Ahmed Mahmoud

Postdoc MIT (he/him)

[Nick Sharp @nmwsharp]

Have you dug deep into an evolving area of visual computing? Consider submitting to the 2027 EG STARs track!

STARs are survey reports providing insight into the current state of the art and future directions. They're impactful and highly-cited!

https://t.co/5d8bBduZDj

[Ticha Sethapakdi @tichaesque]

Thank you for promoting our workshop!! 🥰 If you’re at #SIGGRAPH2026 this year and want to do animation and a little bit of ✨magic✨, drop by our workshop this Thursday from 2-5:15pm (room 501 ABC)! Design, fabricate, and take home animations you can hold in your hands 🪄

[Ahmed Mahmoud @ahdhn]

@chrisoffner3d ops… here is the correct one https://t.co/mJLWxg0MAC

[Ahmed Mahmoud @ahdhn]

Next week at #SIGGRAPH2026, we'll present "Fast Sparse Matrix Permutation for Mesh-Based Direct Solvers," with Behrooz Zarebavani, @ana_dodik, Changcheng Yuan, @serban, @jowens, Maryam Mehri Dehnavi, and @JustinMSolomon

Talk: Wed., July 22, 3:55 PM, Room 403-B (1/6)

[Ahmed Mahmoud @ahdhn]

Our code is open source, it is really simple to use in both C++ and Python—with cuDSS, MKL, and CHOLMOD backends.

Code: https://t.co/Lppvsi6tNd
Python: pip install homapy https://t.co/WTOZIdqHpS
Paper: https://t.co/PtHs9ZBYaL (6/6)

[Ahmed Mahmoud @ahdhn]

While the method is designed for mesh-derived matrices, we also tested it on sparse matrices in the wild. Results are more variable, but large matrices often benefit, with up to 18.2× end-to-end speedup (5/6)

[Ahmed Mahmoud @ahdhn]

The full code and all applications are available in RXMesh which recently received the 2026 SGP Software Award. Come by the talk if you are at SIGGRAPH and interested in GPU systems, autodiff, geometry, and simulation.

Code: https://t.co/rBHDVaYxLY (7/7)

[Ahmed Mahmoud @ahdhn]

Next week at #SIGGRAPH 2026, we'll present "Locality-Aware Automatic Differentiation on the GPU for Mesh-Based Computations," with @__lankylad__, @jrk, and @JustinMSolomon

Talk: Mon., July 20, 2:50 PM, Room 403-B
Paper: https://t.co/zHcIt2MkrK

More in the thread below (1/7):

[Ahmed Mahmoud @ahdhn]

While our main focus is sparse Jacobians/Hessians, our system also outperforms every evaluated autodiff baseline, plus handwritten one, on a (simple) gradient benchmark (6/7)

[Artem Lukoianov @ottogin1]

Join us at @CVPR in Denver for a full-day tutorial about Analytic Understanding of Diffusion Models.

The training objective of diffusion models has a closed-form solution -- yet it only memorizes. How do real models generalize? We'll unpack this paradox and the emerging analytical theory behind it.

@yuancy @CScarvelis @MasonKamb @WangBinxu @vincesitzmann @JustinMSolomon @SuryaGanguli

[sina nabizadeh @sinabiz_]

📢 New England Symposium on Graphics 2026 returns April 12 @MIT! Featuring talks by Profs Frédo Durand (MIT), @AdrianaSchulz7 (Brown), Theodore Kim (Yale), junior talks, posters, and networking. Free to attend!!
Register: https://t.co/uj1rIF67Fc
Details: https://t.co/uXBFUJ3cng

[Justin Solomon @JustinMSolomon]

Announcing SGI 2026! Undergrads and MS students: Apply for 6 weeks of paid summer geometry processing research. No experience needed: 1 week tutorials + 5 weeks of projects. Mentors are top researchers in this emerging branch of graphics/computing/math. https://t.co/TKEXHZwgzo

[The Nobel Prize @NobelPrize]

The day after this year's Nobel Prize in Chemistry was announced, Richard Robson stood before his students at the University of Melbourne, teaching as he always did.

He shared the 2025 chemistry prize with Susumu Kitagawa and Omar Yaghi for the development of metal–organic frameworks.

Later today, the 2025 laureates will meet in Stockholm for the first time, marking the beginning of Nobel Week. Follow along online: https://t.co/XmOhARfots

Photo: Mireille Stahle/University of Melbourne