei has already placed orders over at Chinese SMIC foundry.

The Taichung District Court in Taiwan ruled in favor of Micron Technology in a case dating back to 2017, against semiconductor foundry UMC and two of its former employees. Micron had alleged that UMC and three of Micron former employees stole Micron's trade secrets and conveyed them to Mainland Chinese DRAM maker Fujian Jinhua IC. One of the accused include Stephen Chen, former president for Micron Memory in Taiwan.

The Court ruled that the three former Micron employees serve 4.5-6.5 years in prison, in addition to paying NTD 4-6 million fines, each. As for UMC, it has been ordered to pay NTD 100 million (USD $3.4 million) in damages to Micron Technology, a parking ticket value compared to the commercial damage FJIC will inflict to Micron in the years to come.

The information has been circulating the industry about Apple and its plans to integrate Micro-LED technology into its products for some time now. Micro-LED is a new display technology that allows for much better characteristics of panels like higher brightness, fuller color gamut, and higher dynamic range. The technology also offers a higher refreshing rate, wider viewing angles, and lower power consumption. Being that Micro-LED display is physically thinner, it will allow for more portable devices and it will leave more room for other components.

According to the report from CENS, a Taiwanese media, Apple is investing $330 million into a new Micro-LED factory for Apple devices. A future factory located in Taiwan is seeing major investments from the company. Being that Apple plans to use Micro-LED displays in its future products like iPad Pro and MacBook Pro, it is expected to see this move from them. However, Apple isn't doing it alone. They have partnered with Epistar, Taiwan's biggest LED producer and Taiwanese LCD panel maker, AU Optronics. These two are experienced makers of panels so they should help Apple successfully launch and operate manufacturing facilities.

TSMC has just recently announced that they will be building a semiconductor factory in the US, thanks to the pressure from Trump administration. The 5 nm Fab will be built in Arizona, with construction starting in 2021. It will be finished in the year 2024 when the plant will operate at a capacity of 20,000 wafers per month. This is not a high number as TSMC Fabs usually operate at a rate of 100-150K wafers per month, however, the amazing thing is the location of the Fab. The US Fab in Arizona is set to change the global landscape of the semiconductor supply chain, as per the latest report from DigiTimes Research.

Arizona is a place in the US where lots of companies are building semiconductors. Intel, Raytheon, Microchip, ON Semiconductor, VLSI, Freescale, NXP, STMicroelectronics, Honeywell, Marvel, Amkor, Philips, and Western Digital have their facilities there and Arizona can be considered one of the key places for semiconductor manufacturing in the US. With TSMC adding their manufacturing facilities to that list as well, there could be a change in the supplier ecosystem. In light of the need for TSMC 5 nm Fab, the world's leading OSAT (Outsourced Semiconductor Assembly and Test) suppliers may be encouraged to set up local production in Arizona to help TSMC with its plans. A lot of OSAT providers are headquartered in Taiwan, however, if there is a need, they are possibly going to build their manufacturing facilities in Arizona. This alone could change the way semiconductor manufacturing facilities are supplied, and the US could become a major center of OSAT providers.

TSMC, as one of the largest silicon manufacturers in the world, has been subject to pressure from the Trump administration to build a Fab and manufacture silicon on US soil. The reasoning behind this is that the US government could order chips that are supposed to be used in military applications. For security reasons, they need to be manufactured on US grounds and "checked" by the US government. However, it seems like a Taiwanese company has no concrete plans to realize the building of the US Fab.

Thanks to the report of DigiTimes, TSMC has confirmed that they have resisted requests from the US government, and will not build a Fab on US soil for the government. They haven't dismissed the possibility of building one or silicon manufacturing facilities in the US completely. TSMC chairman Mark Liu has told DigiTimes previously that if the company wants to build a US Fab, it will do so because of consumer demand, not the government demand. And that is understandable. It is much easier to work with regular customers compared to the US government which would force a company to go through rigorous security levels to deliver chips.

Power Logic, a graphics card cooling solution OEM, in an interview with Taiwan tech industry observer DigiTimes, commented that it expects graphics card shipments to rise in the second half of 2020, on the backs of new product announcements from both NVIDIA and AMD, as well as HPC accelerators from the likes of Intel and NVIDIA. NVIDIA is expected to launch its "Ampere" based GeForce RTX 30-series graphics cards, while AMD is preparing to launch its Radeon RX 6000-series "Navi 2#" graphics cards based on the RDNA2 graphics architecture. Power Logic has apparently commenced prototyping certain cooling solutions, and is expected to begin mass-production at its Jiangxi-based plant towards the end of Q2-2020; so it could begin shipping coolers to graphics card manufacturers in the following quarters.

A DigiTimes premium report, interpreted by Chiakokhua, aka Retired Engineer, chronicling NVIDIA's move to contract TSMC for 7 nm and 5 nm EUV nodes for GPU manufacturing, made a startling revelation about NVIDIA's recent foundry diversification moves. Back in July 2019, a leading Korean publication confirmed NVIDIA's decision to contract Samsung for its next-generation GPU manufacturing. This was a week before AMD announced its first new-generation 7 nm products built for the TSMC N7 node, "Navi" and "Zen 2." The DigiTimes report reveals that NVIDIA underestimated the efficiency gains AMD would yield from TSMC N7.

With NVIDIA's bonhomie with Samsung underway, and Apple transitioning to TSMC N5, AMD moved in to quickly grab 7 nm-class foundry allocation and gained prominence with the Taiwanese foundry. The report also calls out a possible strategic error on NVIDIA's part. Upon realizing the efficiency gains AMD managed, NVIDIA decided to bet on TSMC again (apparently without withdrawing from its partnership with Samsung), only to find that AMD had secured a big chunk of its nodal allocation needed to support its growth in the x86 processor and discrete GPU markets. NVIDIA has hence decided to leapfrog AMD by adapting its next-generation graphics architectures to TSMC's EUV nodes, namely the N7+ and N5. The report also speaks of NVIDIA using its Samsung foundry allocation as a bargaining chip in price negotiations with TSMC, but with limited success as TSMC established its 7 nm-class industry leadership. As it stands now, NVIDIA may manufacture its 7 nm-class and 5 nm-class GPUs on both TSMC and Samsung.

Huawei's subsidiary, HiSilicon, which designs the processors used in Huawei's smartphones and telecommunications equipment, has reportedly moved its silicon orders from Taiwan Semiconductor Manufacturing Company (TSMC) to Semiconductor Manufacturing International Corporation (SMIC), according to DigiTimes. Why Huawei decided to do is move all of the 14 nm orders from Taiwanese foundry to China's largest silicon manufacturing fab, is to give itself peace of mind if the plan of the US Government goes through to stop TSMC from supplying Huawei. At least for the mid-tier chips built using 14 nm node, Huawei would gain some peace as a Chinese fab is a safer choice given the current political situation.

When it comes to the high-end SoCs built on 7 nm, and 5 nm in the future, it is is still uncertain how will Huawei behave in this situation, meaning that if US cuts off TSMC's supply to Huawei, they will be forced to use SMIC's 7 nm-class N+1 node instead of anything from TSMC. Another option would be Samsung, but it is a question will Huawei put itself in risk to be dependant on another foreign company. The lack of 14 nm orders from Huawei will not be reflecting much on TSMC, because whenever someone decides to cut orders, another company takes up the manufacturing capactiy. For example, when Huawei cut its 5 nm orders, Apple absorbed by ordering more capacity. When Huawei also cut 7 nm orders, AMD and other big customers decided to order more, making the situation feel like there is a real fight for TSMC's capacity.

TSMC, Taiwan's flagship manufacturer of silicon, has seen a substantial increase in demand for Chip-on-Wafer-on-Substrate (CoWoS) packaging technology, according to the report from DigiTimes. CoWoS is a multi-chip packaging technology that gives an option to build silicon like LEGO, allowing for dies to be placed side by side on interposer that is providing high interconnect density and performance. You can see more about CoWoS in detail here. Some of the examples of CoWoS are NVIDIA's P100 and V100 dies that integrate logic (computing elements), and memory (in the form of HBM) on a single die.

Recently, TSMC updated its CoWoS technology, where this new second-generation parts could scale far larger than the first-generation implementation - up to 1700 squared millimeters of die space, allowing for some very creative solutions to be implemented. This may be the reason that the demand in Q2 has risen so substantially and that TSMC's production lines are now running at full capacity, trying to meet the demand for this packaging technology.