Semiconductor - Focus on Supply Chain Resilience

Over the last five years, the supply chain issues in the semiconductor sector have come to the forefront. The reliance on a small number of suppliers for critical components has led to significant rises in lead times and prices has had a lasting impact on semiconductor production. Lead times have risen from 7 to 8 weeks in the first half of 2016 to around 12-16 weeks by 2019. The Clock and Timing ICs in particular have seen lead times almost double growing from 10 Weeks in 2016 to around 19 weeks by 2019. This dramatic rise has been seen across the discrete component spectrum with lead times growing to over 20 weeks. This rise in lead times has been corresponding with a growth in the semiconductor demand in the latter half of the 2010s. The growing preference for chip functionality over led times or speed of delivery has also led to an increase in lead times. However, the major cause continues to remain unpredictability in material supply. The rising need to meet customer deliveries on time will in particular lead to an increased need for supply chain efficiency improvement.

The supply chain stability has also taken a nosedive in recent years. The rapidly evolving environment in the semiconductor sector is plagued by different problems such as enhancing revenue growth and profitability, delivering new products that match customer requirements, and ability to manage global operations effectively. This has created significant bottlenecks in component manufacturing and procurement that has raised supply chain issues even prior to the Covid-19 pandemic. The front end (FE) cycle Times are typically much higher than back end (BE) cycle times. The typical production period of the Front End cycle is typically 6-8 weeks, whereas back end cycle times are much lower at 1-2 weeks only. This results in complex differing inventory requirements at different production processes, which needs additional planning and tracking. The lack of planning in this sector as well as slow implementation of technology that assists in this aspect have been showcased during the Covid-19 pandemic.

The increase in production of new smartphones leveraging 5G capabilities and the introduction of fast selling new gaming platforms, has resulted in these segments consuming the front-end capacity for semiconductor components available. The earlier cancelations by automotive manufacturers due to falling demand has resulted in semiconductor lead time growing resulting in increased bottlenecks in supply. The global semiconductor market has seen substantial growth in recent years of 5.1% over 2019 and demand is set to further rise in 2021 post Covid-19 at an 8.4% Year over year. The growing demand combined with impact of Covid-19 on the manufacturing facilities has presented new challenges in the semiconductor sector.

The Covid-19 pandemic has thrown these issues to the forefront of semiconductor manufacturing. In particular the lead times have been risen dramatically. Components such as MCUs have seen lead times doubled in 2020 going from 12 weeks in 2019 to around 26 weeks by the end of 2020. There is potential for price and lead time rise in 2021 as well driven by shutdown of electronics and semiconductor production combined with rise in price and issues with supply of materials like copper. Components and materials which are particularly dependent on China will be significant bottlenecks in 2021.

The shutdown in particular in China will have significant impact on semiconductor production which will drive up lead times. In 2019, China accounted for over 50% of worldwide semiconductor consumption. Disruption in China will likely have a significant impact on companies across the globe and up and down the electronics and semiconductor value chain. COVID-19 is having a large impact on both consumers as well as semiconductor manufacturing companies. As of mid-2020, travel restrictions particularly in China as well as across the APAC and European markets are significantly limiting the semiconductor industry.  In addition, critical electronics manufacturing hubs were temporarily shut down to limit COVID-19’s spread in China as well as across the globe. This confluence of events is hitting the global supply chain, impacting primarily suppliers due to shortages of materials, components, and finished goods for which companies are significantly dependent on China. The consumers are also being impacted through reduced spending in China on semiconductor-dependent products. This spending has significantly reduced in China as well as across the global markets. Allied industries such as automotive and consumer electronics are the primary sources of demand reduction. 

As demand has fallen significantly due to the Covid-19 pandemic, automotive manufacturers as well as part suppliers have reduced their inventory both in terms of limiting new orders as well as cancelling current pending supply. Semiconductor component vendors have seen significant cancelled orders from automotive manufacturers. This has led to the internal capacity being significantly halted or slowed either intentionally due to order cancellation or as a result of reduced labour availability due to the pandemic. As the shutdown starts to reduce and companies start to operate at a more regular pace, semiconductor companies have looked for other markets to target in the wake of automotive slowdown. Markets focused on catering to the resulting work at home situations from the Covid-19 pandemic have in particular seen significant rise. Markets ranging from data centers and servers through to Telecom and industrial automation markets have seen a shift of focus from semiconductor manufacturers. It however remains to be seen how much focus will be on this sector post Covid-19 as the Automotive and consumer electronics markets are projected to witness a quick recovery. 

The impact across the electronics and semiconductor value chain, from materials to final products, will likely be far reaching. The impact will be particularly hard hitting for companies involved with semiconductor manufacturing due to the significance of the cancelled orders. The COVID-19 pandemic has highlighted the vulnerability and risks of the current semiconductor supply chain model and the semiconductor industry has been increasingly required to transform its global supply chain model. Beyond 2021, semiconductor companies may have more difficulty predicting demand because even greater uncertainty abounds in the Post Covid-19 world. As companies create long-term plans and evaluate potential scenarios the changing market demand scenario will need to be handled in a quick and efficient manner if the semiconductor manufacturing market is to witness stability once again. This will require focusing on new emerging markets and technologies as well as adoption of internal solutions to improve supply chain resilience and efficiency.

As the semiconductor sector struggles to cope with the Covid-19 global pandemic, the procurement teams have been diligently to secure raw materials and components and protect supply lines despite global response challenges. However, vital procurement and supply chain information is often not available or accessible across global teams but rather focused on a single procurement head team. As a result, their response to the disruption has been reactive and uncoordinated, and the impact of the crisis is hitting many of their companies full force. This has led to companies investing to ensure the supply chain is resilient to these disruptions. This includes utilizing the following methods to achieve supply chain resilience.

Supply chains of semiconductor companies are globally fragmented, with key operations dispersed in places as diverse as U.S (design), China (foundry), India (software), and Malaysia (assembly & testing). Globalization has also introduced a change in the customer mix and created diverse product requirements, driven by emerging economies such as China and India. This complex supply chain has resulted in more complex processes and procurement policies. This has led to the need to maintain a global supply chain that is efficient and meets the customer requirements despite the complex nature. In addition to this rapid product commoditization enabled by transition to digital technology and use of modular designs has resulted in shrinking product lifecycles — from 24 months to as low as 9 months. Companies must meet increasing on-time product demand from customers which highlights faster time-to-market requirements, thus requiring companies to have processes that enhance agility in making changes, and allow flexibility to configure or localize products while meeting changing regulatory scenarios. This involves ensuring the supply chain is able to meet market demand either through the current suppliers or offering a way to integrate additional suppliers while ensure time to market and cost do not rise significantly. However, the presence of manually intensive processes in supply chain planning and execution have created gaps that need to be fixed to address the increasing need for supply chain modernization and resilience. Lack of process automation and standardization of design and production has led to increase in delivery and customer commit lead times, and reduced on-time delivery performance even prior to Covid-19. Thus, with semiconductor companies focusing on improving the supply chain planning and execution processes, technology adoption has been shifted to the forefront of industry investment scenarios. This includes technology adoption across the methods of improving supply chain resilience listed above.

In recent years the need for advanced inventory and capacity management has come to the forefront. Classic inventory planning techniques and levels will not work when there is a need to manage the sudden spurts in demand that can occur with consumer devices and changing automotive contracts. Chip vendors need to determine the optimal inventory levels, taking into account short product lifecycles and high demand. The significant lead times in semiconductor supply also need to be taken into account for proper palling of the production cycles. Chip vendors will need to strike a fine balance between having excessive inventory and creating bottlenecks. Inventory Levels will need to be adjusted continuously to keep pace with the rapid phase-outs of existing models. Semiconductor companies will have to expand their capacity in order to cope with the distinct characteristics of the NPI cycles for consumer device and automotive OEMs which include reduced lead times and potentially exponential sales growth. Component lead times have risen from 12-to-15 weeks in 2019 to 30 or more weeks midway through 2020 with some manufacturers reporting 52-week lead times for passive components as well. If these lead times are to be adjusted better planning needs to be included in automated solutions to ensure inventory and demand is met in an efficient production cycle. The need for better Inventory Management has been showcased due to Covid-19 related demand issues in particular. The cancelling of supply contracts by automotive companies have showcased the dire need for better inventory and capacity management capabilities. This will primarily involve automation of the inventory and capacity management process including everything from automation of warehouse and logistics operations through to inventory tracking and new processes and models being deployed in inventory management.

Warehouse Automation has been one of the premier technologies that is being adopted in the semiconductor sector to meet the growing need for better inventory management. The semiconductor industry is highly competitive and cyclical. In order to gain an edge over competition, constant innovation and the ability to adapt quickly, are very important. While on the one hand, semiconductor manufacturers are investing heavily in research and development, there is also a race to improve efficiency in design, manufacturing and operations. For better inventory management, the need to automate the warehouse process is paramount. However semiconductor companies are thinking bigger and looking to adopt enterprise wide automation including automation of data collection process that is needed to improve efficiency. Companies are looking to implement an integrated, company-wide data warehouse that can leverage the organization’s complete data assets and increase operational efficiency. Semiconductor companies have so far focused on adoption of robotics, Automated Guided Vehicles and Automated Material handling systems. However this is changing with more comprehensive solutions being deployed. This includes complete automation in the warehouse aspects focusing on warehouse management solutions as well as complete autonomy solution provision. 

Apart from this, tracking of inventory and parts has also come to the forefront. This includes adoption of hardware to track location and storage of inventory as well as software solutions that track and optimize the inventory management. While generalized inventory management solutions are being adopted significantly due to lower cost, companies have also started looking into adopting manufacturing and warehouse management solutions specifically designed for semiconductor industry that allow better inventory tracking despite. These systems are integrated with hardware such as RFID/RTLS solutions that can track serial numbers and lot numbers, use bin tracking, integration with accounting and order tracking solutions, create sub-assemblies, scan barcodes, and generate sales orders. This is lowing companies to more efficiently and effectively handle inventory.

One of the primary results of the Covid-19 crunch has been adoption of Lean Inventory Management models. Semiconductor companies are implementing lean inventory management techniques to reduce costs, improve flexibility and have more time to focus on their customers. The uncertain demand has led to the need for semiconductor companies to better plan their sales and operations and check the inventory management practices especially in light of growing lead times in the industry. The cost reduction and improved flexibility offered through this lean inventory management model will allow semiconductor companies to plan better sales and serve customers at reduced cost despite the high lead times in the industry.

In the current market, China accounts for a significant portion of the Semiconductor manufacturing sector. The country is especially a prevalent supplier of critical components used in Semiconductor and Electronics sector. However to improve resilience and as a result of the brewing China-U.S trade war, there has been a growing demand to shift manufacturing bases from China to other APAC markets as well as local European and U.S markets. With advanced semiconductors key to powering a wide range of potentially transformative technologies, cutting edge computer chips have become a heated area of geopolitical competition for the 21st century. The brewing U.S-China trade war in particularly targets Chinese semiconductor industry and could have significant repercussions to the Chinese semiconductor manufacturing sectors. However the potential changes in the trade tensions due to the new incoming U.S president could cool down these tensions and limit the shift of the manufacturing base from China. Apart from the trade tensions, the impact of Covid-19 also cannot be understated. As a result of the Covid-19 pandemic, the significant dependence of semiconductor and electronics manufacturing industry on Chinese component supply has been identified as the lead times rose significantly and supply irregularities led to issues with supply of semiconductor and electronics products. Apple has been one of the primary proponents of this shift away from China as the company has in recent years expanded their manufacturing base to India. Suppliers of Apple such as Foxconn and TSMC have also followed Apple to expand their production facilities into South East Asia, particularly Vietnam in a bid to continue to supply key customers such as Apple and mitigate risks from the Chinese manufacturing base. 

This instability in supply chain procurement has led many companies to look for ways to diversify their supply chain that allows them to continue with their production cycle despite instability in component supply and future proof their production. Critical components in particular will need to be diversified through multishoring and supplier diversity practices. Although companies have previous focused on forming long-term relationships with a small pool of reliable vendors, particularly from China, more buyers are exploring the use of multi-sourcing to make sure their orders are covered in cases of future instability. A mix of a diversified supply chain, as well as reliable partners, will be the way forward in the semiconductor sector. Another important proponent to reduce this dependence has been nearshoring. Companies, particularly in Europe and U.S, have been nearshoring as a tactic to dissuade dependence on external suppliers. The U.S government in particular has been pushing for this with an amendment to the National Defense Authorization Act, shows Washington recognizes that boosting domestic manufacturing and R&D is key for national security and global competitiveness in the semiconductor sector. Apart from this the U.S Government provides support for semiconductor-based research grants to U.S. universities or private companies, either directly or through agencies like the Defense Advanced Research Projects Agency (DARPA). It also calls for the creation of a National Semiconductor Technology Center, which could serve as a clearinghouse and organizational control point for U.S.-based chip manufacturing, as well as a critical new research and development facility. Such developments are intended to increase on-shoring and development of local manufacturing supply chains in the U.S

The manufacturing chain for any given semiconductor is extraordinarily complex and relies on as many as 300 different inputs, including raw wafers, commodity chemicals, specialty chemicals, and bulk gases. These components and materials all are processed and analysed a significant number of processing equipment and tools. Those tools and components are sourced from around the world, and are typically highly engineered. Further, a large portion of the equipment utilized in semiconductor manufacturing, such as lithography, wafer processing and metrology machines, rely on complex supply chains that are also highly optimized and complex, where a significant number of different companies deliver modules and components for production. Thus, the complex supply chain needs to be monitored to reduce risk and enhance the production cycle to improve efficiency. This is where Supply Chain Risk management and Blockchain solutions come into the picture. 

It is becoming increasingly important to Reduce and Monitor Risk in the Supply Chain. Solutions such as Data analytics are being utilized to reduce risk and follow a more proactive rather than reactive model to risk management and assessment. Due diligence related to potential suppliers as well as diversification of these suppliers is one of the methods followed. The presence of significant amount of data regarding the manufacturing process and product design and production has also allowed companies to fully integrate data analytics into the risk management process and create a more complex but agile supply chain network. To manage this complex network blockchain technology is being used to ensure smooth functioning of the supply chain and sourcing practices. There are a number of potential use cases where the Blockchain technology can be a value addition in the Semiconductor industry. This primarily has to do with component procurement where counterfeit parts identification is one of the major risk mitigation factors. Blockchain allows for provenance tracking of any asset in the value chain as well allowing improved inventory management. However the major application of blockchain remains in the supply chain where the integration of blockchain in procurement will make the business operations pertaining to the same more efficient by eliminating the intermediaries and the archaic processes. This will also enable better ability to track and manage supplier relationships as well as inventory to enable semiconductor companies to efficiently handle the supply chain and increase resilience.

The Plant Execution management has been a focus area for the semiconductor industry for a long time. However technology adoption in recent years as well as in the short term future will be a major advocate of this need for a highly efficient manufacturing process. One of the drivers has been the significant adoption of automation and IoT in recent years to improve the process efficiencies. However this has been largely focused on hardware integration. The focus in the near term for a number of companies will be on ways to utilize the new data generated from these automation solutions and IoT devices. Soon see the more widespread adoption of advanced analytics in semiconductors. This will be utilized not just in computing solutions but to also enable a number of process efficiencies through the utilization of the significant amount of generated data to ensure adoption of predictive maintenance and bottleneck assessment. This will also allow reduction of lead times in the semiconductor sector as well as ensuring the production efficiency which allows on time delivery for customers. 

Generative design has also been a solution that is being implemented to improve efficiency. Artificial intelligence and machine learning algorithms quickly produce multiple iterations of a design based on predefined specifications, including structural composition, materials, and available manufacturing methods that allows the creation of standardized designs that can maximize the production cycle as well as functionality. Covid-19 has created an additional impetus for speed that could accelerate the adoption of generative design approaches. The adoption of 3D printing, in particular high speed sintering, to shore up vulnerabilities highlighted by the pandemic and improve production efficiency and cost in conjunction with generative and standardized design will be a major part of future manufacturing cycles in the semiconductor sector. The rapid prototyping and production cycle through the usage of High Speed Sintering will reduce time to market and ensure on time delivery in a post Covid-19 world where demand will rise significantly for the semiconductor sector. Adoption of new technologies such as light based manufacturing and nanotechnology in conjunction with the aforementioned technologies will also drive improvements in the semiconductor production in the mid to long term as well.

The supply chain mapping in the current semiconductor sector is typically done manually. However this makes it a very time consuming and complex process. The semiconductor companies are looking to focus on an Automation of Mapping process through the utilization of advanced analytics and blockchain solutions. The utilization of analytics and blockchain has allowed companies to utilize internal data and supplier relationships to map out these suppliers and sub suppliers even with a diversified supply base that enhances the transparency in an industry which is shifting to a more visible supply chain. The urgent need for change has been primarily driven by the U.S-China trade war. This trade war has resulted in a significant need to understand the sourcing process of the procured components to allow companies to understand the sourcing from China. This has in particular been exacerbated by the Covid-19 pandemic and thus advanced analytics and blockchain have been increasingly adopted. In the procurement space, sophisticated algorithms can analyse vendors based on price, quality, and cycle time to ascertain the optimal mix of procurement. Advanced analytics also help determine the supply chain visibility issues to allow a more transparent supply chain in an industry which is looking to be less reliant on China. 

In a post Covid-19 world, the semiconductor industry is increasingly focusing on enhancing supply chain resilience. This will require a complete overhaul of the current manual supply chain processes as well as adoption of new models to enhance the visibility and procurement process. New technologies such as 3D printing, blockchain, analytics and automation will be at the forefront of the investment over the next few years as brewing geopolitical tensions along with component procurement lead times and instability will drive adoption.