Talk Around Town: A Mystery Site Claims to Have Rounded Up Every Free Processor Die Photo Ever Taken

What Folks Are Chattering About

Well, shoot — word is spreading around Hacker News faster than gossip at a church potluck about a data-visualization project sitting quietly out at sheets.works. According to the sheets.works project page, somebody — no named author, no named organization, just a lonesome website standing in a digital field — has gone and rounded up 356 freely licensed microscope photographs of processor dies, spanning all the way from 1972 clear up to 2024. The site claims every last one of those images is free to use, like a pile of firewood left at the end of a neighbor's driveway with a 'help yourself' sign nailed to it.

A community summary over at daily.dev independently describes the same project details, including the photo count, the contributor breakdown, and a special comparison feature the site calls a 'true scale' view. That gives us at least two sources describing the same observable facts, which is a start — though neither one is an independent editorial newsroom doing original verification, so hold your horses before you go betting the farm on any of these figures.

What the Project Page Actually Says

According to the sheets.works project page, getting one of these die photographs in the first place is not exactly a lazy Sunday afternoon activity. The site explains that you have to either sand down or chemically dissolve the chip's outer package — the little black square you see on a circuit board — and then peer at the exposed silicon innards through a microscope. It is, the site says, a finicky process, like trying to skin a catfish with a pair of tweezers while wearing oven mitts.

The sheets.works page attributes the lion's share of the collection's images to a small volunteer crew. According to the project, Pauli Rautakorpi contributed the most with 162 photos, followed by Fritzchens Fritz at 68, John McMaster at 36, Martijn Boer at 20, Cole L at 16, and ZeptoBars at 10, among others. The daily.dev summary corroborates this contributor breakdown. These are real people putting in real elbow grease for zero dollars, which is either admirable or baffling depending on how you feel about microscopes.

The project's headline trick, according to the sheets.works page and confirmed by the daily.dev summary, is a view the site calls 'true scale.' The page claims it renders the 114 dies whose physical dimensions are actually known all at the same uniform scale, so that a chip from the 1970s supposedly appears as nothing more than a tiny speck sitting next to a hulking modern processor. It is, by the site's own description, a visual gut-punch that shows just how much silicon engineering has crammed into smaller and smaller packages over fifty years.

According to the daily.dev summary, each photograph in the collection keeps its original license intact and links back to its source, with origins including Wikimedia Commons, Flickr, ZeptoBars, Siliconpr0n, and the CPU Graveyard wiki. The sheets.works project makes no commercial claims, which at least keeps the snake-oil risk low, even if the anonymous authorship keeps us from knowing exactly whose barn we're standing in.

What We Actually Know for Certain

Here is the solid ground, the stuff that does not shift when you poke it. The broader die-photography hobbyist world is independently documented and real as a tractor pull. The CPU Graveyard wiki over at happytrees.org catalogs more than 1,500 die shots from 167 manufacturers across 651 chip families — that is an independently verifiable resource that exists regardless of whatever sheets.works is or isn't doing.

Wikipedia separately documents the practice of die photography and notes something worth chewing on: chip manufacturers are often reluctant to let these images loose in the wild, because a competitor might peek at a die shot and glean some insight into how technologically sophisticated a chip is. Wikipedia does note, however, that the resolution of typical die photography is generally too low to actually reproduce the circuits themselves — so it is more like seeing someone's general floor plan than getting their blueprints. That tension between openness and competitive caution helps explain why freely licensed die photos are a rare and somewhat precious commodity.

What Remains Unverified and Murky as a Catfish Pond

Nobody outside of the sheets.works project itself has gone and independently counted those 356 photographs, verified the contributor attributions, or checked the math behind the true-scale measurements. No independent newsroom has touched this story. The site has no named author and no named organization behind it. We are, in effect, taking the word of an anonymous website, with a community aggregator summary as the only additional voice — and that summary is itself just describing what the site says, not independently auditing it.

It is also worth noting that the sheets.works collection and the CPU Graveyard wiki are not the same thing and cannot be directly compared. The sheets.works page claims 356 openly and freely licensed photographs only. The CPU Graveyard's 1,500-plus shots reportedly include images under various license terms, not all of them free for general use. So those two numbers are measuring different things, like comparing a jar of homemade jam to a whole grocery store shelf.

One more wrinkle worth flagging: the CPU Graveyard's resources page, according to available information, lists several prominent die photographers — including Pauli Rautakorpi, who the sheets.works project credits with the most contributions — as inactive, while listing others like Fritzchens Fritz, Cole L, and ZeptoBars as still active. If that characterization is accurate, the hobby community may be slowly shrinking even as archival projects try to gather and preserve what has already been created.

Our Analysis: Small Project, Real Niche, Real Unknowns

This is analysis, not reporting. The sheets.works project, whatever its ultimate accuracy, is pointing at something genuinely interesting: a tiny, scattered community of hobbyists has quietly produced hundreds of freely licensed scientific images of some of the most consequential objects in modern technological history. The fact that so few people are doing it — and that some of the most prolific contributors may have already stepped back — makes the archival impulse behind a project like this feel less like a quirky side project and more like a preservation effort for a disappearing craft.

The Hacker News attention is a soft signal that the technical community finds both the collection and the true-scale comparison feature compelling, but low overall spread — this cluster has only one independent discussion channel and a below-50 cluster score — means this has not yet broken through to a wider audience. It may stay a niche curiosity, or it may eventually attract enough eyeballs to prompt the kind of independent verification that would let us say something more definitive. Right now, though, it is a promising-looking barn with nobody standing outside to tell you whose land you're on.