Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Overview

The Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) market size is estimated to reach $XXX million by 2026. Furthermore, it is poised to grow at a CAGR of XX% over the forecast period of 2021-2026. The Hybrid Memory Cube (HMC) is the high-performance RAM interface through-silicon via (TSV) technology by connecting the multiple memory arrays to the top of one another. It uses standard DRAM cells for memory implementation. The High-bandwidth Memory (HBM) is also the RAM interface but for 3D stacked SDRAM. It is used in combination with network devices and graphic accelerators. Both memory bus is capable of providing higher bandwidth with lower power consumption. It is widely applicable in graphics designing, in the field of networking, and others.

Major factors driving the growth of the HMC and HBM market include the growing need for high-bandwidth, low power consuming, and highly scalable memories, increasing adoption of artificial intelligence, and rising trend of miniaturization of electronic devices.

By product type, the market for APU to grow at the highest CAGR during the forecast period.

The market for APUs is expected to grow at the highest rate during the forecast period. HBM-based APUs are a recent innovation by AMD (US) developed to meet the requirements of high-performance computing. APUs integrate both GPU and CPU capabilities on a single SoC. This further improves the overall energy efficiency of APUs by eliminating connections between chips. APUs can also be used for graphics applications. Moreover, AMD (US), the leading manufacturer of APUs, demonstrated an APU with integrated HBM and stacked non-volatile memory cells. This will also serve to drive the adoption of APUs in computing applications.

By Application, HMC and HBM market for graphics applications to grow at the highest CAGR during the forecast period

The HMC and HBM market for graphics applications is expected to grow at the highest CAGR during the forecast period. A majority of the products available in the market with the HBM technology are GPU products. HBM was initially adopted in GPUs specifically for graphics applications. For instance, AMD (US) developed the HBM technology along with SK Hynix (South Korea) to be used in GPUs. Along with GPUs, APUs have also been introduced in the market that is increasingly being used for gaming applications. The increasing adoption of HMC and HBM in gaming is largely due to the increasing requirements to process vast amounts of pixels for larger screens and support higher compute rates for more stabilization to support high-end gaming.

North America to account for the largest market share during the forecast period.

The high adoption of HMC and HBM memories in North America is largely due to the growth in high-performance computing (HPC) applications that require high-bandwidth memory solutions for fast data processing. The demand for HPC in North America is growing owing to the increasing market for AI, machine learning, and cloud computing. In addition, major HPC-based CPU and processor providers, such as Intel, are based in North American countries. Other key tech companies such as Google, Amazon, and Microsoft are also headquartered in the US and have served to drive the demand for high-performing CPUs in servers and supercomputers. The major drivers for the rapid growth of the HMC and HBM market in the APAC are the growing number of data centres and servers, increasing shipments of network equipment, and the rising number of manufacturing activities in the enterprise storage and consumer electronics sectors. The strong economic growth and growing demand for high-density memories are also expected to drive the HMC and HBM market in the APAC region.

Market Dynamics

Driver: Growing need for high-bandwidth, low power consuming, and highly scalable memories

The development of various 3D-stacked memories is mainly driven by the rising need for memories offering high bandwidth, low power consumption, and high scalability. With the emergence of Big Data, the Internet of Things (IoT), and other data-intensive applications, there is a growing demand for technologies that can efficiently process and store more information. Also, there is a strong need for memories with high efficiency and performance in the network system for data packet buffering, data packet processing, and storage applications beyond 100 Gbps. HMC and HBM, which provide a bandwidth of more than 100 Gbps, can be a feasible replacement for DRAM as they achieve competitive speeds with much lower power consumption. Also, the market is witnessing continuous advancements in these technologies. For instance, HBM 2 provides 256 Gbps of bandwidth, significantly improving over HBM 1 (128 Gbps of bandwidth).

It has also been observed that HMC and HBM consume about 70% less energy per-bit than traditional DRAM-based memory technologies. For example, at 128 Gbps of bandwidth, HBM 1 consumes only 3.30 watts, which is considerably lower than DDR4 (its preceding memory technology), which consumes 4.48 watts at the same bandwidth. Hence, the demand for HBM and HMC is expected to increase in line with the growing demand for high-bandwidth, low power consuming, and highly scalable memories.

Restraint: Thermal issues caused by high levels of integration

HMC and HBM are a stack of DRAM chips connected internally to TSVs and externally to one or more chips using micro bumps and TSV. Even though these technologies offer several advantages, thermal issues caused by the high level of integration and their impact on the overall module present a major challenge to manufacturers. These technologies offer highly dense multi-level integration per unit footprint, which creates challenges for thermal management. Listed below are some of the major complications associated with these technologies:

• The high level of integration leads to high on-chip temperatures.
• As the memory lies between the heat sink and the logic die, the heat generated from the logic die raises the temperature of the memory.
• Although the compact package offers a shorter route for a signal to travel, the heat generated in the module slows down the movement of the signal once it hits the thermal limit (to protect the circuitry)

A small temperature change has a big impact on the lifespan of the circuit, as metal lines burn out as a result of the increased heat. The typical operating temperature for DRAMs is under 85°C. When the temperature exceeds 85°C, the performance of the memory starts reducing, and it starts consuming more power than usual. Also, more pronounced thermal effects are higher power densities and greater thermal resistance along the heat dissipation paths. Therefore, thermal issues associated with the 3D integration of memories is a major factor restraining the growth of the HMC and HBM market.

Opportunity: Growing Big Data

The introduction of the Internet of Things has led to the generation of a large volume of data. Also, Big Data applications, such as business analytics, scientific computing, financial transactions, social networking, and search engines, are increasing rapidly. All these applications handle large datasets and require high-performance IT infrastructures to achieve fast-processing throughput. Based on the data provided at the National Association of Software and Services Companies (NASSCOM) Big Data and Analytics Summit 2016, about 2.5 quintillion bytes of data are created every day; 90% of the data globally has been created in the last two years. Also, according to the National Institute of Standards and Technology (NIST), a non-regulatory federal agency within the US Department of Commerce, about 45 zettabytes of data is expected to be generated worldwide by 2020. HMC and HBM are well-suited for Big Data applications because they offer ultra-fast storage performance. These technologies also effectively address the performance, bandwidth, and endurance requirements of Big Data applications.

Challenge: Design complexities associated with HMC and HBM

HMC and HBM are still in the early commercialization stage. Currently, there are many manufacturing and fabrication challenges associated with these products, including the placement and routing of ICs. Also, there are new layout rules and layout layers (such as the back-side redistribution layer) and challenges associated with the wafer test for HBM and HMC. The power noise impact from HBM I/Os is one of the most critical design challenges. Although the power per pin is low, there are many I/Os generating noise in parallel, which leads to significant power consumption. These testing and design complexities increase the cost of HMC and HBM memories. As a result, unless these memories are mass-produced, their cost would continue to be higher than the existing memories available in the market. The impact of this challenge is expected to be medium during the forecast period.

Key Market Players

Samsung (South Korea), Micron (US), SK Hynix (South Korea), Intel (US), Advanced Micro Devices (AMD) (US), Xilinx (US), Fujitsu (Japan), NVIDIA (US), IBM (US), Open-Silicon, Inc. (US)

Recent Developments:

• In December 2017, Intel (US) launched the Stratix 10 MX FPGA with integrated high-bandwidth memory HBM2. The product provides a maximum memory bandwidth of 512 gigabytes per second and targets high-end applications such as high-performance computing (HPC), network function virtualization (NFV), and broadcast applications.
• In November 2017, AMD collaborated with Intel (US), a leading developer of advanced and integrated digital technology platforms, to integrate semi-custom GPUs with a multi-chip processor package.
• In July 2017, Samsung started the mass production of its 8-gigabyte (GB) HBM2. HBM2 offers 256 Gbps of memory bandwidth and meets the growing market requirements across many applications, including artificial intelligence, high-performance computing, advanced graphics, network systems, and enterprise servers.

Chapter 1 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Overview

1.1 Product Overview and Scope of Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM)

1.2 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Segmentation By Type

1.2.1 APU

1.3 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Segmentation by End Use

1.3.1 HMC

1.3.2 HBM

1.4 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Segmentation by Regions

1.4.1 North America

1.4.2 China

1.4.3 Europe

1.4.4 Southeast Asia

1.4.5 Japan

1.4.6 India

1.5 Global Market Size (Value) of Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) (2015-2027)

Chapter 2 Global Economic Impact on Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Industry

2.1 Global Macroeconomic Environment Analysis

2.1.1 Global Macroeconomic Analysis

2.1.2 Global Macroeconomic Environment Development Trend

2.2 Global Macroeconomic Environment Analysis by Regions

Chapter 3 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Competition by Manufacturers

3.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production and Share by Manufacturers (2021 and 2021)

3.2 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Revenue and Share by Manufacturers (2021 and 2021)

3.3 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Average Price by Manufacturers (2021 and 2021)

3.4 Manufacturers Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Manufacturing Base Distribution, Production Area and Product Type

3.5 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Competitive Situation and Trends

3.5.1 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Concentration Rate

3.5.2 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Share of Top 3 and Top 5 Manufacturers

3.5.3 Mergers & Acquisitions, Expansion

Chapter 4 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue (Value) by Region (2015-2021)

4.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production by Region (2015-2021)

4.2 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production Market Share by Region (2015-2021)

4.3 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Revenue (Value) and Market Share by Region (2015-2021)

4.4 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.5 North America Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.6 Europe Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.7 China Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.8 Japan Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.9 Southeast Asia Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

4.10 India Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue, Price and Gross Margin (2015-2021)

Chapter 5 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Supply (Production), Consumption, Export, Import by Regions (2015-2021)

5.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Consumption by Regions (2015-2021)

5.2 North America Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

5.3 Europe Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

5.4 China Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

5.5 Japan Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

5.6 Southeast Asia Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

5.7 India Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption, Export, Import by Regions (2015-2021)

Chapter 6 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue (Value), Price Trend by Type

6.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production and Market Share by Type (2015-2021)

6.2 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Revenue and Market Share by Type (2015-2021)

6.3 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Price by Type (2015-2021)

6.4 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production Growth by Type (2015-2021)

Chapter 7 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Analysis by Application

7.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Consumption and Market Share by Application (2015-2021)

7.2 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Consumption Growth Rate by Application (2015-2021)

7.3 Market Drivers and Opportunities

7.3.1 Potential Applications

7.3.2 Emerging Markets/Countries

Chapter 8 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Manufacturing Cost Analysis

8.1 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Key Raw Materials Analysis

8.1.1 Key Raw Materials

8.1.2 Price Trend of Key Raw Materials

8.1.3 Key Suppliers of Raw Materials

8.1.4 Market Concentration Rate of Raw Materials

8.2 Proportion of Manufacturing Cost Structure

8.2.1 Raw Materials

8.2.2 Labour Cost

8.2.3 Manufacturing Expenses

8.3 Manufacturing Process Analysis of Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) 

Chapter 9 Industrial Chain, Sourcing Strategy and Downstream Buyers

9.1 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Industrial Chain Analysis

9.2 Upstream Raw Materials Sourcing

9.3 Raw Materials Sources of Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Major Manufacturers in 2021

9.4 Downstream Buyers

Chapter 10 Marketing Strategy Analysis, Distributors/Traders

10.1 Marketing Channel

10.1.1 Direct Marketing

10.1.2 Indirect Marketing

10.1.3 Marketing Channel Development Trend

10.2 Market Positioning

10.2.1 Pricing Strategy

10.2.2 Brand Strategy

10.2.3 Target Client

10.3 Distributors/Traders List

Chapter 11 Market Effect Factors Analysis

11.1 Technology Progress/Risk

11.1.1 Substitutes Threat

11.1.2 Technology Progress in Related Industry

11.2 Consumer Needs/Customer Preference Change

11.3 Economic/Political Environmental Change

Chapter 12 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Market Forecast (2021-2027)

12.1 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Revenue Forecast (2021-2027)

12.2 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production, Consumption Forecast by Regions (2021-2027)

12.3 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Production Forecast by Type (2021-2027)

12.4 Global Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Consumption Forecast by Application (2021-2027)

12.5 Hybrid Memory Cube (HMC) and High-bandwidth Memory (HBM) Price Forecast (2021-2027)

Chapter 13 Appendix