According to Counterpoint Research’s latest Global Smartphone Quarterly Shipment Forecasts, total units shipped for 2021 are expected to grow by only 6% annually to 1.41bn units; Counterpoint had previously called for 9% annual growth to 1.45bn units. Tom Kang, Research Director at Counterpoint Research said in a statement, “the semiconductor shortage seems to affect all brands in the ecosystems. Samsung, Oppo, Xiaomi have all been affected and we are lowering our forecasts. But Apple seems to be the most resilient and least affected by the AP shortage situation.”
The smartphone industry was set for a strong rebound this year after COVID-19 had hit the market hard in 2020. Smartphone vendors placed large component orders from the end of last year, and consumer demand coming from delayed replacement purchases buoyed the market in the first quarter. However, some smartphone OEMs and vendors are reporting they had only received 80% of their requested volumes on key components during Q2 2021, and the situation seems to be getting worse as we move through Q3 2021. Some smartphone makers are now saying they are only receiving 70% of their requests, creating multiple problems. Counterpoint Research believes 90% of the industry is affected and this will impact the second half forecast for 2021.
Semiconductor shortages had been plaguing the market since Q4 2020, but the smartphone industry had managed to grow despite shortages in components like DDIs and PMICs. This was done by advanced planning and order placing along with hoarding of certain components like Application Processors (AP) and camera sensors which are typically much higher value than DDIs or PMICs.
Semiconductor shortages continue and despite foundries running at full capacity for several quarters, the smartphone industry is being affected. Components that were once fully stored in the warehouse are bottoming and new components are not coming as requested.
In the case of application processors, one of the most crucial elements in smartphones, the shortage was triggered by low yield rates in newly established fab lines. With the situation seeming to persist it caused a chain reaction throughout the industry. AP vendors like Qualcomm and Mediatek rely on these foundries and manufacturing problems result in fewer processors supplied which in turn affects smartphone OEMs.
MacDailyNews Take: Karma works in mysterious ways.
Apple was screwed over by a Samsung and as a result built up TSMC to turn them into a world beating semiconductor fabrication facility and as a consequence, Apple is treated as a priority customer by TSMC.
Just imagine how precarious Apple’s situation would be these days if it still had to rely on Samsung for most of its semiconductors. I can’t imagine Samsung prioritising Apple if it’s own phones needed chips.
Smart phone after all is a tiny gadget compared to say, auto assembling. The heart of iPhone is all assembled on SoC, which I assume are assembled by TSMC. Assembling of subaasemblies of certain camera components (not necessarily the lens alone) may have been affected in countries like Vietnam, but I do not hear any complaints from Sony. Compared with auto industry, which is a typical assembly business that has so many supply chains including their own (subsidiaries etc) and independent contractors, smart phone assemblers are relatively small (by scale, not sales). Their components (large, small, heavy, light, complicated shape of sheet metals, machine components, semiconductors for vehicle control units as well as for power control, you name it) are so many times more than what is required to assemble smart phone. Beside, they typically practice just-in-time system to reduce inventory and warehousing cost, on top of eliminating assembly downtime.
A smart phone assembler like Apple can stock up and warehouse qty of required components (I believe that’s what they do) and gradually roll out assembled products while watching how sales goes. iPhone roll out has ALWAYS been slow and that’s what Apple (or any other assemblers) does. Once Apple knows the data to plan which model to ramp up, congestion gradually disappears.