Fair enough. My point is, spinning up fabs in America relies on Taiwanese people so people should be more mindful of saying their quiet things out loud about being ok with Taiwan being annexed and the rest of the world destabilized.
AI HARDWARE
-
Taiwan semiconductor independence strategic necessity America allies
By
–
Even if you all don’t give a shit about what happens to Taiwanese people, the pragmatic response would be that it’s in the interest of America and its allies that Taiwan remains free so an authoritarian country doesn’t control the world’s supply of advanced semiconductors.
-
TSMC Arizona Fab Requires 500 Engineers Knowledge Transfer
By
–
That TSMC fab in Arizona required 500 Taiwanese engineers to come over and help ramp things up while they sent 500 American engineers over to train. Given that it’s going to take about 50 advanced semiconductor fabs to catch up with Taiwan and about 5 years and $40B each to
-
Government GPU Clusters Gap in AI Leadership
By
–
Guess how many large GPU clusters the government owns or operates? None. https://
oodaloop.com/archive/2024/1
0/27/securing-the-future-why-the-government-must-invest-in-large-scale-gpu-clusters-to-lead-in-ai/
… -
Datacenter Teams Working Graveyard Shift Operations
By
–
Shoutout to our amazing datacenter teams working the graveyard shift
-
GenAI Power Consumption: No Need to Worry
By
–
Relax, no need to worry about the power needs of GenAI. https://t.co/IWQlnSAJLa
— Gary Marcus (@GaryMarcus) 26 octobre 2024Relax, no need to worry about the power needs of GenAI.
-
Cerebras Sets New Speed Record for Llama 3.1-70B Serving
By
–
Congrats @andrewdfeldman and @CerebrasSystems for a huge leap forward and setting a new speed record for serving Llama 3.1-70B. 2100 tokens/sec is blazingly fast for a 70B model. This is great for agentic AI! https://t.co/LmIAXKQb3W
— Andrew Ng (@AndrewYNg) 26 octobre 2024Congrats @andrewdfeldman and @cerebras for a huge leap forward and setting a new speed record for serving Llama 3.1-70B. 2100 tokens/sec is blazingly fast for a 70B model. This is great for agentic AI!
-
Groq Offers Dedicated Whisper Large v3 Instances Beyond Free Tier
By
–
If you’re interested in moving beyond the free Groq dev tier into a dedicated instance of Whisper Large v3, Whisper Large v3 Turbo or Distil-Whisper, learn more in our ASR Model Guide or reach out to us at sales@groq.com. https://
hubs.la/Q02VRN1g0 -
Boston Dynamics’ Electric Atlas Robot Demonstrates Strength with Pushups
By
–
Boston Dynamics' new electric version of Atlas shows off its strength and flexibility by doing pushups. @IntEngineering #Robot #AI #IoT #5G #CES2025 @Hal_Good @gvalan @enilev @Analytics_699 @FinMKTG
-
GPU Cluster Utilization Remains Below 50% in Practice
By
–
“From our conversations and other people’s experience, the real world utilization of large GPU cluster is often sub 50%!”