According to a report by Taiwan’s Commercial Times, JPMorgan’s latest analysis reveals that AI demand will remain robust in the second half of the year. Encouragingly, TSMC’s CoWoS capacity expansion progress is set to exceed expectations, with production capacity projected to reach 28,000 to 30,000 wafers per month by the end of next year.

The trajectory of CoWoS capacity expansion is anticipated to accelerate notably in the latter half of 2024. This trend isn’t limited to TSMC alone; other players outside the TSMC are also actively expanding their CoWoS-like production capabilities to meet the soaring demands of AI applications.

Gokul Hariharan, Head of Research for JPMorgan Taiwan, highlighted that industry surveys indicate strong and unabated AI demand in the latter half of the year. Shortages amounting to 20% to 30% are observed with CoWoS capacity being a key bottleneck and high-bandwidth memory (HBM) also facing supply shortages.

JPMorgan’s estimates indicate that Nvidia will account for 60% of the overall CoWoS demand in 2023. TSMC is expected to produce around 1.8 to 1.9 million sets of H100 chips, followed by significant demand from Broadcom, AWS’ Inferentia chips, and Xilinx. Looking ahead to 2024, TSMC’s continuous capacity expansion is projected to supply Nvidia with approximately 4.1 to 4.2 million sets of H100 chips.

Apart from TSMC’s proactive expansion of CoWoS capacity, Hariharan predicts that other assembly and test facilities are also accelerating their expansion of CoWoS-like capacities.

For instance, UMC is preparing to have a monthly capacity of 5,000 to 6,000 wafers for the interposer layer by the latter half of 2024. Amkor is expected to provide a certain capacity for chip-on-wafer stacking technology, and ASE Group will offer chip-on-substrate bonding capacity. However, these additional capacities might face challenges in ramping up production for the latest products like H100, potentially focusing more on older-generation products like A100 and A800.

(Photo credit: TSMC)

As semiconductor process technology nears known physical limits, the spotlight among major industry players is shifting towards the development of advanced packaging. Concurrently, the rise of applications like artificial intelligence and AIGC has propelled the concept of advanced packaging into a new technological wave. In the midst of the semiconductor industry’s global competition, securing more orders has become a core objective for major players.

A Competitive Landscape in Advanced Packaging

The competition in advanced packaging technology is intensifying, with companies pouring substantial investments into the field, resulting in a landscape of vigorous competition. Various packaging technologies have emerged, with notable offerings from industry giants such as TSMC, Intel, and Samsung.

TSMC introduced 3DFabric, an integration of its TSMC-SoIC front-end technology with CoWoS and InFO back-end technologies, providing maximum flexibility for diverse innovative product designs.

Intel, on the other hand, features its 2.5D EMIB and 3D Foveros packaging technologies. EMIB is applied in the connection of logic chips and high-bandwidth memory, as seen in the Intel Xeon Max series and Intel Data Center GPU Max series.

Foveros allows top dies to overcome size limitations and accommodate more top and base dies, connected through copper pillars to reduce potential interference from through-silicon vias (TSVs).

Samsung also exhibits strong competitiveness in advanced packaging, with its 2.5D I-Cube4 and H-Cube, along with 3D X-Cube packaging technologies, achieving breakthroughs in multi-chip interconnects and integration.

Samsung’s I-Cube4, for example, integrates four HBM stack dies and one core compute IC on the silicon interposer layer, while H-Cube enhances packaging area through the stacking of HDI PCBs to accommodate designs with six or more HBM stack dies.

Advantages of the Three Giants

In recent years, the three semiconductor giants have directed substantial capital expenditure towards advanced packaging. Their diverse technological developments and marketing strategies are poised to ignite a global battle in the semiconductor advanced packaging industry.

TSMC holds the advantage with its dominant wafer process technology and an end-to-end comprehensive service approach. Coupled with Taiwan’s robust semiconductor ecosystem, TSMC leads the way in the advanced packaging domain.

Intel, while slightly trailing TSMC in advanced process technology, matches it in advanced packaging capabilities. Emphasizing flexible foundry services, Intel allows clients to mix and match its wafer manufacturing and packaging offerings. With manufacturing facilities scattered worldwide, Intel leverages geographic advantages, particularly in Western countries, to expand capacity and services, leading to anticipated gains in the future.

Samsung, like TSMC, offers end-to-end services, but its packaging technology lags behind TSMC’s. It secures a share in constrained supply situations. Notably, Samsung, in June 2022, was ahead of TSMC in unveiling the innovative GAA 3nm process, and is poised to combine it with 3D packaging technology, potentially marking a pivotal point in the next semiconductor generation.

With semiconductor technology’s continuous evolution and surging market demand, the competition among the three giants in advanced packaging will remain fierce. While wafer fabs currently prioritize processes, the next three to five years are expected to witness a gradual shift towards advanced packaging. Different packaging technologies and marketing strategies will ultimately determine companies’ positions and influence in the market.

(Photo credit: TSMC)

Read more:

• An In-Depth Explanation of Advanced Packaging Technology: CoWoS
• ASE, Amkor, UMC and Samsung Getting a Slice of the CoWoS Market from AI Chips, Challenging TSMC

Charlie Boyle, Vice President of NVIDIA’s DGX Systems, recently addressed the issue of limited GPU production at the company.

Boyle clarified that the current GPU shortage is not a result of NVIDIA misjudging demand or constraints in Taiwan Semiconductor Manufacturing Company’s (TSMC) wafer production. The primary bottleneck for GPUs lies in the packaging process.

It’s worth noting that the NVIDIA A100 and H100 GPUs are currently manufactured by TSMC using their advanced CoWoS (Chip-on-Wafer-on-Substrate) packaging technology. TSMC has indicated that it may take up to a year and a half, including the completion of additional wafer fabs and expansion of existing facilities, to normalize the backlog of packaging orders.

Furthermore, due to the significant strain on TSMC’s CoWoS capacity, there have been reports of overflow of NVIDIA GPU packaging orders to other manufacturers.

Sources familiar with the matter have revealed that NVIDIA is in discussions with potential alternative suppliers, including Samsung, as secondary suppliers for the 2.5D packaging of NVIDIA’s A100 and H100 GPUs. Other potential suppliers include Amkor and the Siliconware Precision Industries Co., Ltd. (SPIL), a subsidiary of ASE Technology Holding.

In December 2022, Samsung established its Advanced Packaging (AVP) division to seize opportunities in high-end packaging and testing. Sources suggest that if NVIDIA approves of Samsung’s 2.5D packaging process yield, a portion of AI GPU packaging orders may be placed with Samsung.

TrendForce’s research in June this year indicated that driven by strong demand for high-end AI chips and High-Bandwidth Memory (HBM), TSMC’s CoWoS monthly capacity could reach 12,000 units by the end of 2023. Particularly, demand from NVIDIA for A100 and H100 GPUs in AI servers has led to nearly a 50% increase in CoWoS capacity compared to the beginning of the year. Coupled with the growth in demand for high-end AI chips from companies like AMD and Google, the second half of the year is expected to witness tighter CoWoS capacity. This robust demand is projected to continue into 2024, with advanced packaging capacity potentially growing by 30-40% if the necessary equipment is in place.

(Photo credit: NVIDIA)

Over the past few decades, semiconductor manufacturing technology has evolved from the 10,000nm process in 1971 to the 3nm process in 2022, driven by the need to increase the number of transistors on chips for enhanced computational performance. However, as applications like artificial intelligence (AI) and AIGC rapidly advance, demand for higher core chip performance at the device level is growing.

While process technology improvements may encounter bottlenecks, the need for computing resources continues to rise. This underscores the importance of advanced packaging techniques to boost the number of transistors on chips.

In recent years, “advanced packaging” has gained significant attention. Think of “packaging” as a protective shell for electronic chips, safeguarding them from adverse environmental effects. Chip packaging involves fixation, enhanced heat dissipation, electrical connections, and signal interconnections with the outside world. The term “advanced packaging” primarily focuses on packaging techniques for chips with process nodes below 7nm.

Amid the AI boom, which has driven demand for AI servers and NVIDIA GPU graphics chips, CoWoS (Chip-on-Wafer-on-Substrate) packaging has faced a supply shortage.

But what exactly is CoWoS?

CoWoS is a 2.5D and 3D packaging technology, composed of “CoW” (Chip-on-Wafer) and “WoS” (Wafer-on-Substrate). CoWoS involves stacking chips and then packaging them onto a substrate, creating a 2.5D or 3D configuration. This approach reduces chip space, while also lowering power consumption and costs. The concept is illustrated in the diagram below, where logic chips and High-Bandwidth Memory (HBM) are interconnected on an interposer through tiny metal wires. “Through-Silicon Vias (TSV)” technology links the assembly to the substrate beneath, ultimately connecting to external circuits via solder balls.

The difference between 2.5D and 3D packaging lies in their stacking methods. 2.5D packaging involves horizontal chip stacking on an interposer or through silicon bridges, mainly for combining logic and high-bandwidth memory chips. 3D packaging vertically stacks chips, primarily targeting high-performance logic chips and System-on-Chip (SoC) designs.

When discussing advanced packaging, it’s worth noting that Taiwan Semiconductor Manufacturing Company (TSMC), rather than traditional packaging and testing facilities, is at the forefront. CoW, being a precise part of CoWoS, is predominantly produced by TSMC. This situation has paved the way for TSMC’s comprehensive service offerings, which maintain high yields in both fabrication and packaging processes. Such a setup ensures an unparalleled approach to serving high-end clients in the future.

Applications of CoWoS

The shift towards multiple small chips and memory stacking is becoming an inevitable trend for high-end chips. CoWoS packaging finds application in a wide range of fields, including High-Performance Computing (HPC), AI, data centers, 5G, Internet of Things (IoT), automotive electronics, and more. In various major trends, CoWoS packaging is set to play a vital role.

In the past, chip performance was primarily reliant on semiconductor process improvements. However, with devices approaching physical limits and chip miniaturization becoming increasingly challenging, maintaining small form factors and high chip performance has required improvements not only in advanced processes but also in chip architecture. This has led to a transition from single-layer chips to multi-layer stacking. As a result, advanced packaging has become a key driver in extending Moore’s Law and is leading the charge in the semiconductor industry.

(Photo credit: TSMC)

Looking at the impact of AI server development on the PCB industry, mainstream AI servers, compared to general servers, incorporate 4 to 8 GPUs. Due to the need for high-frequency and high-speed data transmission, the number of PCB layers increases, and there’s an upgrade in the adoption of CCL grade as well. This surge in GPU integration drives the AI server PCB output value to surpass that of general servers by several times. However, this advancement also brings about higher technological barriers, presenting an opportunity for high-tech PCB manufacturers to benefit.

TrendForce’s perspective: 

• The increased value of AI server PCBs primarily comes from GPU boards.

Taking the NVIDIA DGX A100 as an example, its PCB can be divided into CPU boards, GPU boards, and accessory boards. The overall value of the PCB is about 5 to 6 times higher than that of a general server, with approximately 94% of the incremental value attributed to the GPU boards. This is mainly due to the fact that general servers typically do not include GPUs, while the NVIDIA DGX A100 is equipped with 8 GPUs.

Further analysis reveals that CPU boards, which consist of CPU boards, CPU mainboards, and functional accessory boards, make up about 20% of the overall AI server PCB value. On the other hand, GPU boards, including GPU boards, NV Switch, OAM (OCP Accelerator Module), and UBB (Unit Baseboard), account for around 79% of the total AI server PCB value. Accessory boards, composed of components such as power supplies, HDD, and cooling systems, contribute to only about 1% of the overall AI server PCB value.

• The technological barriers of AI servers are rising, leading to a decrease in the number of suppliers.

Since AI servers require multiple card interconnections with more extensive and denser wiring compared to general servers, and AI GPUs have more pins and an increased number of memory chips, GPU board assemblies may reach 20 layers or more. With the increase in the number of layers, the yield rate decreases.

Additionally, due to the demand for high-frequency and high-speed transmission, CCL materials have evolved from Low Loss grade to Ultra Low Loss grade. As the technological barriers rise, the number of manufacturers capable of entering the AI server supply chain also decreases.

Currently, the suppliers for CPU boards in AI servers include Ibiden, AT&S, Shinko, and Unimicron, while the mainboard PCB suppliers consist of GCE and Tripod. For GPU boards, Ibiden serves as the supplier, and for OAM PCBs, Unimicron and Zhending are the suppliers, with GCE, ACCL, and Tripod currently undergoing certification. The CCL suppliers include EMC. For UBB PCBs, the suppliers are GCE, WUS, and ACCL, with TUC and Panasonic being the CCL suppliers.

Regarding ABF boards, Taiwanese manufacturers have not yet obtained orders for NVIDIA AI GPUs. The main reason for this is the limited production volume of NVIDIA AI GPUs, with an estimated output of only about 1.5 million units in 2023. Additionally, Ibiden’s yield rate for ABF boards with 16 layers or more is approximately 10% to 20% higher than that of Taiwanese manufacturers. However, with TSMC’s continuous expansion of CoWoS capacity, it is expected that the production volume of NVIDIA AI GPUs will reach over 2.7 million units in 2024, and Taiwanese ABF board manufacturers are likely to gain a low single-digit percentage market share.

(Photo credit: Google)