still need to test it in concurrency + try it in nvfp4 hopefully will be able to compare perplexity in bf16 / fp8 / nvfp4 as well as tp performance jumps from 2 -> 4 -> 8 nodes across all 3
AI HARDWARE
-
MiniMax M2.7 on 4x DGX Sparks with vLLM and OpenCode
By
–
MiniMax M2.7 at home running on 4x DGX Sparks
— Ahmad (@TheAhmadOsman) 12 avril 2026
vLLM serving full BF16 weights, 200k context
OpenCode having the model monitor its own hardware and report thermals, tokens/sec, TTFT, and other runtime stats in real time
What benchmarks / workflows / things do you wanna see next? https://t.co/iCR8UsLs8E pic.twitter.com/I04c7j6UkgMiniMax M2.7 at home running on 4x DGX Sparks vLLM serving full BF16 weights, 200k context OpenCode having the model monitor its own hardware and report thermals, tokens/sec, TTFT, and other runtime stats in real time What benchmarks / workflows / things do you wanna see next?
-
MIT Cheetah Quadruped Robot Redefining High-Performance Legged Locomotion
By
–
MIT Cheetah: The Agile Quadruped #Robot Redefining High-Performance Legged Locomotion
— Ronald van Loon (@Ronald_vanLoon) 12 avril 2026
by @IntEngineering
#Robotics #Engineering #ArtificialIntelligence #Innovation #Technology pic.twitter.com/9A8MwVNpc2MIT Cheetah: The Agile Quadruped #Robot Redefining High-Performance Legged Locomotion
by @IntEngineering #Robotics #Engineering #ArtificialIntelligence #Innovation #Technology -
Iran and Houthis Threaten Global AI Chip Supply Chain
By
–
Iran and Huthis: How two Middle East chokepoints could strangle the Global AI Chip supply chain For days I've been trying to make explain the Iran war and the unfolding supply chain problems. The current situation has worsened. Iran is now threatening to draw the Houthis into
-
AI models debate power usage of local OpenClaw on Mac
By
–
And if you are interested in the answer, Claude and ChatGPT 5.4 Pro think running OpenClaw with local inference on your Mac uses more total power. Gemini doesn't think so (but appears to have not thought about that too hard)
-
JCBOT Wheeled Humanoid Combines Speed AI Interaction
By
–
JCBOT Wheeled Humanoid Blends Speed, #AI Interaction and Home-Friendly Design
— Ronald van Loon (@Ronald_vanLoon) 12 avril 2026
via @XRoboHub#Robotics #Engineering #ArtificialIntelligence #Innovation #Technology pic.twitter.com/VS4J9xIzfcJCBOT Wheeled Humanoid Blends Speed, #AI Interaction and Home-Friendly Design
via @XRoboHub #Robotics #Engineering #ArtificialIntelligence #Innovation #Technology -
Advanced Autonomous Agents Interacting With Physical World Mid-2030s
By
–
I see this as a route for advanced autonomous agentic AI (neurosymbolic & neuromorphic) to experience and interact with the physical world in the mid 2030s onwards
-
Scientists Hijacked Nvidia CUDA for Deep Learning Research
By
–
Nvidia did not "sell its gaming architecture to a handful of scientists" A handful of scientists such as Patrice Simard et MSR, Andrew Ng, and later Geoff Hinton, hijacked CUDA and implemented convnets and backprop on them. It took several years for Nvidia to realize there was
-
Training Neural Networks with Optical Backpropagation Technology
By
–
Training Neural Network on Optical Backpropagation! #BigData #Analytics #DataScience #AI #MachineLearning #IoT #IIoT #PyTorch #Python #RStats #TensorFlow #Java #JavaScript #ReactJS #GoLang #CloudComputing #Serverless #DataScientist #Linux #Programming #Coding #100DaysofCode References Linnainmaa, S. I. (1976). Taylor expansion of the accumulated rounding error. BIT Numerical Mathematics, 16(2), 146–160. Published June 1976. Retrieved March 10, 2025, from doi.org/10.1007/BF01931367 Lvovsky, A. I., Guo, X., & Spall, J. (2025). Training neural networks with end-to-end optical backpropagation. Advanced Photonics, 7(1), 016004. Published February 4, 2025. Retrieved March 10, 2025, from doi.org/10.1117/1.AP.7.1.016… Nielsen, M. A. (2015). How the backpropagation algorithm works. In Neural networks and deep learning. Published 2015. Retrieved March 10, 2025, from neuralnetworksanddeeplearnin… Spall, J., Guo, X., & Lvovsky, A. I. (2025). The optical implementation of backpropagation (Oxford, Lumai). Semiconductor Engineering. Published February 4, 2025. Retrieved March 10, 2025, from semiengineering.com/the-opti…
→ View original post on X — @gp_pulipaka, 2026-04-12 14:26 UTC