The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
The Neural Information Processing Systems (NeurIPS) is the largest artificial intelligence (AI) conference in the world. NeurIPS 2023 will be taking place December 10-16 in New Orleans, USA.Teams from across Google DeepMind are presenting more than 150 papers at the main conference and workshops.Read More
Now available in the U.S., NotebookLM has new features to help you easily read, take notes and organize your writing projects.Read More
Generative AI has revolutionized how we create text and images. How about designing novel materials? We at Microsoft Research AI4Science are thrilled to announce MatterGen, our generative model that enables broad property-guided materials design.
The central challenge in materials science is to discover materials with desired properties, e.g., high Li-ion conductivity for battery materials. Traditionally, this has been done by first finding novel materials and then filtering down based on the application. This is like trying to create the image of a cat by first generating a million different images and then searching for the one with a cat. In MatterGen, we directly generate novel materials with desired properties, similar to how DALL·E 3 tackles image generation.
MatterGen is a diffusion model specifically designed for generating novel, stable materials. MatterGen also has adapter modules that can be fine-tuned to generate materials given a broad range of constraints, including chemistry, symmetry, and properties. MatterGen generates 2.9 times more stable (≤ 0.1 eV/atom of our training + test data convex hull), novel, unique structures than a SOTA model (CDVAE). It also generates structures 17.5 times closer to energy local minimum. MatterGen can directly generate materials satisfying desired magnetic, electronic, mechanical properties via classifier-free guidance. We verify generated materials with DFT-based workflows.
Additionally, MatterGen can keep generating novel materials that satisfy a target property like high bulk modulus while screening methods instead saturate due to the exhaustion of materials in the database.
MatterGen can also generate materials given target chemical systems. It outperforms substitution and random structure search baselines equipped with MLFF filtering, especially in challenging 5-element systems. MatterGen also generates structures given target space groups. Finally, we tackle the multi-property materials design problem of finding low-supply-chain risk magnets. MatterGen proposes structures that have both high magnetic density and a low supply-chain risk chemical composition.
We believe MatterGen is an important step forward in AI for materials design. Our results are currently verified via DFT, which has many known limitations. Experimental verification remains the ultimate test for real-word impact, and we hope to follow up with more results.
None of this would be possible without the highly collaborative work between Andrew Fowler, Claudio Zeni, Daniel Zügner, Matthew Horton, Robert Pinsler, Ryota Tomioka, Tian Xie and our amazing interns Xiang Fu, Sasha Shysheya, and Jonathan Crabbé, as well as Jake Smith, Lixin Sun and the entire AI4Science Materials Design team.
We are also grateful for all the help from Microsoft Research, AI4Science, and Azure Quantum.
The post MatterGen: Property-guided materials design appeared first on Microsoft Research.
We’re celebrating a milestone this week with 500 RTX games and applications utilizing NVIDIA DLSS, ray tracing or AI technologies. It’s an achievement anchored by NVIDIA’s revolutionary RTX technology, which has transformed gaming graphics and performance.
The journey began in 2018 at an electrifying event in Cologne. In a steel and concrete music venue amidst the city’s gritty industrial north side, over 1,200 gamers, breathless and giddy, erupted as NVIDIA founder and CEO Jensen Huang introduced NVIDIA RTX and declared, “This is a historic moment … Computer graphics has been reinvented.”
This groundbreaking launch, set against the backdrop of the world’s largest gaming expo, Gamescom, marked the introduction of the GeForce RTX 2080 Ti, 2080 and 2070 GPUs.
The most technically advanced games now rely on the techniques that RTX technologies have unlocked.
Ray tracing, enabled by dedicated RT Cores, delivers immersive, realistic lighting and reflections in games.
The technique has evolved from games with only a single graphics element executed in ray tracing to games such as Alan Wake 2, Cyberpunk 2077, Minecraft RTX and Portal RTX that use ray tracing for all the light in the game.
And NVIDIA DLSS, powered by Tensor Cores, accelerates AI graphics, now boosting performance with DLSS Frame Generation and improving RT effects with DLSS Ray Reconstruction in titles like Cyberpunk 2077: Phantom Liberty.
Beyond gaming, these technologies revolutionize creative workflows, enabling real-time, ray-traced previews in applications that once required extensive processing time.
Ray tracing, a technique first described in 1969 by Arthur Appel, mirrors how light interacts with objects to create lifelike images.
Ray tracing was once limited to high-end movie production. NVIDIA’s RTX graphics cards have made this cinematic quality accessible in real-time gaming, enhancing experiences with dynamic lighting, reflections and shadows.
High engagement rates in titles like Cyberpunk 2077, NARAKA: BLADEPOINT, Minecraft with RTX, Alan Wake 2 and Diablo IV, where 96% or higher of RTX 40 Series t gamers use RTX ON, underscore this success.
To commemorate this milestone, 20 $500 Green Man Gaming gift cards and exclusive #RTXON keyboard keycaps are up for grabs. Participants must follow GeForce’s social channels and comply with the sweepstakes rules.
Stay tuned for more RTX 500 giveaways.
NVIDIA’s advancement from the first RTX graphics card to powering 500 RTX games and applications with advanced technologies heralds a new gaming and creative tech era. And NVIDIA continues to lead, offering unparalleled experiences in gaming and creativity.
Stay tuned to GeForce News for more updates on RTX games and enhancements.
