Corporate AI, Lithography & Hardware R&D Leaderboard + Upstream Foundry Demand
Track verified technical R&D velocity across 51 public technology and semiconductor leaders — spanning frontier AI systems, custom accelerators, computational lithography, EUV optics, and 3D packaging — paired with mid-quarter upstream wafer foundry & AI server rack demand read-through mapped directly to US equities.
Live Corporate AI, Lithography & Process R&D Leaderboard
Public technology, semiconductor, foundry, EDA, and wafer-equipment companies ranked by active technical R&D output across AI compute kernels, optical interconnects, High-NA lithography, and advanced packaging. Click for interactive charts or for company fundamentals.
| Rank | Company & Tools | Lab Output | Active R&D Focus Themes | Latest Technical Disclosure | First-Entry & 365d Velocity | Valuation & Analysis |
|---|---|---|---|---|---|---|
| #1 | AVGO |
180
|
Co-packaged optics (CPO)
|
Active hardware R&D program | Unlock Alert |
$347.01
· P/E 18.4x
|
| #2 | VRT |
129
|
Datacenter liquid cooling
|
Active hardware R&D program | Unlock Alert |
$202.07
· P/E 21.1x
|
| #3 | NVIDIA |
52
51 AI/Sys · 1 Litho/Fab
|
Robot Manipulation
World Models / Embodied AI
CUDA Kernel Engineering
HBM / Memory Stacking
+14 more
|
A Scalable Sharp-Interface Immersed Boundary Method for Moving Boundary Flow Simulation on Multi-GPU Sys…
|
Unlock Alert |
$162.85
· P/E 11.0x
|
| #4 | CAMT |
40
|
Silicon Carbide (Research) & Power Wafer Defect Inspection
|
Active hardware R&D program | Unlock Alert |
$101.73
· P/E 21.2x
|
| #5 | Samsung |
37
11 AI/Sys · 26 Litho/Fab
|
Inference Serving / Speculative Decoding
HBM / Memory Stacking
World Models / Embodied AI
5G mmWave / MIMO
+11 more
|
Overcoming BEOL patterning challenges for the ArF immersion extension
|
Unlock Alert |
|
| #6 | Microsoft |
34
34 AI/Sys
|
LLM Efficiency / Architecture
Inference Serving / Speculative Decoding
Robot Manipulation
Hardware-Agnostic ML Compilers
+8 more
|
Beyond Expert Rankings: Contrast Calibration for Mixed-Precision MoE Quantization
|
Unlock Alert |
$523.88
· P/E 25.8x
|
| #7 | Alibaba |
32
32 AI/Sys
|
KV-Cache / Attention Memory Compression
LLM Efficiency / Architecture
Mixture-of-Experts / Sparse Models
HBM / Memory Stacking
+4 more
|
OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling
|
Unlock Alert |
$94.38
· P/E 16.3x
|
| #8 | Amazon |
29
29 AI/Sys
|
LLM Efficiency / Architecture
Mixture-of-Experts / Sparse Models
World Models / Embodied AI
Inference Serving / Speculative Decoding
+6 more
|
Hybrid Generative-Tagging Decoding for Efficient Product Attribute-Value Extraction
|
Unlock Alert |
$266.66
· P/E 24.6x
|
| #9 | Google/DeepMind |
28
28 AI/Sys
|
LLM Efficiency / Architecture
Mixture-of-Experts / Sparse Models
KV-Cache / Attention Memory Compression
LLM / Agent Security
+10 more
|
Two For The Price Of One: Curvature-Aware Pruning Preserves Influence Functions Too
|
Unlock Alert |
$295.77
· P/E 19.7x
|
| #10 | ASML |
25
1 AI/Sys · 24 Litho/Fab
|
EUV Lithography
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
+5 more
|
Simulation-driven lithography innovation: honoring the legacy of Prof. Andrew R. Neureuther
|
Unlock Alert |
$1495.16
· P/E 23.4x
|
| #11 | IBM |
22
16 AI/Sys · 6 Litho/Fab
|
KV-Cache / Attention Memory Compression
Quantum Error Correction
HBM / Memory Stacking
Hardware-Agnostic ML Compilers
+11 more
|
DropKV: Decoupling Residual-Output Perturbation for Near-Optimal KV-Cache Eviction
|
Unlock Alert |
$247.13
· P/E 18.6x
|
| #12 | Tencent |
21
21 AI/Sys
|
LLM Efficiency / Architecture
World Models / Embodied AI
Low-Precision Inference (FP4/FP8)
Mixture-of-Experts / Sparse Models
+4 more
|
BELT : BEhavior-Aligned Low-Rank KV Cache Compression with Learned Top-k Rank Allocation
|
Unlock Alert |
|
| #13 | BESIY |
19
|
Hybrid bonding (Besi W2W/D2W)
|
Active hardware R&D program | Unlock Alert |
$171.53
· P/E 23.9x
|
| #14 | Intel |
15
5 AI/Sys · 10 Litho/Fab
|
Low-Precision Inference (FP4/FP8)
Edge AI / Inference Chips
KV-Cache / Attention Memory Compression
World Models / Embodied AI
+9 more
|
Advances in multibeam mask writing with MBMW-401
|
Unlock Alert |
$59.66
|
| #15 | Synopsys |
14
14 Litho/Fab
|
Photomask Defect Mitigation
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
EUV Mask Blank & Multilayer Optics
+4 more
|
Production high-NA EUV stitch coloring and physical verification for physical design and mask synthesis
|
Unlock Alert |
$455.20
· P/E 25.5x
|
| #16 | Tokyo Electron |
14
14 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
Computational & Inverse Lithography (ILT)
+1 more
|
EUV quantum yield of organotin molecular resists: a first principles study
|
Unlock Alert |
|
| #17 | Siemens |
14
1 AI/Sys · 13 Litho/Fab
|
LLM / Agent Security
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
Computational & Inverse Lithography (ILT)
+2 more
|
Beyond Manhattan OPC: improving contact array fidelity using curvilinear OPC
|
Unlock Alert |
$154.09
|
| #18 | Apple |
11
11 AI/Sys
|
KV-Cache / Attention Memory Compression
World Models / Embodied AI
Inference Serving / Speculative Decoding
AI Drug Discovery
+2 more
|
MS-ECG-FM: Towards a More Universal Electrocardiogram Foundation Model for Health Monitoring using Multi…
|
Unlock Alert |
$356.27
· P/E 37.2x
|
| #19 | KLA |
9
9 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
Computational & Inverse Lithography (ILT)
+3 more
|
ML-driven context monitoring for inline CDU error analysis
|
Unlock Alert |
$152.32
· P/E 28.1x
|
| #20 | Applied Materials |
9
1 AI/Sys · 8 Litho/Fab
|
ALD / Thin Film Deposition
3D Wafer-on-Wafer & Hybrid Bonding
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
+2 more
|
P20 High-NA enablement through pattern shaping technology: tip-to-tip engineering, roughness, and defect…
|
Unlock Alert |
$370.50
· P/E 22.1x
|
| #21 | SK Hynix |
9
3 AI/Sys · 6 Litho/Fab
|
HBM / Memory Stacking
Custom HBM / Base Die Logic
KV-Cache / Attention Memory Compression
3D Wafer-on-Wafer & Hybrid Bonding
+5 more
|
Die warpage and overlay control using die stress management (DSM)
|
Unlock Alert |
|
| #22 | Meta |
9
9 AI/Sys
|
LLM Efficiency / Architecture
KV-Cache / Attention Memory Compression
Inference Serving / Speculative Decoding
LLM / Agent Security
+3 more
|
Generative Neural Retargeting for Human-to-Robot Dexterous Manipulation
|
Unlock Alert |
$641.69
· P/E 18.2x
|
| #23 | Lam Research |
8
8 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
Next-Gen Photoresist Chemistry
+1 more
|
Application of ALD and isotropic etch to 3D patterning of advanced logic and memory devices
|
Unlock Alert |
$225.92
· P/E 24.4x
|
| #24 | Merck KGaA/EMD |
8
8 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Next-Gen Photoresist Chemistry
|
Negative-tone EUV photoresist system for bridge technology between positive CAR and MOR
|
Unlock Alert |
$33.07
|
| #25 | AMD |
7
7 AI/Sys
|
Hardware-Agnostic ML Compilers
Inference Serving / Speculative Decoding
Edge AI / Inference Chips
KV-Cache / Attention Memory Compression
+2 more
|
Architecting for Hardware Diversity: Meta's End-to-End Co-Design for High-Performance AI Training on AMD…
|
Unlock Alert |
$534.06
· P/E 32.9x
|
| #26 | Micron |
6
6 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
Euv Source
Actinic Inspection & E-Beam Metrology
Computational & Inverse Lithography (ILT)
+3 more
|
ML-driven context monitoring for inline CDU error analysis
|
Unlock Alert |
$778.31
· P/E 4.7x
|
| #27 | Fujifilm |
5
5 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Computational & Inverse Lithography (ILT)
Next-Gen Photoresist Chemistry
|
PFAS-Free 193 immersion resist for 28 nm node metal-1 Layer: electrical performance meets POR quality
|
Unlock Alert |
|
| #28 | Baidu |
5
5 AI/Sys
|
LLM Efficiency / Architecture
Inference Serving / Speculative Decoding
KV-Cache / Attention Memory Compression
LLM / Agent Security
|
FlashMask-3: Efficient and Expressive Mask-Aware Distributed Attention
|
Unlock Alert |
$78.14
|
| #29 | Qualcomm |
5
5 AI/Sys
|
Advanced Packaging / Chiplets
Autonomous Vehicles / LiDAR
LLM / Agent Security
LLM Efficiency / Architecture
+1 more
|
How Do Hardware Faults Manifest Differently in LLM Agents?
|
Unlock Alert |
$172.47
· P/E 17.8x
|
| #30 | Dai Nippon Printing |
4
4 Litho/Fab
|
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
EUV Mask Blank & Multilayer Optics
Computational & Inverse Lithography (ILT)
+2 more
|
High-NA EUV mask CD-SEM metrology matching, and contour-based comparison of simulation result and wafer …
|
Unlock Alert |
|
| #31 | Nokia/Bell Labs |
4
4 AI/Sys
|
Low-Precision Inference (FP4/FP8)
Co-Packaged Optics
|
Capability Scaling-Down Laws for LLM Compression
|
Unlock Alert |
$8.07
· P/E 19.3x
|
| #32 | Cadence |
3
3 Litho/Fab
|
Computational & Inverse Lithography (ILT)
|
Demonstration of an efficient, high-performance repair flow for any-angle post-OPC layouts
|
Unlock Alert |
$323.01
· P/E 33.6x
|
| #33 | Screen Holdings |
3
3 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
High-NA Lithography (0.55 NA)
Actinic Inspection & E-Beam Metrology
Next-Gen Photoresist Chemistry
|
Overcoming pattern collapse in metal oxide resists using a new rinse technology for EUV lithography
|
Unlock Alert |
|
| #34 | Ericsson |
3
3 AI/Sys
|
LLM Efficiency / Architecture
Neuromorphic Computing
World Models / Embodied AI
|
Design Space Exploration of In-Memory Computing Implementations for Discrete Fourier Transform
|
Unlock Alert |
$10.56
· P/E 15.9x
|
| #35 | Infineon |
3
3 AI/Sys
|
Silicon Carbide (Research)
Gallium Nitride (GaN)
|
Watching a GaN Transistor Switch: Real-Time Nanoscale Strain and Heat Dynamics
|
Unlock Alert |
$55.71
|
| #36 | Canon |
2
2 Litho/Fab
|
Next-Gen Photoresist Chemistry
|
Atomic layer etching pitch splitting (APSTM) for nanoimprint lithography: combining simplicity, precisio…
|
Unlock Alert |
|
| #37 | GlobalFoundries |
2
2 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
Computational & Inverse Lithography (ILT)
Co-Packaged & Silicon Photonics
|
Innovating at scale: toward diversified manufacturing for purpose-built chips
|
Unlock Alert |
$41.47
· P/E 16.8x
|
| #38 | TSMC |
2
2 Litho/Fab
|
Atomic Layer Etch & Deposition (ALE/ALD)
Multi-Beam Mask Writing
N2 / N2P & A16 Nanosheet GAA + Backside Power
|
Extendibility of multibeam mask writer MBM-2000
|
Unlock Alert |
$398.73
· P/E 18.7x
|
| #39 | Analog Devices |
2
2 AI/Sys
|
Robot Manipulation
Ultra-Precision Data Converters (ADC/DAC) for WFE, ATE & Surgical Robotics
|
DuoMind: Enabling Distributed Multi-Robot Coordination with Semantic Communication
|
Unlock Alert |
$397.97
· P/E 24.3x
|
| #40 | Sony |
2
2 AI/Sys
|
Humanoid Robots
World Models / Embodied AI
|
LiteNWM: Efficient Latent World Models for Onboard Visual Navigation in the Wild
|
Unlock Alert |
$27.07
· P/E 18.2x
|
| #41 | AGC |
1
1 Litho/Fab
|
EUV Mask Blank & Multilayer Optics
|
Exploring multilayer mirror designs for future EUV lithography
|
Unlock Alert |
$6.20
|
| #42 | Entegris |
1
1 Litho/Fab
|
Next-Gen Photoresist Chemistry
|
A novel purifier for acid-sensitive applications: suppressing dimerization and metal impurities in solve…
|
Unlock Alert |
$147.41
· P/E 29.5x
|
| #43 | Hoya |
1
1 Litho/Fab
|
Actinic Inspection & E-Beam Metrology
EUV Mask Blank & Multilayer Optics
|
Long-term stability of actinic metrology and EUV mask blanks
|
Unlock Alert |
|
| #44 | Mycronic |
1
1 Litho/Fab
|
Multi-Beam Mask Writing
|
The SLX laser mask writer: advancing performance and expanding application space
|
Unlock Alert |
|
| #45 | Onto Innovation |
1
1 Litho/Fab
|
3D Wafer-on-Wafer & Hybrid Bonding
Actinic Inspection & E-Beam Metrology
|
Buried void inspection in high bandwidth memory with picosecond laser acoustics
|
Unlock Alert |
$203.62
· P/E 18.2x
|
| #46 | Photronics |
1
1 Litho/Fab
|
Computational & Inverse Lithography (ILT)
|
Enhancing wafer process window performance with curvilinear ILT correction for DRAM contact layer
|
Unlock Alert |
$27.55
· P/E 13.3x
|
| #47 | Corning |
1
1 AI/Sys
|
Solid State Batteries
|
Local Structure Dictates Ionic Transport and Mechanical Properties in Glassy Solid Electrolytes for Lith…
|
Unlock Alert |
$114.08
· P/E 27.0x
|
| #48 | Boston Dynamics (HYMTF) |
1
1 AI/Sys
|
Humanoid Robots
|
HuMBLE: Human Motion-Driven Behavior Learning for Embodied Locomotion
|
Unlock Alert |
|
| #49 | IonQ |
1
1 AI/Sys
|
Quantum Error Correction
|
Fast logical operations in quantum LDPC codes using simple resource states
|
Unlock Alert |
$28.88
|
| #50 | Marvell |
1
1 AI/Sys
|
KV-Cache / Attention Memory Compression
|
A Photonic-CXL Memory Appliance for Scalable KV Cache Management in LLM Inference
|
Unlock Alert |
$156.76
· P/E 22.7x
|
| #51 | ON |
1
|
EliteSiC M3e 1200V Silicon Carbide MOSFETs & Modules
|
Active hardware R&D program | Unlock Alert |
$71.06
· P/E 15.5x
|
Mid-Quarter Upstream Foundry & AI Server Demand Tracker
Monthly upstream wafer fabrication and high-density AI server rack assembly momentum provides a high-signal mid-quarter read-through on silicon utilization, CoWoS/HBM packaging throughput, and hyperscale data center rack deliveries weeks before US semiconductor and server companies report quarterly results.
Direct US AI Silicon, Foundry Equipment & Server Infrastructure Read-Through (19 Equities)
Tracks real-time mid-quarter leading-edge 3nm/5nm wafer output and 3D advanced packaging utilization ahead of quarterly earnings cycles.
Upstream wafer allocation, interposer packaging throughput, and high-density liquid-cooled AI rack assembly provide a real-time mid-quarter read on Data Center compute shipments.
Data center GPU accelerators and server CPU chiplets scale directly with upstream leading-edge foundry utilization and cloud server integration.
Hyperscale custom AI accelerators and high-radix Ethernet switching silicon track upstream leading-edge wafer and advanced packaging throughput.
Custom mobile, desktop, and cloud silicon nodes ramp first on leading-edge foundry capacity alongside high-volume global hardware assembly.
Flagship mobile, AI PC, and automotive cockpit SoC shipments track upstream leading-edge foundry wafer starts.
800G/1.6T optical interconnect DSPs and custom cloud accelerators depend directly on advanced-node foundry throughput.
High-Bandwidth Memory cube shipments are co-packaged directly alongside AI compute dies and integrated into high-density server racks.
Coolant distribution units (CDUs) and high-density rack power infrastructure co-move with upstream AI server rack assembly velocity.
Enterprise AI server backlog conversion follows upstream accelerator tray availability and high-density rack assembly.
Rack-scale liquid-cooled system shipments track upstream GPU package allocation and server baseboard output.
Back-end AI cluster Ethernet switch deployments scale with upstream networking silicon output and rack-level server installations.
AI server revenue conversion tracks upstream accelerator supply and ODM system integration throughput.
Per-chip architecture royalties scale with leading-edge mobile SoC and custom cloud/AI host CPU wafer shipments.
Sustained upstream leading-edge wafer revenue expansion drives lithography scanner utilization, service upgrades, and next-node tool bookings.
Rising advanced-node wafer starts and larger AI die sizes directly increase optical and e-beam metrology intensity.
Nanosheet transistor transitions and 3D through-silicon-via packaging expansion drive etch and deposition tool demand.
High foundry capacity utilization accelerates interconnect, epitaxy, and advanced packaging equipment pull-ins.