Custom AI chip design

and are nearly even for custom AI silicon design and NPU evaluation.

It depends

’s answer sentences instead focus on custom-chip co-design for infrastructure and AI workloads.

Since Jul 5Updated Nov 14, 2026Based on 63 AI answersReviewed by

How we measure

Brands AI mentions

#BrandMentioned in

Its Maia accelerators target high-throughput AI and data-center workloads.

Its custom-chip work spans infrastructure partnerships and AI processor co-design.

Its MTIA accelerators address internal workloads, including recommendation-model inference.

Mentioned in, on ChatGPT and Google AI Mode

Questions and answers

  • ChatGPTNov 7, 2026

    1. 1Broadcom
    2. 2Marvell
    3. 3Alchip Technologies
    4. 4MediaTek
    5. 5VeriSilicon
    6. 6Socionext
    7. 7Faraday Technology
    8. 8Google
    9. 9Amazon
    10. 10Microsoft
    11. 11MAIA
    12. 12Meta
    13. 13MTIA
    14. 14Apple
    15. 15Qualcomm
    16. 16Huawei
    17. 17Cambricon
    18. 18Arm
    19. 19Synopsys
    20. 20Cadence
    21. 21Siemens EDA
    22. 22CEVA
    23. 23Samsung Foundry
    24. 24ASE Technology
    25. 25Global Unichip Corporation
    26. 26Google Cloud TPU
    27. 27AWS
    28. 28AWS Inferentia
    29. 29Arm Neoverse
    30. 30Taiwan Semiconductor Manufacturing Company (TSMC)

    The custom AI silicon market includes two distinct groups: companies that design complete custom NPUs/AI ASICs and engineering partners that help customers turn an architecture into manufacturable silicon.

    For an NPU evaluation, I would focus first on the second group: ASIC design partners, independent design houses, and silicon IP providers.

    1. Leading custom AI silicon designers

    PRIORITY COMPANIES TO EVALUATE

    CompanyHeadquartersCore capabilities
    BroadcomUSCustom AI ASICs, XPU co-design, high-speed interconnect
    MarvellUSCustom AI accelerators, networking and memory interfaces
    Alchip TechnologiesTaiwanAdvanced-node ASIC design and physical implementation
    Global Unichip (GUC)TaiwanCustom ASIC design, HBM integration, advanced packaging
    MediaTekTaiwanCustom SoCs, AI accelerators and edge AI
    VeriSiliconChinaNPU IP, AI ASIC architecture and turnkey design services
    SocionextJapanCustom SoCs and ASIC development
    Faraday TechnologyTaiwanASIC design services and semiconductor IP

    Broadcom and Marvell are the two particularly important players in hyperscale custom AI silicon. Alchip and GUC are important alternatives for outsourced ASIC engineering.

    2. Companies designing their own AI processors

    These are potential technology competitors, partners, or benchmarks.

    CompanyAI silicon platform
    GoogleTPU
    AmazonTrainium and Inferentia
    MicrosoftMaia
    MetaMTIA
    AppleNeural Engine
    QualcommHexagon NPU
    HuaweiAscend
    CambriconMLU AI processors

    These companies have varying degrees of internal design capability and external design-partner involvement.

    3. Silicon IP and design-tool providers

    These companies may be especially relevant if you are developing an NPU rather than purchasing a finished accelerator.

    CompanyWhat it provides
    ArmCPU architecture, compute IP and Neoverse
    SynopsysDesign IP, RTL tools, verification and implementation
    CadenceEDA tools, simulation, verification and chip design
    Siemens EDAPhysical verification and semiconductor design tools
    CEVADSP and AI processor IP
    VeriSiliconProprietary NPU and AI accelerator IP

    www.verisilicon.comchipsupplychain.orgarmkeil.blob.core.windows.net

    4. Manufacturing partners

    Don't overlook the manufacturing ecosystem:

    • 台積電法說會報佳音 7關鍵重點一次看 - 新唐人亞太電視台

    TSMC

    Leading advanced-node foundry and advanced packaging provider.

    • Samsung, SK Hynix shares leap after OpenAI chip deal

    Samsung Foundry

    Advanced-node manufacturing and packaging.

    • Taiwan ASE plant listed as 'lighthouse factory' by WEF
| Taiwan News | Jan. 16, 2023 13:31

    ASE Technology

    Advanced packaging, assembly and testing.

    5. My recommended shortlist for NPU evaluation

    I would prioritize these six companies:

    CompanyWhy evaluate
    BroadcomLarge-scale AI ASIC co-design
    MarvellCustom accelerator engineering
    AlchipASIC implementation expertise
    GUCAdvanced-node and HBM integration
    VeriSiliconNPU IP and end-to-end design
    SocionextCustom SoC engineering

    An important distinction: Broadcom, Marvell and the design houses are potential engineering partners. Google, Amazon and Microsoft are examples of organizations developing proprietary AI silicon. These are different business models.

    For your evaluation, the next important question is whether you need an NPU architecture developed from scratch, licensed NPU IP integrated into your SoC, or a complete turnkey ASIC design.

    I can also prepare a competitive landscape comparing these companies by NPU architecture, design capabilities, licensing model, process nodes and suitability for custom NPU development.

    Compare the six leading NPU design companies Identify turnkey NPU design partners

    Google AI ModeNov 7, 2026

    1. 1NVIDIA
    2. 2Vera Rubin
    3. 3AMD
    4. 4Instinct MI
    5. 5Intel
    6. 6Google
    7. 7Amazon
    8. 8AWS
    9. 9Microsoft
    10. 10MAIA
    11. 11Meta
    12. 12MTIA
    13. 13Broadcom
    14. 14Marvell
    15. 15Tenstorrent
    16. 16SambaNova
    17. 17Etched
    18. 18d-Matrix
    19. 19Rebellions
    20. 20Samsung
    21. 21Intel Foundry
    22. 22NVIDIA Blackwell
    23. 23AMD Ryzen
    24. 24Intel Core
    25. 25Google Cloud TPU
    26. 26AWS Inferentia
    27. 27Cerebras Systems
    28. 28Taiwan Semiconductor Manufacturing Company (TSMC)

    Custom AI silicon and neural processing unit (NPU) design is split across major merchant vendors, hyperscale cloud giants (often partnering with ASIC specialists), and specialized startups.

    Market analysis from AI Windes and AIMultiple outlines the primary players designing custom AI and NPU silicon:

    1. Merchant AI & NPU Designers (Sell Openly to the Market)

    • Nvidia : Dominates data center and edge AI with its GPU and custom system architectures (such as Hopper, Blackwell, and the Vera Rubin platform).
    • AMD : Designs merchant AI accelerators (like the Instinct MI series) and embeds XDNA-based NPUs into its Ryzen PC processors.
    • Intel : Builds merchant data center accelerators (Gaudi) alongside client-side NPU silicon integrated directly into Core and Core Ultra client processors.

    2. Hyperscalers (Design In-House / Custom Silicon)

    • Google : Designs its own Tensor Processing Units (TPUs) and custom inference/training ASICs for internal workloads.
    • Amazon (AWS) : Designs custom AI chips like Trainium (for training) and Inferentia (for inference).
    • Microsoft : Develops custom silicon projects (such as the Maia accelerator series) to power its Azure cloud and internal AI models.
    • Meta : Builds custom inference accelerators like the MTIA (Meta Training and Inference Accelerator) family.

    3. Custom ASIC Co-Design Partners

    • Broadcom : Acts as the primary silicon co-designer and execution partner for hyperscalers—co-developing Google's TPUs, Meta's MTIA, and custom chips for other major cloud and AI firms.
    • Marvell : Provides custom ASIC design services and high-speed connectivity IP for cloud service providers building proprietary AI silicon.

    4. Independent AI & NPU Startups

    • Tenstorrent : Designs RISC-V based AI and NPU chiplets and hardware scaling solutions.
    • Cerebras : Builds massive wafer-scale processors (WSE) tailored for high-speed AI compute.
    • SambaNova Systems : Focuses on reconfigurable dataflow units (RDUs) for enterprise AI inference and training.
    • Etched, d-Matrix, & Rebellions : Build domain-specific ASICs and rack-scale inference accelerators targeting lower-latency or transformer-specific workloads.

    (Note: The vast majority of these designers do not own fabrication plants; instead, they outsource physical manufacturing almost exclusively to TSMC , with smaller volumes handled by Samsung or Intel Foundry.)

    Are you evaluating NPUs for edge/client devices (like AI PCs or phones) or server/data center workloads , and do you have a specific target performance or power envelope in mind?

  • ChatGPTNov 1, 2026

    1. 1AMD
    2. 2Samsung Foundry
    3. 3GlobalFoundries
    4. 4PyTorch
    5. 5ONNX
    6. 6Xcelium
    7. 7Synopsys
    8. 8Cadence
    9. 9Cadence Innovus
    10. 10Synopsys Fusion Compiler
    11. 11JasperGold
    12. 12VC Formal
    13. 13Siemens Calibre
    14. 14Llvm
    15. 15ESP (Embedded Systems Platform)
    16. 16AMD Vitis
    17. 17Taiwan Semiconductor Manufacturing Company (TSMC)
    18. 18Questa Capital
    19. 19Verified Carbon Standard (VCS)
    20. 20AMD Vivado Design Suite
    21. 21Altera

    How to design, develop, and manufacture custom AI accelerator silicon

    Creating a custom silicon accelerator for specialized AI workloads is a multidisciplinary engineering process that combines AI architecture, computer engineering, semiconductor physics, chip design, software development, and advanced manufacturing.

    The goal is to build a processor whose hardware is specifically optimized for a particular class of AI computations, rather than relying on a general-purpose CPU or GPU.

    For example, imagine you want to develop a custom AI chip for:

    • Large language model (LLM) inference.
    • Transformer-based neural networks.
    • Computer vision and image recognition.
    • Autonomous robotics.
    • Speech recognition and real-time audio processing.
    • Scientific AI and simulation.
    • Edge AI devices with extremely low power consumption.

    The process is broadly similar across these applications, although the architecture, memory system, precision requirements, and manufacturing choices can differ substantially.

    • Meta: Our next-generation Meta Training and Inference Accelerator | SemiWiki
    • Il silicio è il nuovo petrolio? Chi potrebbe vincere davvero la corsa all'intelligenza artificiale
    • AMD’s Next GPU Is a 3D-Integrated Superchip - IEEE Spectrum
    • Inside Amazon's AI Chip Architecture
    • FuriosaAI’s RNGD at Hot Chips 2024: Accelerating AI with a More Flexible Primitive
    • Kurnal (@Kurnalsalts) on X
    • Q&A: How can advanced chip packaging help redesign the future of semiconductors? | Penn State University
    • More Efficient AI Training With A Systolic Neural CPU
    • How Qualcomm Can Compete With Nvidia For Datacenter AI Inference
    • Rebellions AI Puts Together An HBM And Arm Alliance To Take On Nvidia
    • AI Chip Startup Puts Inference Cards on the Table
    • Why Is Japan Still Investing In Custom Floating Point Accelerators?

    I'll walk through the process from the initial idea through architecture design, RTL development, verification, fabrication, packaging, software, and commercial production.


    1. Define the AI workload before designing the chip

    (Starting point)

    The biggest mistake in custom silicon development is starting with the physical chip design before understanding the computational workload.

    You first need to identify exactly what the accelerator will do.

    Consider three possible products:

    AcceleratorPrimary workloadImportant design priorities
    LLM inference ASICTransformer inferenceMemory bandwidth, matrix multiplication, latency
    Computer vision ASICCNNs and vision transformersParallel processing, image throughput, power
    Edge AI ASICSmall neural networksLow power, small die, low cost

    You should characterize the AI models you intend to support.

    For example, with an LLM accelerator, you would examine:

    • Model parameter count.
    • Number of transformer layers.
    • Attention head dimensions.
    • Sequence length.
    • Batch size.
    • KV-cache requirements.
    • Weight storage requirements.
    • Quantization requirements.
    • Target tokens per second.
    • Maximum acceptable inference latency.

    Translate AI operations into hardware requirements

    Neural networks perform large numbers of mathematical operations.

    Some of the most important are:

    • Matrix multiplication.
    • Vector multiplication.
    • Convolution.
    • Attention calculations.
    • Normalization.
    • Activation functions.
    • Reduction operations.

    For example, matrix multiplication:

    math
    C=A\times B

    is one of the fundamental operations in modern AI.

    A specialized accelerator can implement thousands of multiplication-and-accumulation operations in parallel.

    You should establish measurable design targets:

    MetricExample target
    Performance100 TOPS
    Power consumption50 W
    Memory bandwidth1 TB/s
    Numerical precisionINT8 / BF16
    Manufacturing process5 nm-class
    ApplicationLLM inference

    These are illustrative targets, not specifications for a particular chip.

    Deliverable: A workload specification describing what the chip must compute, how quickly it must compute it, how much energy it can consume, and what it will cost.

    2. Select the accelerator architecture

    This is where you decide what kind of processor you are actually building.

    • SAPER-AI Accelerator: A Systolic Array-Based Power-Efficient Reconfigurable AI Accelerator | Newswise
    • Neural Processing Unit (NPU) Explained - Utmel
    • Reimplementing FlashAttention for performance and giggles | AmineDiro
    • How NPU Works: The AI Chip Engine Explained | Soon Ee Ong posted on the topic | LinkedIn
    • How a GPU Actually Works (I Wrote Kernels to Prove It) — chaoticengineer.dev
    • Sparse Systolic Tensor Array for Efficient CNN Hardware Acceleration | DeepAI
    • 구글, 엔비디아, 딥시크의 최신 동향: HBM을 넘어서
    • CSDL | IEEE Computer Society

    There are several major architectural approaches.

    A. Systolic array architecture

    A systolic array uses a grid of processing elements that exchange data in a coordinated pattern.

    Each processing element performs arithmetic operations, typically multiply-accumulate operations.

    Advantages:

    • Excellent matrix multiplication efficiency.
    • High computational density.
    • Predictable data movement.
    • Good energy efficiency for regular workloads.

    This is particularly useful for transformer and neural-network matrix operations.

    B. Vector processing architecture

    Vector processors operate on multiple data elements simultaneously.

    They are useful for:

    • Element-wise operations.
    • Activation functions.
    • Normalization.
    • Reductions.
    • General tensor operations.

    C. Spatial dataflow architecture

    A spatial architecture distributes computations across multiple specialized hardware units.

    Instead of repeatedly moving data between a centralized processor and memory, the design attempts to keep data close to where computation occurs.

    This can significantly reduce energy consumption.

    D. Hybrid architecture

    For many specialized AI accelerators, a hybrid architecture makes the most sense.

    For example:

    Illustrative AI accelerator architecture

    Host interface / command processor

    PCIe or another system interface

    Matrix engines

    Tensor operations

    Vector engines

    Element-wise operations

    On-chip SRAM

    Fast local storage

    Memory controller

    External memory

    High-bandwidth memory / external memory

    Conceptual architecture; actual implementations vary.

    The architecture should be driven by the workload rather than by a desire to include as many processing units as possible.

    3. Design the memory subsystem

    This is one of the most important parts of an AI accelerator.

    In many AI workloads, the bottleneck isn't the speed of the arithmetic units. It's getting data to those units quickly enough.

    For example, an LLM might have enough computational capacity to generate tokens quickly, but its performance can be limited by moving model weights and KV-cache data through memory.

    You need to design several levels of memory.

    • GPU MODE Lecture 8: CUDA Performance Checklist – Christian Mills
    Memory typePurposeCharacteristics
    RegistersImmediate calculationsExtremely fast, very small
    SRAMLocal tensor storageFast, on-chip
    L1/L2 buffersReusable dataLow-latency access
    HBMLarge model dataVery high bandwidth
    External DRAMAdditional capacityLarger but slower

    Important design decisions

    On-chip SRAM: Determines how much data can remain close to the processing engines.

    HBM: High-bandwidth memory is especially useful for data-center AI accelerators.

    Memory controllers: Coordinate data movement between processing units and external memory.

    Interconnect: Connects compute engines, memory controllers, and other subsystems.

    Dataflow: Determines how tensors move through the processor.

    For LLM inference, you must also consider:

    • Weight streaming.
    • KV-cache storage.
    • Attention data movement.
    • Prefetching.
    • Memory allocation.
    • Multiple concurrent inference requests.

    A good accelerator architecture balances compute throughput with memory bandwidth and capacity.

    4. Select numerical precision and implement AI-specific compute engines

    Modern AI chips don't necessarily need to perform every operation using conventional 32-bit floating-point arithmetic.

    Different AI workloads benefit from different numerical formats.

    FormatTypical application
    FP32High-precision computation
    BF16AI training and inference
    FP16Neural-network computation
    FP8Efficient large-model computation
    INT8Quantized inference
    INT4Highly compressed inference

    For example, an LLM inference accelerator might use:

    • INT4 for quantized model weights.
    • INT8 for selected intermediate computations.
    • BF16 for operations requiring greater numerical range or precision.

    The design team needs to establish acceptable numerical error and validate model accuracy under each supported format.

    Develop specialized compute units

    The main compute blocks might include:

    1. Matrix multiplication engines.
    2. Vector arithmetic units.
    3. Attention processing units.
    4. Normalization units.
    5. Activation-function units.
    6. Tensor movement engines.
    7. Reduction and accumulation units.

    The objective is to implement frequently used AI operations efficiently in hardware.

    An important architectural consideration is programmability: you don't want a chip that performs one particular model extremely well but cannot accommodate future models or software updates.

    5. Develop the actual digital hardware (RTL design)

    (Major engineering milestone)

    Now you translate the architecture into hardware descriptions.

    The primary technologies include:

    • Verilog.
    • SystemVerilog.
    • VHDL.
    • Chisel.
    • SystemC.
    • High-Level Synthesis (HLS).

    RTL stands for Register Transfer Level.

    It describes how digital hardware stores, transfers, and processes information over clock cycles.

    For example, a simplified multiply-accumulate operation is:

    math
    ACC_{next}=ACC+(A\times B)

    A real accelerator contains many such operations operating in parallel.

    What engineers build

    The RTL design is divided into manageable hardware modules.

    ModuleResponsibility
    Compute enginePerforms tensor operations
    SchedulerDistributes operations
    DMA engineMoves data
    Memory controllerCoordinates memory
    InterconnectConnects hardware blocks
    Command processorExecutes instructions
    Security controllerProtects system resources
    PCIe interfaceConnects accelerator to host

    Engineers also integrate existing intellectual property where appropriate, such as processor cores, memory interfaces, PCIe controllers, and standard communication interfaces.

    You generally don't want to reinvent every component.

    Recommended development approach

    Develop the architecture in progressively more detailed models:

    1. Python-based performance model.
    2. Cycle-accurate architectural simulator.
    3. Hardware-oriented C++ or SystemC model.
    4. RTL implementation.
    5. FPGA prototype.

    This allows you to discover architectural mistakes before committing to expensive custom silicon.

    A useful educational resource is Columbia University's ESP platform. Its documentation covers accelerator RTL development, integration, and high-level synthesis.

    6. Verify and validate the hardware

    This is one of the most expensive and time-consuming parts of semiconductor development.

    You must establish that the design actually works before manufacturing it.

    A silicon defect can cost millions of dollars to correct.

    A. Functional verification

    Engineers simulate the hardware and verify that it behaves according to its specifications.

    They test:

    • Arithmetic correctness.
    • Memory access.
    • Data movement.
    • Instruction execution.
    • Clock behavior.
    • Reset behavior.
    • Error handling.
    • Multiple concurrent operations.

    B. Formal verification

    Formal verification uses mathematical techniques to establish that specified hardware properties hold across defined conditions.

    It complements simulation, rather than replacing it.

    C. Performance verification

    You must establish that the chip can achieve its intended:

    • Clock frequency.
    • Compute throughput.
    • Memory bandwidth.
    • Latency.
    • Power consumption.

    D. FPGA prototyping

    • Alveo—An FPGA Board for the Cloud and Data Center | Electronic Design

    Before producing custom silicon, you can implement parts of the architecture on an FPGA.

    An FPGA allows you to test real hardware behavior without commissioning a custom semiconductor fabrication run.

    You can evaluate:

    • Data movement.
    • Hardware/software integration.
    • Command execution.
    • Driver functionality.
    • Model execution.
    • Architectural assumptions.

    An FPGA will not necessarily reproduce the speed, power consumption, or memory characteristics of the eventual ASIC.

    However, it is extremely useful for validating functionality.

    AMD's Vitis documentation provides a practical example of the RTL-to-FPGA workflow, including simulation, synthesis, implementation, timing analysis, and application integration.

    7. Synthesize the design and perform physical implementation

    (From logical design to physical silicon)

    Once the RTL is sufficiently verified, the design moves into physical implementation.

    This is where your logical hardware becomes a physical arrangement of transistors and wires.

    • From RTL to GDSII: Demystifying the ASIC Design Flow
    • Floorplanning
|
vlsi4freshers
    • Tutorial 7: ASIC Back-End Flow - ECE 6745 Complex Digital ASIC Design
    • Need to Fit Billions of Transistors on a Chip? Let AI Do It | WIRED
    • GitHub - suhasshirol/Advanced-Physical-Design-Using-OpenLANE-SKY130 · GitHub
    • PPT - ASIC Front-End Design PowerPoint Presentation, free download - ID:3955086
    • Printing my first ASIC's GDS files | Zero to ASIC Course
    • RTL Synthesis — Advanced Digital Systems Design Fall 2024  documentation
    • ZeptoBars
    • Synthesis RTL- Gate level Netlist | by Medha Kadam | Medium

    There are several major steps.

    A. Logic synthesis

    EDA software converts RTL into a gate-level netlist.

    This involves mapping the design into actual logic cells available in the selected semiconductor manufacturing process.

    The result is a representation of the hardware using gates, flip-flops, and other standard cells.

    B. Floorplanning

    Engineers determine where major blocks should physically reside on the chip.

    For an AI accelerator, this includes:

    • Matrix engines.
    • SRAM blocks.
    • Memory controllers.
    • Interconnect.
    • PCIe interfaces.
    • Clock and power infrastructure.

    The objective is to optimize physical area, power consumption, timing, and routing.

    C. Placement and routing

    The EDA software determines the physical locations of standard cells and creates the interconnections between them.

    This is a difficult optimization problem because physical distance affects signal delay, power, congestion, and heat.

    D. Timing analysis

    Static timing analysis checks whether signals can propagate through the design within the required clock periods.

    If the design cannot meet its timing requirements, engineers may need to modify the architecture, placement, or implementation.

    E. Physical verification

    Before fabrication, engineers conduct important sign-off checks, including:

    • Design Rule Checking (DRC).
    • Layout Versus Schematic (LVS).
    • Static timing analysis.
    • Power integrity analysis.
    • Electromigration analysis.
    • Signal integrity analysis.
    • Thermal analysis.

    A typical digital ASIC flow proceeds from RTL through synthesis, physical implementation, verification, and tapeout. CERN publishes examples of technology-specific RTL-to-GDS implementation flows.

    8. Choose a semiconductor foundry and manufacturing process

    Now you need to determine where and how your accelerator will be manufactured.

    Most new AI hardware companies operate using a fabless model.

    This means they design the chip but contract an external semiconductor foundry to manufacture it.

    Some major manufacturing providers include:

    • TSMC projects sustained AI-driven growth amid capacity constraints and expanding global footprint | Noah Intelligence

    TSMC

    A major advanced semiconductor foundry serving fabless chip companies.

    Official website


    • Potential Dependency Oversight in U.S.-South Korea Chip Policy | Council on Foreign Relations

    Samsung Foundry

    Offers semiconductor manufacturing and advanced packaging capabilities.

    Official website


    • Semiconductor manufacturing | GlobalFoundries

    GlobalFoundries

    Provides specialty and mature-node semiconductor manufacturing.

    Official website

    Choosing the process node

    You might encounter manufacturing processes described as:

    • 180 nm.
    • 65 nm.
    • 28 nm.
    • 16/14 nm.
    • 7 nm.
    • 5 nm.
    • 3 nm.

    Smaller process nodes can offer advantages in transistor density and potentially power and performance, but they also introduce greater design complexity and cost.

    Importantly, a smaller node isn't automatically better for every accelerator.

    For instance:

    • A small edge inference chip might be economically attractive on a mature node.
    • A high-performance data-center accelerator may justify an advanced process.
    • A memory-intensive design may benefit more from packaging and memory architecture improvements than from a smaller transistor node.

    The foundry design process

    You need access to a Process Design Kit (PDK).

    A PDK contains the technology-specific information necessary to design and verify a chip for a particular manufacturing process.

    It includes information about:

    • Standard-cell libraries.
    • Transistor models.
    • Design rules.
    • Manufacturing constraints.
    • Physical layout requirements.
    • Electrical characteristics.

    You must design within the foundry's manufacturing rules.

    The foundry and chip-design team work together to ensure that the proposed design can be manufactured reliably.

    9. Tapeout: Send the completed design for fabrication

    • GPT Image 2 vs Nano Banana Pro: 10 Fields Tested

    Tapeout is the formal milestone at which the finalized chip layout is delivered for manufacturing.

    The main physical design deliverable is typically a GDSII or OASIS layout database, accompanied by the required manufacturing and sign-off data.

    The foundry uses this design to prepare the manufacturing process.

    For a conventional monolithic digital ASIC, the manufacturing process broadly involves:

    1. Preparing silicon wafers.
    2. Depositing and processing material layers.
    3. Patterning structures using photolithography.
    4. Etching and doping selected regions.
    5. Building transistor structures.
    6. Creating metal interconnections.
    7. Performing wafer-level inspection and testing.

    The exact process depends on the foundry and manufacturing technology.

    Why tapeout is such a major milestone

    Before tapeout, you can modify the digital design relatively easily.

    After tapeout, correcting a fundamental silicon design defect may require another fabrication run.

    That can mean significant additional engineering expense and schedule delays.

    This is why extensive pre-silicon verification is essential.

    10. Package the silicon and integrate high-bandwidth memory

    • #vlsi #advancedpackaging #2_5dpackaging #semiconductor #chippackaging #interposer #hbm #heterogeneousintegration #aihardware #hpc #datacenterchips #gpu #asic #semiconductorindustry #vlsilearning… | Chaithanya Kuncha
    • 高带宽内存的下一步是什么 - Embedded AI Data Center
    • AMD GPUOpen Developer Portal - AMD GPUOpen
    • What's all the noise in the AI basement?
    • 첨단 이종 집적화 | Foundry | 삼성반도체
    • 【大原雄介の半導体業界こぼれ話】同じ「チップレット」でも、明暗が分かれたIntelとAMD  - PC Watch
    • AI Pushes High-End Mobile From SoCs To Multi-Die
    • Atalla Institute for Advanced System Integration and Packaging (ASIP) - Purdue University
    • Imec pushes 3D HBM-on-GPU technology to extreme limits
    • Why is HBM so Hard to Manufacture? - by Vikram Sekar
    • Beyond CoWoS: What Are Your Real Advanced Packaging Options for AI Chips in Late 2026? - SupplyICs

    For modern AI accelerators, manufacturing the silicon die is only part of the process.

    You also need to package it.

    Packaging provides:

    • Electrical connections.
    • Physical protection.
    • Thermal management.
    • Power delivery.
    • External interfaces.

    For high-performance AI chips, advanced packaging is particularly important.

    Common packaging approaches

    PackagingGeneral use
    Conventional 2DStandard packaged chips
    2.5DMultiple dies connected through an interposer or similar substrate
    3D stackingDies vertically stacked
    ChipletsMultiple specialized dies integrated into one package

    Why chiplets are interesting

    Instead of manufacturing one enormous monolithic processor, you can divide functionality among multiple dies.

    For example:

    • Compute die.
    • I/O die.
    • Memory interface die.
    • Specialized accelerator die.

    These can be integrated into a single package.

    The Universal Chiplet Interconnect Express (UCIe) standard is designed to provide standardized die-to-die connectivity and facilitate chiplet-based systems.

    For an AI accelerator, a chiplet strategy can offer greater flexibility in combining compute, memory interfaces, and different manufacturing processes.

    However, it adds packaging, interconnect, testing, and integration complexity.

    11. Develop the software stack

    (Essential for commercial success)

    A custom AI accelerator is not a complete product just because the silicon works.

    You need software that allows AI developers to actually use it.

    This is one of the most underestimated aspects of building an AI chip.

    A chip with excellent theoretical performance can struggle commercially if developers cannot easily run their models on it.

    Your software stack should include several layers.

    AI accelerator software stack

    AI applications

    LLMs, vision, speech, robotics

    AI frameworks

    PyTorch, ONNX and supported model interfaces

    Model compiler and optimizer

    Graph optimization, quantization, operator mapping

    Runtime and libraries

    Scheduling, memory management, kernel libraries

    Device drivers

    Hardware communication and resource management

    Custom AI accelerator silicon

    The major software components

    1. Device drivers

    These allow the operating system to communicate with the accelerator.

    2. Runtime system

    Manages hardware resources, memory allocation, execution queues, and synchronization.

    3. Compiler

    Converts AI model operations into instructions and kernels supported by your hardware.

    4. Kernel libraries

    Provide optimized implementations of matrix multiplication, attention, normalization, and other operations.

    5. Framework integration

    Enables developers to use your accelerator through familiar machine-learning frameworks and APIs.

    6. Model optimization tools

    Support quantization, graph transformations, operator fusion, and model compilation.

    A particularly important decision: Build your own compiler or adapt an existing ecosystem?

    For a first-generation accelerator, building an entirely new AI programming ecosystem is generally an expensive mistake.

    A better strategy is often to support established model representations and frameworks, while developing a compiler backend and hardware-specific runtime.

    For example, you could support ONNX model import and PyTorch-based workflows.

    The software architecture should be planned alongside the hardware architecture, not after the silicon is completed.

    12. First silicon bring-up and post-silicon validation

    • Turnkey ASIC Supply Chain Management & Long-Term Availability
    • 系统级测试 (SLT) - Teradyne
    • ADVANTEST T 5581 P Final Testing Equipment used for sale price #293643267 > buy from CAE
    • FormFactor - Cascade CM300xi-SiPh - 300 mm semi-/ fully-automated probe system with Autonomous Silicon Photonics Measurement Assistant | PSI Solutions, Inc.
    • integrated circuit testing | SEMI
    • AI Drives Higher Demands For Chip Testing
    • Fraunhofer IIS, division EAS Dresden - Saxony Trade & Invest
    • Advantest expands device handler offerings -  News

    Once the manufactured chips return, you begin testing the actual physical silicon.

    This is called post-silicon validation.

    The first objective is to establish that the chip powers on and operates correctly.

    Initial bring-up

    Engineers typically work through:

    1. Inspecting the packaged devices.
    2. Installing the chip onto a test board.
    3. Checking power rails and electrical behavior.
    4. Establishing clock and reset functionality.
    5. Configuring the device.
    6. Testing host communication.
    7. Running basic memory tests.
    8. Executing simple compute operations.
    9. Running complete AI workloads.

    What you measure

    MeasurementWhat it tells you
    Clock frequencyActual operating speed
    Power consumptionEnergy requirements
    TemperatureThermal behavior
    Memory bandwidthData movement capability
    Compute throughputActual AI performance
    LatencyResponse time
    Model accuracyNumerical correctness
    StabilityReliability under sustained operation

    You should compare actual silicon performance with your pre-silicon simulations.

    Sometimes the first silicon revision reveals unexpected limitations:

    • Timing problems.
    • Memory bottlenecks.
    • Excessive power consumption.
    • Thermal issues.
    • Software compatibility problems.
    • Manufacturing defects.

    Some problems can be addressed through firmware or software. Others require a hardware revision.

    13. Manufacturing test, yield optimization, and mass production

    A working prototype is not necessarily a commercially manufacturable product.

    You need to establish that the chip can be manufactured consistently and economically.

    A. Wafer testing

    Before individual dies are packaged, manufacturers perform electrical testing to identify defective devices.

    B. Dicing

    The wafer is separated into individual dies.

    C. Packaging

    The dies are assembled into their final packages.

    D. Final testing

    The packaged devices undergo electrical, functional, and performance testing.

    E. Reliability testing

    Depending on the application, qualification can include:

    • Temperature cycling.
    • High-temperature operating life.
    • Thermal stress.
    • Electrical stress.
    • Package reliability.
    • Long-duration operation.

    F. Yield optimization

    Yield represents the proportion of manufactured devices that meet the required specifications.

    For example, if a wafer produces 1,000 dies and 800 meet the required criteria, the initial yield is 80%.

    Yield has a major influence on chip economics.

    You must work with the foundry and packaging partners to improve manufacturing consistency, identify defects, and establish acceptable production specifications.


    14. Build the actual AI accelerator product

    The final chip is typically only one component of the product.

    For a data-center accelerator, you might need:

    • AMD Instinct MI350P: ultra-fast AI accelerator as a PCI Express card | heise online
    • Microsoft представила ИИ-ускоритель Maia 200 с 216 Гбайт HBM3e
    • Atlas 200I A2 加速模块
    • SYSONE
    • Tenstorrent will disable 20 cores on already sold Blackhole p150 accelerators.
    • Metis M.2 AI Inference Acceleration Card | Axelera AI
    • Product - Alpha Data
    • Systolic Arrays and the TPU
    • SiMa.ai and Enclustra to push boundaries of physical AI
    • Varium Compute Accelerator Card - V-C1100-P00G-PQ-G
    • Intel Vision 2024: Gaudi 3 soll NVIDIAs H100 und H200 starke Konkurrenz machen - Hardwareluxx
    • Custom accelerator silicon.
    • HBM or external memory.
    • Power delivery circuitry.
    • PCB design.
    • Cooling system.
    • PCIe or other host interface.
    • Firmware.
    • Operating-system drivers.
    • AI runtime.
    • Developer tools.

    You then need to integrate everything into a production-ready board or server.

    For an edge AI accelerator, the final product might instead be a compact module integrated into a camera, robot, industrial controller, or embedded computer.

    15. Understand the economics of producing custom silicon

    This is where the engineering project becomes a semiconductor business.

    Custom AI silicon has substantial upfront costs.

    Your expenses can include:

    ExpenseWhat it covers
    ArchitectureSystem design and modeling
    EngineeringRTL, verification and physical design
    EDA licensingCommercial chip-design software
    IP licensingProcessor, interface and memory IP
    PrototypingFPGA platforms and validation
    MasksManufacturing photomasks
    Wafer fabricationManufacturing silicon
    PackagingAssembly and advanced packaging
    TestingValidation and manufacturing tests
    SoftwareDrivers, compilers and runtime
    ProductionBoards, systems and supply chain

    Three possible business models

    Model A: Fabless AI chip company

    You design and own the accelerator and outsource fabrication.

    This is the traditional approach used by many semiconductor companies.

    Model B: Custom accelerator IP company

    You develop reusable accelerator intellectual property and license it to other companies.

    This can reduce the need to manufacture and sell complete physical products.

    Model C: Vertical AI hardware company

    You design the chip, manufacture the accelerator through external partners, build the software stack, and sell complete AI systems or cloud inference services.

    This allows you to control more of the customer experience.

    The importance of cost per inference

    For a commercial AI accelerator, theoretical TOPS is not enough.

    You need to understand the actual economics of running AI models.

    Useful metrics include:

    math
    \text{Energy per token}=
    \frac{\text{Energy consumed}}{\text{Tokens generated}}
    math
    \text{Cost per token}=
    \frac{\text{Total operating cost}}{\text{Tokens generated}}

    For an LLM inference chip, these can be more commercially meaningful than peak arithmetic throughput.

    A chip that delivers fewer theoretical operations per second but produces tokens more efficiently could be a better commercial product.


    16. How long does the entire process take?

    The development schedule depends on architecture complexity, manufacturing technology, team experience, and whether you reuse existing IP.

    A reasonable planning estimate for a first commercial ASIC is:

    Illustrative development timeline

    Example planning schedule, not a guaranteed industry benchmark.

    Several stages overlap. Verification, software development and testing continue throughout the project.

    A sophisticated high-performance accelerator can take considerably longer, especially if it requires advanced packaging, new memory technology, or multiple silicon revisions.

    17. What tools and technologies would you need?

    Here is a practical overview of the engineering ecosystem.

    CategoryExamples
    Hardware descriptionSystemVerilog, Verilog, Chisel
    Architecture modelingPython, C++, SystemC
    AI developmentPyTorch, ONNX
    RTL simulationQuesta, VCS, Xcelium
    Logic synthesisSynopsys Design Compiler, Cadence Genus
    Physical designCadence Innovus, Synopsys Fusion Compiler
    Formal verificationJasperGold, VC Formal
    FPGA prototypingAMD Vivado, Intel Quartus
    Physical verificationSiemens Calibre
    Chip layoutGDSII, OASIS
    Software developmentC++, Python, LLVM
    Model compilationMLIR-based compiler infrastructure

    These are examples rather than a mandatory toolchain. Commercial ASIC development generally requires foundry-compatible EDA tools, licensed IP, and appropriate technology access.

    For learning, three particularly useful resources are:

    1. CERN ASIC Support — Practical RTL-to-GDS implementation flows.
    2. Columbia ESP documentation — Open-source SoC and accelerator integration.
    3. AMD hardware development documentation — FPGA prototyping and accelerator development.

    For additional background, AWS Semiconductor Design whitepaper describes the front-end and back-end chip-design process.


    18. My recommendation: How I would approach building a specialized AI accelerator from scratch

    I would not begin by trying to create a massive GPU competitor.

    I would start with a narrowly defined AI workload and develop an accelerator that delivers a meaningful improvement over existing hardware.

    Here is the approach I would take.

    Suggested first-generation AI accelerator

    (Practical starting concept)

    • 韩国 FuriosaAI 今年量产 2 万颗 RNGD 芯片,计划云与端侧双轨并行 - 产业链 - 光通信Pro

    Product concept: Specialized LLM inference accelerator

    Target a specific category of transformer-based inference rather than trying to support every AI workload.

    Initial architecture goals

    • Matrix multiplication acceleration.
    • INT8 and INT4 inference support.
    • BF16 support where necessary.
    • On-chip SRAM.
    • External memory or HBM integration, depending on the product.
    • Dedicated data-movement engines.
    • Standard host connectivity.
    • Programmable execution and scheduling.

    Software goals

    • PyTorch model integration.
    • ONNX support.
    • Model quantization.
    • Custom inference runtime.
    • Optimized attention and matrix multiplication kernels.

    A staged development strategy

    Phase 1 — Prove the concept

    Build a Python performance simulator and benchmark actual AI models. Establish whether the proposed architecture has a meaningful advantage.

    Phase 2 — Build a hardware prototype

    Implement the important compute engines on an FPGA. Validate the hardware/software interface and test real model workloads.

    Phase 3 — Develop the ASIC

    Create RTL, verify the design, perform synthesis and physical implementation, and prepare for fabrication.

    Phase 4 — Manufacture and validate

    Work with a foundry and packaging provider to manufacture the chip, bring up the first silicon, and validate its performance.

    Phase 5 — Commercialize

    Build accelerator boards or modules, release the compiler and runtime, establish production, and sell to customers.

    The critical principle is to validate the workload, architecture, and software before spending heavily on custom silicon.


    19. The most important strategic question

    There are several very different ways to approach this industry, and they require different levels of investment and expertise.

    For example, designing a small edge AI accelerator is fundamentally different from developing a high-performance HBM-equipped data-center inference processor.

    I'd suggest deciding which direction you want to pursue first:

    1. What type of AI accelerator interests you most?

    LLM and generative AI accelerators

    Computer vision accelerators

    Low-power edge AI accelerators

    Robotics and autonomous systems accelerators

    General-purpose AI accelerator


    2. What is your primary objective?

    Build a commercial semiconductor product/company

    Develop a research prototype

    Learn the complete chip-design process

    Develop and license proprietary accelerator IP


    3. How would you like to approach development?

    Design everything from the ground up

    Start with open-source hardware and modify it

    Combine existing IP with custom silicon

    Start with an FPGA prototype and transition to ASIC

    [Develop my accelerator blueprint]

    With those decisions, I can take this beyond the general semiconductor lifecycle and develop a concrete engineering blueprint, including the actual chip architecture, compute units, memory subsystem, software stack, manufacturing strategy, development team, and estimated investment required.

    Google AI ModeNov 1, 2026

    1. 1Synopsys ZeBu
    2. 2Synopsys
    3. 3Cadence Palladium
    4. 4Cadence
    5. 5Intel Foundry
    6. 6PyTorch
    7. 7Taiwan Semiconductor Manufacturing Company (TSMC)
    8. 8Google Gemini

    Creating and producing a custom silicon accelerator—an Application-Specific Integrated Circuit (ASIC)—optimized for specialized AI workloads requires a multi-year, highly specialized engineering pipeline.

    The end-to-end journey from a clean sheet to physical, production-ready AI chips is broken down into sequential phases below:

    1. Workload Analysis & Architectural Definition
      • Target Identification: Analyze specific AI algorithms (e.g., LLM inference, diffusion models, sparse transformers) to pinpoint bottlenecks like memory bandwidth, matrix multiplication throughput (M A C cap M cap A cap C𝑀𝐴𝐶 operations), or specific quantization formats (F P 8 cap F cap P 8𝐹𝑃8, I N T 4 cap I cap N cap T 4𝐼𝑁𝑇4).
      • Architectural Specification: Define the core compute topology, such as systolic arrays, tensor cores, on-chip SRAM hierarchy, and interconnect networks. Set strict targets for power envelope (Thermal Design Power, or T D P cap T cap D cap P𝑇𝐷𝑃), die area, and target performance metrics (TOPS or TFLOPs per watt).
    2. RTL Design & Microarchitecture Implementation
      • Register-Transfer Level (RTL) Coding: Translate the architecture into hardware description languages (HDLs) like SystemVerilog or Chisel.
      • IP Integration: Integrate specialized third-party or internal intellectual property (IP) blocks, including high-speed interface controllers like PCIe Gen6, CXL, and high-bandwidth memory controllers (HBM3e/HBM4).
    3. Functional Verification & Simulation
      • Testbench Development: Build comprehensive verification environments using Universal Verification Methodology (UVM).
      • Software-Hardware Co-Simulation: Emulate the RTL design on enterprise hardware emulation systems (such as Synopsys ZeBu or Cadence Palladium ) running actual AI workloads and software stacks before any silicon is cut.
    4. Logical Synthesis & Design for Test (DFT)
      • Logic Synthesis: Use EDA tools (from vendors like Synopsys or Cadence ) to convert RTL code into a gate-level netlist mapped to a specific foundry's standard cell library.
      • DFT Insertion: Embed test structures like boundary scan chains and Built-In Self-Test (BIST) engines to enable post-manufacturing fault screening.
    5. Physical Design & Timing Closure (RTL-to-GDSII)
      • Floorplanning & Placement: Map out macro blocks, memory arrays, and IO pads on the physical die, managing power distribution networks (IR-drop analysis).
      • Clock Tree Synthesis (CTS): Route clock signals uniformly across the die to minimize skew.
      • Routing & Signoff: Connect all logic gates while satisfying strict Static Timing Analysis (STA), Design Rule Checking (DRC), and Layout Versus Schematic (LVS) verification. The final output is the GDSII/OASIS layout database.
    6. Tape-Out & Foundry Manufacturing
      • Tape-Out: Deliver the final GDSII database to an advanced semiconductor foundry (such as TSMC or Intel Foundry).
      • Wafer Fabrication: The foundry processes multi-patterning extreme ultraviolet (EUV) lithography across months, building dozens of metal layers on silicon wafers.
    7. Advanced Packaging & Chiplet Integration
      • 2.5D/3D Packaging: For high-performance AI accelerators, mount the compute die alongside HBM stacks onto a silicon interposer using advanced packaging technologies (like TSMC CoWoS).
    8. Post-Silicon Validation & Bring-Up
      • Laboratory Bring-Up: Power on the first physical chips on custom evaluation boards using logic analyzers and oscilloscopes.
      • Software Stack Tuning: Calibrate firmware, low-level drivers, and runtime compilers (integrating with PyTorch/TensorFlow backends) to achieve functional stability and peak performance.

    If you want to dig deeper, let me know if you would like to explore:

    The economics and NRE (Non-Recurring Engineering) costs associated with advanced process nodes (e.g., 3nm/2nm)Specific architectural choices for memory vs. compute-bound AI models

Also asked as

  • I need to understand the complete lifecycle of developing and manufacturing bespoke AI chips tailored to unique workloads. What does that workflow look like?