Intel 1971
Intel 4004
Informational Summary Areax86Process10 μmTransistor2KTotal Data Processing0.5 W
Family Tree Researcher for Microchips
Explore 55 years of microprocessor history: from Intel 4004 (1971, 2,300 transistors) to Apple A18 Pro and NVIDIA B200 (2024, over 200 billion). View lineage predecessor → successor, filter by architecture and process, explore the Chip War.
FreeClient-side processingPlease log in to continue29 Chip · 1971–2026
Static dataset based on Intel ARK, Apple Newsroom, NVIDIA white paper, Chris Miller "Chip War" (2022), Dwarkesh Patel podcast "The Chip Race" (2023-2024), AnandTech reviews and Wikipedia Transistor Count. Approximate transistor count and nm for hybrid generations (multi-die). Not technical or financial advice.
29Microchip
3 nmMinimum Node
208.0BMaximum Transistor
6Active Manufacturers
Timeframe: 1971-2026
Intel
AMD
Apple Silicon
ARM Holdings
NVIDIA
SiFive
Chip Federico Calo/Federico Calo2929
Intel 1974
Intel 8080
Informational Summary Areax86Process6 μmTransistor6KTotal Data Processing1 W
Has a predecessor
Intel 1978
Intel 8086
Informational Summary Areax86Process3 μmTransistor29KTotal Data Processing2.5 W
Has a predecessor
Intel 1985
Intel 80386
Informational Summary Areax86Process2 μmTransistor275KTotal Data Processing4 W
Has a predecessor
Intel 1993
Intel Pentium
Informational Summary Areax86Process800 nmTransistor3.1MTotal Data Processing10 W
Has a predecessor
AMD 1997
AMD K6
Informational Summary Areax86Process350 nmTransistor8.8MTotal Data Processing18 W
AMD 1999
AMD Athlon (K7)
Informational Summary Areax86Process250 nmTransistor22.0MTotal Data Processing42 W
Has a predecessor
Intel 2000
Intel Pentium 4 (Willamette)
Informational Summary Areax86Process180 nmTransistor42.0MTotal Data Processing55 W
Has a predecessor
Intel 2006
Intel Core Duo (Yonah)
Informational Summary Areax86Process65 nmTransistor151.0MTotal Data Processing31 W
Has a predecessor
ARM 2007
ARM Cortex-A8
Informational Summary AreaARMProcess65 nmTransistor166.0MTotal Data Processing0.8 W
Intel 2008
Intel Core i7-920 (Nehalem)
Informational Summary Areax86-64Process45 nmTransistor731.0MTotal Data Processing130 W
Has a predecessor
Apple 2010
Apple A4
Informational Summary AreaARMProcess45 nmTransistor153.0MTotal Data Processing2 W
Apple 2011
Apple A5
Informational Summary AreaARMProcess45 nmTransistor800.0MTotal Data Processing3 W
Has a predecessor
ARM 2012
ARM Cortex-A53
Informational Summary AreaARMProcess28 nmTransistor30.0MTotal Data Processing0.3 W
Has a predecessor
Apple 2013
Apple A7 (Cyclone)
Informational Summary AreaARMProcess28 nmTransistor1.0BTotal Data Processing4 W
Has a predecessor
AMD 2017
AMD Ryzen 7 1800X (Zen 1)
Informational Summary Areax86-64Process14 nmTransistor4.8BTotal Data Processing95 W
Has a predecessor
Apple 2017
Apple A11 Bionic
Informational Summary AreaARMProcess10 nmTransistor4.3BTotal Data Processing6 W
Has a predecessor
NVIDIA 2017
NVIDIA Tesla V100
Informational Summary AreaGPU-CUDAProcess12 nmTransistor21.1BTotal Data Processing300 W
SiFive 2018
SiFive U54 (Freedom U540)
Informational Summary AreaRISC-VProcess28 nmTransistor6.0MTotal Data Processing8 W
Intel 2020
Intel Core i9-10900K (Comet Lake)
Informational Summary Areax86-64Process14 nmTransistor10.0BTotal Data Processing125 W
Has a predecessor
Apple 2020
Apple A14 Bionic
Informational Summary AreaARMProcess5 nmTransistor11.8BTotal Data Processing7 W
Has a predecessor
ARM 2020
ARM Cortex-A78
Informational Summary AreaARMProcess5 nmTransistor45.0MTotal Data Processing5 W
Has a predecessor
NVIDIA 2020
NVIDIA A100 (Ampere)
Informational Summary AreaGPU-CUDAProcess7 nmTransistor54.2BTotal Data Processing400 W
Has a predecessor
AMD 2022
AMD EPYC 9654 (Genoa)
Informational Summary Areax86-64Process5 nmTransistor80.0BTotal Data Processing360 W
Has a predecessor
NVIDIA 2022
NVIDIA H100 (Hopper)
Informational Summary AreaGPU-CUDAProcess4 nmTransistor80.0BTotal Data Processing700 W
Has a predecessor
Apple 2023
Apple A17 Pro
Informational Summary AreaARMProcess3 nmTransistor19.0BTotal Data Processing8 W
Has a predecessor
Apple 2024
Apple A18 Pro
Informational Summary AreaARMProcess3 nmTransistor22.0BTotal Data Processing9 W
Has a predecessor
NVIDIA 2024
NVIDIA B200 (Blackwell)
Informational Summary AreaGPU-CUDAProcess4 nmTransistor208.0BTotal Data Processing1000 W
Has a predecessor
SiFive 2024
SiFive P870 (Intel Horse Creek)
Informational Summary AreaRISC-VProcess7 nmTransistor500.0MTotal Data Processing25 W
Has a predecessor
Come utilizzare Microchip Genealogy Explorer
Explore the SVG Timeline 1971-2026
The top timeline shows each chip as a colored point by manufacturer, placed on the year axis. Click a point to open the chip's detail card.
Filter by manufacturer, architecture and year
In the sidebar select one or more manufacturers (Intel, AMD, Apple, ARM, NVIDIA...) and architectures ISA (x86, ARM, RISC-V, GPU-CUDA...), then restrict the year interval with the sliders From/To.
Open chip detail sheet
Clicking on a die in the grid or timeline shows transistor count, process nm, clock, TDP, historical significance, sources and the lineage of predecessor -> successor if available.
Arrange your grill chips
Use the buttons "Year", "Transistor", or "nm Process" to sort the filtered chip grid according to your preferred comparison criteria.
Suggerimenti
- Filter by architecture RISC-V to see the evolution of the fully open-source ISA from early Berkeley cores to latest SiFive chips.
- Sort by "Transistor" to quickly view the trend of Moore's Law from 4004 (2,300 transistors) to modern GPUs (over 200 billion).
- Use combined manufacturer filters (e.g. Intel + AMD) to compare the x86 architectures of the two main historical competitors.
Domande frequenti
Where do the data on 30 historic chips come from?
The dataset is static and built from publicly sourced materials: Intel ARK, Apple Newsroom, NVIDIA white paper, "Chip War" by Chris Miller (2022), "The Chip Race" podcast by Dwarkesh Patel, AnandTech, and Wikipedia Transistor Count. Each chip sheet lists the specific sources.
What does the "lineage" chain shown on the chip card mean?
The lineage connects a chip to its direct predecessor (predecessorId in the dataset), tracing back along the chain until the oldest available chip is shown, only when the selected chip has at least one registered predecessor in the dataset.
How precise are the values for transistor count and process nm?
Values are approximate, especially for hybrid multi-die generations (e.g., modern GPU and CPU with multiple chiplets), where transistor counting aggregates separate dies. Always refer to official data from the cited sources on each card.
Are data loaded from a server or are they client-side?
The dataset of 30 chips is statically included in the app's bundle: no API calls. Filters, sorting and timelines are fully calculated in the browser.
Why do some chips lack lineage or predecessors?
Some chips in the dataset represent the start of a new architectural family (e.g. Intel 4004 in 1971) or do not have a direct predecessor tracked in the current dataset, which covers 30 selected chips and not the entire history of microprocessors.