Take it from @techradar SambaNova runs @deepseek_ai R1 at 198 tokens/sec using 16 custom chips! Our SN40L RDU chip is 3X faster & 5X more efficient than GPUs 5X speed boost coming soon, with 100X capacity on cloud. https://
bit.ly/3Xatj6k
HARDWARE
-
SambaNova Achieves 198 Tokens Per Second With Custom AI Chips
By
–
-
DeepSeek-R1 Achieves 95% Chip Reduction Industry-First Performance
By
–
@techradar reported that SambaNova’s @deepseek_ai
-R1 achieved industry-first #AI performance with 95% fewer chips. @RodrigoLiang told @Waynewill1
: “We’re collapsing inference costs and slashing hardware demands—from 40 racks to just one.” #GPU #DeepSeek -
Microsoft’s topological qubit arrives seven years late
By
–
Microsoft’s ‘topological qubit’ is seven years late and a million bits short Microsoft told me they would have a “topological qubit” in 2018. It’s only just arrived this week — maybe. There’s skepticism among scientists as to whether Microsoft’s paper in Nature actually proves
-
Grok’s H100 GPU Training: Missing Transparency on Key Metrics
By
–
The number of H100 GPUs used for training helps estimate parameters, data size, and training duration, so it’s an important detail. But Grok’s team hasn’t been clear about key factors like benchmark performance and the exact number of GPUs used for training, which leads to
-
Grok-3 Training Hardware: 100k vs 200k H100s Clarification
By
–
I want to clarify this. Was Grok-3 trained on 200k H100s, or is it being served using them? I remember hearing in the live Q&A that it was trained on 100k H100s.
-
Microsoft Creates New State of Matter for Quantum Computers
By
–
Microsoft Says It Has Created a New State of Matter to Power Quantum Computers
-
AI CUDA Engineer: Agentic Kernel Discovery and Optimization
By
–
The AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Compositionhttps://t.co/fMDhdIkNIo
— hardmaru (@hardmaru) 20 février 2025
In principle, AI can be as resource-efficient as our brains 🧠. To achieve this, we can use AI to make AI more efficient! Eventually, AI will become a million times faster. https://t.co/84rZGBKHnsThe AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Composition https://
pub.sakana.ai/ai-cuda-engine
er/paper/
… In principle, AI can be as resource-efficient as our brains . To achieve this, we can use AI to make AI more efficient! Eventually, AI will become a million times faster. -
Artificial Intelligence Increasing Resource Demands Significantly
By
–
Artificial Intelligence Is Putting Ever-Increasing Demands on Our Resources
#AI #AIio #BigData #ML #NLU #Futureofwork @lexfridman @sama @kaifulee @ID_AA_Carmack @karpathy @2morrowknight @ylecun -
New State of Matter Powers First Topological Qubits
By
–
What a breakthrough! Published in Nature today, this is "a new state of matter" to power the first topological qubits. With scalable quantum computing, we'll soon be able to solve previously impossible problems in chemistry, biology and life sciences. Truly exciting times.
-
Ultra-Scale Playbook: 5D Parallelism and CUDA Optimization Guide
By
–
After 6+ months in the making and burning over a year of GPU compute time, we're super excited to finally release the "Ultra-Scale Playbook" Check it out here: http://
hf.co/spaces/nanotro
n/ultrascale-playbook
… A free, open-source, book to learn everything about 5D parallelism, ZeRO, fast CUDA kernels,