A personal look back, from Saarbrücken in 2010 to SIGGRAPH 2026. The full numbers are in SIGGRAPH 2026, Laid Bare.
Share
I started my PhD at the Max Planck Institute for Informatics in Saarbrücken in 2008. Two of my papers made it into SIGGRAPH Asia: MovieReshape, in Seoul in 2010, and Material Memex, in Singapore in 2012.
Meanwhile, this was happening in the world.
My AI agents had nothing to do, so I got them to go back and study twenty years of SIGGRAPH. They counted every paper by where its first author works. The full numbers are in SIGGRAPH 2026, Laid Bare. This is the view from my seat.
SIGGRAPH Asia 2010: papers led, by the country of the first author's institution (49 papers placed). Hover or click a number for the institutions behind it.
At SIGGRAPH Asia 2010, America led 23 of 49 papers. Germany led five. My institute led four of them, and tied for the most papers in the world. China led three. India led none.
Singapore in 2012 looked much the same. America led 33 papers, China 11, Germany 8. Max Planck and ETH Zurich tied for the most, with five each. India led none again. In 2008, the year I started, my institute led 11 papers. All of China led 12.
In 2006, America led 69% of the papers. This year it led 22%. China went from 5% to 37%, and passed America in 2024. Germany fell from 12% in 2008 to 3%.
Hover over the chart to see any year.
Source: ACM's records, 2006–2025 (SIGGRAPH Asia from its first meeting, 2008); the official programs for 2026. Hong Kong is counted apart from China.
Both 2026 meetings together: papers led, by the country of the first author's institution (618 papers). Hover or click a number for the institutions behind it.
China leads the new topics: generative AI, 3D generation and Gaussian splatting. America still leads physics simulation and geometry. Laid Bare has the split, and a wall of this year's paper videos.
India is getting better, from almost nothing. In twenty years, 15 papers had a first or last author in India. The first author is usually the student; the last is usually the professor. Only eight had both in India. All eight are from 2019 or later: one from DA-IICT in Gandhinagar, five from IISc in Bengaluru and two from IIIT Hyderabad. Five professors wrote all eight. Rajiv Soundararajan's lab at IISc wrote half.
Before 2019, India shows up only through individuals. Niloy Mitra wrote Shadow Art and Emerging Images in 2009, during a short time at IIT Delhi. He then moved to KAUST and later UCL. Rahul Narain, at IIT Delhi, was last author on three papers from 2018 to 2021. Their first authors were at the University of Minnesota.
Every SIGGRAPH and SIGGRAPH Asia paper with a first or last author in India. Names in amber worked in India. Hover for all authors. Click to pin it and open the paper.
2009
2010–2017: no paper with an India-based first or last author
2018
2019
2020
2021
2022: no paper with an India-based first or last author
2023
2024
2025
2026
Source: ACM's records (2006–2025) and the official 2026 programs. Not counted: India as a second affiliation, short papers (Technical Communications), and ACM TOG papers presented at the meetings.
The good thing is that SIGGRAPH is still small. SIGGRAPH 2026 accepted 309 of 1,124 submissions. SIGGRAPH Asia accepted 309 of 1,303. NeurIPS accepted 9,100, about thirty times as many.
This is one conference that truly values the idea and its novelty.
If the idea has been done, they ask how it was done. After Niloy Mitra published Shadow Art, you can't just write another shadow art paper. You need a new idea, or a truly new way to do an old one.
All twenty years, every institution and the videos are in SIGGRAPH 2026, Laid Bare.
How I counted: each paper counts once, for the institution of its first author (the rule The Economist uses for NeurIPS), except the India timeline, which also counts the last author. 2006 to 2025 come from ACM's records of SIGGRAPH and SIGGRAPH Asia technical papers, 2026 from the official programs. Hong Kong is counted apart from China.
Incorporate AI ML into your workflows to boost efficiency, accuracy, and productivity. Discover our artificial intelligence services.
View All
Chinese first authors led 100 SIGGRAPH 2026 papers to America's 80, even in Los Angeles. Twenty years of SIGGRAPH and SIGGRAPH Asia, 618 papers from 2026 and a wall of their videos.
Tsinghua is #1, KAIST #2, and 8 of the top 10 institutions are in Asia: who wrote NeurIPS 2026's 9,100 papers, where they work and what they argue about, in 32 interactive charts.
China passed America at ICRA in 2026, and its robots are in a third of US-led papers that name one; US-led papers wrote half of the most-cited 1%. 19,273 papers from IROS and ICRA since 2021, in interactive charts and a directory.
© Copyright Fast Code AI 2026. All Rights Reserved