Significant progress over the last 18 months, but many challenges remain before ramping up for large volume consumer applications.
- Technology status and trends: epitaxy, microLED efficiency vs OLED, assembly processes and equipment, light management and color conversion
- Yield and defect management: status and roadmaps, analysis of the different strategies, cost analysis
- MicroLED display driving: analog vs. digital, TFT vs. microdriver ICs
- Applications: why are smartphones so hard? Is TV possible? Are smartwatches coming soon?
- Updated adoption roadmap and volume forecast for displays and epiwafers
Key features of the report
- MicroLED technology status
- Competitive landscape and keyplayer’s profiles
- Intellectual property trends
- Supply chain
- MicroLED yield management and repair strategies analysis
- Cost analysis: TV, smartwatch and smartphones
- MicroLED display applications: strength, weakness opportunity and threat (SWOT) analysis, roadmap and forecast for TVs, smartphones, wearables, augmented reality, laptops, tablets and monitors
- MicroLED display panel forecast through 2027
- Wafer demand forecast
Objectives of the report
Understand the status of microLED technologies:
- Recent progress
- What are the remaining pinch points?
- What are the best yield management and repair strategies? Current status and roadmaps
Which applications could microLED display address and when?
- Detailed analysis and roadmaps for major display applications
- Cost analysis
Competitive landscape and supply chain
- Identify the key players and IP owners in technology development and manufacturing. Who’s taking the lead? Key partnerships
- Scenario for a microLED display supply chain
- Impact on the display supply chain
Table of content
What we got right, what we got wrong 11
Executive summary 12
Introduction to microLED displays 49
Recent trends 61
MicroLED display adoption forecast 73
AR/MR and HUDs 82
Tablets, laptops, monitors 111
Epiwafer forecast 121
Intellectual property trends 127
Supply chain 142
Competitive landscape 153
- Major players by technology node (nonexhaustive)
Large company ecosystems 170
- Overview of large companies and their microLED ecosystems
- Taiwan microLED ecosystem
- China microLED ecosystem
Technology trends and status 199
MicroLED efficiency 214
Beam shaping and light management 223
Transfer and assembly 228
Deterministic assembly 231
Pixel and display driving 270
Yield management and repair 297
Yield management and repair strategies 306
Cost aspects 317
Monolithic microdisplays 329
MICROLED: PROGRESSING ON ALL FRONTS
Microscopic light emitting diodes (microLEDs) are drawing an increasing amount of attention. Startups have raised more than $800M to date, including at least $100M in 2019. Apple has spent $1.5-$2B over the last five years. Panel makers such as Samsung, LG, AUO or Innolux have also significantly increased their efforts.
Patent filings are growing exponentially and technology is progressing on all fronts. The external quantum efficiency of blue and green microLED chips has more than doubled over the past 24 months. Some transfer and assembly processes are reaching performance close to what is required to enable some microLED consumer applications.
Progress is also visible in the proliferation of prototypes presented over the last 18 months by close to 20 companies. The demos cover a broad range of display types, sizes and technologies. Native RGB or color converted displays on Thin Film Transistor (TFT) backplanes are offered by many companies, with some examples including Playnitride, CSOT, Samsung, LG, glō, AUO, eLux, and Kyocera. Lumiode has developed native RGB or color converted displays on monolithicaly integrated Low-Temperature Polysilicon (LTPS). Meanwhile such displays on CMOS backplanes are on offer from companies including Plessey, glō, Lumens, JB Display, Sharp and Ostendo. Finally, discrete microdriver ICs have been developed by X-Display. The multiple prototypes based on TFT backplanes give credence to the idea that microLED displays can leverage existing panel maker capacity, thereby simplifying and streamlining the supply chain.
Equipment makers have taken notice and are starting to develop microLED-specific tools for assembly, bonding, inspection, testing and repair. LED makers are also showing interest, with San’an planning to invest $1.8B to set up a mini and micro-LED manufacturing base. Osram, Seoul Semiconductor, Nichia or Lumileds are also increasing their activity and Playnitride is completing its first microLED pilot line.
SIGNIFICANT ROADBLOCKS STILL IN PLACE FOR KEY APPLICATIONS
For many applications, economics is pushing die size requirements below 10μm. This compounds efficiency, transfer and manufacturability challenges and despite significant improvement, small die efficiency remains low. Display efficiency based on this technology still can’t match OLED. Significant effort is therefore needed to further improve the internal quantum efficiency, light extraction and beam shaping of green and red microLED chips.
Epitaxy and chip fabrication are no longer seen as roadblocks, but solid yield management and repair strategies must be implemented. Transfer and assembly processes need to evolve from table-top experiments to robust high-volume production tools. The proliferation of technology paths creates some confusion and delays. Equipment makers can be reluctant to commit. A piece of equipment developed for certain processes or architectures therefore often won’t work with others. Developing process-agnostic tools is challenging. Choosing a technology today is risky, but so is waiting too long to get in the game with an increasingly crowdedintellectual property (IP) landscape
For microLED companies, the first few prototypes provide strong returns in terms of experience, but maturing toward consumer-grade displays could require thousands more. Startups are entering the ‘valley of death’. Many might fail to raise enough money to successfully go through this more capital- and resource-intensive phase. Support and partnership with large display makers or
original equipment manufacturers (OEMs), either as strategic investors or development partners is critical.
The situation is less challenging for microdisplays. Many prototypes can be built from a wafer run, and setting up the supply chain is easier as a lot of steps can be outsourced. Small foundry runs are expensive, however, and non-recurring engineering costs can be significant.
A STRONG CASE FOR AR, SMARTWATCHES AND AUTOMOTIVE, BUT WILL PHONES AND TVS MATERIALIZE?
Smartwatches are a perfect ‘beach-head’. Low volumes, small displays with high price elasticity make it possible to use larger dies and more redundancy. Apple could push high volume manufacturing and make smartwatches a stepping stone to overcoming supply chain obstacles and improve technology toward other applications. Other companies could enter the market sooner with lower volume, lower specification devices. glō is partnering with Kyocera to set up its supply chain and Playnitride expects to ship
passive-matrix wearable displays with its partner RiTdisplay by the end of 2020.
There is also a strong case for augmented reality (AR) and head-up display (HUD) microdisplays where microLEDs could be the only technology delivering the right combination of brightness, efficiency and form factor. More work however is needed to deliver full-color displays and efficient coupling to waveguide optics. For automotive, microLEDs offer a unique and compelling combination of high brightness, contrast, ruggedness and environmental stability, while enabling freeform, conformable displays. Higher price elasticity means microLED could be technology-ready rapidly, but lengthy qualification cycles will delay adoption past 2023.
The TV market is more challenging. OLEDs are progressing and might leave little room for differentiation by the time microLEDs are ready. TV sizes up to 75” will be commoditized by then, but larger panels with modular builds present an opportunity. Companies like Samsung could test the water as early as 2020-2021 with low volume “luxury” models aimed at “mansion” home theatres or high-end retail. Smaller dies, below 5μm, are needed to address consumer markets, which will require at least two more years.
For smartphones, OLED will be a mature, highperformance, cost-effective solution by the time microLED is ready. MicroLED can’t match OLED’s cost. Differentiating performance and features still to be invented are required to compete. Die sizes below 5μm are needed to remain within an acceptable cost bracket, and high volumes require massive investments in the supply chain.
Apple still appears the best positioned to enable high volume microLED smartphones. This could happen 2-3 years after introduction in smartwatches but also raises an existential question for the industry. What happens if Apple pulls the plug on microLEDs?
Aixtron, Aledia, Allos Semiconductor, Apple, Applied Materials, AUO, BOE, CEA-LETI, CIOMP, Columbia University, corning, CSOT, eLux, Epiled, Epistar, Facebook, Foxconn, Galnt Precision Machinery, glō, Goertek, Google, HKUST, IMEC, Instrument System, Intel, ITRI, Jade Bird Display, Kansas State University, KIMM, Konika Minolta, Kookmin U., Kyocera, Lextar, LG, Lumens, Lumileds, Lumiode, LuxVue, Macroblock, Marketech, Microsolar, Mikro Mesa, NCTU, Mojo Vision, Nichia, Nitride Semiconductors, Nth Degree, Oculus, Optovate, Osram, Ostendo, PlayNitride, Plessey, PSI Co, QMAT, Rohinni, Samsung, Sanan, Sapien, Saultech, SelfArray, Semprius, Sharp, Smart Equipment Technology, Seoul Semiconductor, Sharp, Sony, Soitec, Strathclyde University, SUSTech, Sun Yat-sen University, Sxaymiq Technologies, Tesoro Scientific, Texas Tech, Tianma, Topconn, TSMC, Toray, Tyndall National Institute, Uniqarta, U. Of Hong Kong, U. of Illinois, Ultra Display, Veeco, VerLASE, Viewtrix, V-Technology, VueReal, Vuzix, X-Display, Xerox PARC, and more.
Related Reports & Monitors
MACOM NPA1008 RF Power Amplifier with GaN-on-Si HEMT
Reverse Costing - Structural, Process & Cost Report
Automotive Advanced Front-Lighting Systems 2019
Market & Technology
Need to discuss?
We are open for discussionContact us