Department of Commerce on Semiconductors
In September 2021 the Biden administration, concerned that the semiconductor shortage had the potential to extend into 2022 and beyond, petitioned the Department of Commerce to request information from the semiconductor supply chain in order to understand where the bottlenecks and other issues might be causing supply problems, with the intent of creating a better semiconductor supply strategy for the US. The DOC received over 150 responses from almost every major semiconductor producer along with response from many semiconductor industrial consumers. Here’s a summary of what they found:
While we expect the information gathering was well intentioned, organizations like SEMI and the SIA, along with a bit of research, could have likely pinpointed the same weak spots across the semiconductor supply chain, but the real problem with government sponsored bean-counting is that it takes forever to reach a meaningful conclusion and even longer to implement a solution, should one be available. The semiconductor supply chain is global and in most cases is governed by private industry (China the exception), and as such it moves quickly toward any venue that will produce profits, far faster than any government, so government semiconductor spending mandates are always going to be behind both the technology and the market itself.
It is our belief that rather than trying to protect and rapidly rebuild the US semiconductor business to make it self-sufficient, we should be seeding the parts of the US semiconductor industry that the US dominates and using that expertise to lock in the foreign semiconductor capacity we need. The US dominates the EDA market and is on the forefront in terms of new silicon process innovation. By partnering with foundries who need those innovations the US can gain the quid pro quo capacity it needs when demand is strong while growing its own capacity on a more normalized basis, rather than trying to capture growth cycles that tend to be similar to how long it takes for new capacity expansion. Stop wasting time and taxpayer money collecting facts that are obvious, and spend the money on semiconductor innovation where the risk is high but the reward is greater and carries considerable leverage. JOHO
“We’re going to capitalize on this new information to engage industry on node-specific problem-solving in the coming weeks, and we will continue the Early Alert System to monitor and take action related to pandemic-related disruptions to the supply chain.”
[1] Inventory levels for 25 intermediate/end users of semiconductors measured in days of supply. 2021 data is as of September 2021. Based on the 160 products with inventory data identified by respondents that ‘present the greatest challenge for your organization to acquire’.
- Average semiconductor demand was ~17% higher in 2021 than it was in 2019 although buyers saw little change in what they were able to receive from suppliers.
- Median semiconductor inventory[1] has fallen from ~40 days in 2019 to less than 5 days in 2021.
- The primary bottleneck appears to be wafer production capacity, followed by material, assembly, test, and packaging, with certain semiconductor application specific products (legacy logic, analog semis, and opto-electrical semiconductors) in very short supply.
- Specific production node products at 40nm, 90nm, and 180nm, along with broader issues at other mid and higher nodes, seem to be lacking, with no mention of capacity issues at smaller node geometries.
While we expect the information gathering was well intentioned, organizations like SEMI and the SIA, along with a bit of research, could have likely pinpointed the same weak spots across the semiconductor supply chain, but the real problem with government sponsored bean-counting is that it takes forever to reach a meaningful conclusion and even longer to implement a solution, should one be available. The semiconductor supply chain is global and in most cases is governed by private industry (China the exception), and as such it moves quickly toward any venue that will produce profits, far faster than any government, so government semiconductor spending mandates are always going to be behind both the technology and the market itself.
It is our belief that rather than trying to protect and rapidly rebuild the US semiconductor business to make it self-sufficient, we should be seeding the parts of the US semiconductor industry that the US dominates and using that expertise to lock in the foreign semiconductor capacity we need. The US dominates the EDA market and is on the forefront in terms of new silicon process innovation. By partnering with foundries who need those innovations the US can gain the quid pro quo capacity it needs when demand is strong while growing its own capacity on a more normalized basis, rather than trying to capture growth cycles that tend to be similar to how long it takes for new capacity expansion. Stop wasting time and taxpayer money collecting facts that are obvious, and spend the money on semiconductor innovation where the risk is high but the reward is greater and carries considerable leverage. JOHO
“We’re going to capitalize on this new information to engage industry on node-specific problem-solving in the coming weeks, and we will continue the Early Alert System to monitor and take action related to pandemic-related disruptions to the supply chain.”
[1] Inventory levels for 25 intermediate/end users of semiconductors measured in days of supply. 2021 data is as of September 2021. Based on the 160 products with inventory data identified by respondents that ‘present the greatest challenge for your organization to acquire’.
RSS Feed