You can run llama 70b and load it all into memory. tried it last night and it worked quite well on my m2 ultra! Also tend to prefer it for adobe tools – is more stable. But yeah when it comes to things like radiance fields and photogrammetry, or gaming — PC FTW
AI HARDWARE
-
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
— Harold Sinnott #MWC26 (@HaroldSinnott) 25 août 2024
pic.twitter.com/CEcKgCs0YrBoston 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
-
AI-Powered Evacuation System Converts Balcony Railings into Escape Ladders
By
–
This evacuation system quickly turns balcony railings into escape ladders! @lexfridman @KirkDBorne @Ronald_vanLoon @erikbryn @antgrasso @sallyeaves @Nicochan33 @HaroldSinnott @mvollmer1 @marcusborba @CatherineAdenle @vinod1975 @YuHelenYu @ingliguori
-
Groq Supports Open Source Agent Hackathon with Andrew Ng
By
–
There's great energy at the Open Source Agent Hackathon with @AndrewYNg at @agihouse_org
. Can't wait to see what our developer community builds on Groq today. -
CUDA Installation Challenge and Potential Interview Questions
By
–
"Install CUDA without bricking your personal laptop. You have 6 hours." Jokes aside, sounds like a great initiative. What type of questions could they ask?
-
Grok-2-mini Speed Upgrade Through Advanced Inference Optimization
By
–
https://t.co/SDW9dmflJ7 pic.twitter.com/sXmhTXhuQJ
— Elon Musk (@elonmusk) 24 août 2024Grok-2-mini just got a speed upgrade. Over the past few days, we have substantially improved our inference stack. These gains come from using custom algorithms for computation and communication kernels, along with more efficient batch scheduling and quantization. Our inference x.com/ibab/status/18…
-
Plaud Voice Recorder with GPT-4o Integration Costs $159
By
–
you mean you don't want the Plaud with GPT-4o integration? it costs $159, considerably more than other popular voice recorders
-
Grok 2 Trained at Oracle Data Center
By
–
Special thanks to @Oracle at whose data center Grok 2 was trained
-
ExpertsInt8: Novel Quantization Technique for Jamba-1.5-Large Efficient Serving
By
–
To support efficient serving of Jamba-1.5-Large, we developed a novel quantization technique – ExpertsInt8. We quantize the MoE and MLP weights to INT8 in order to store them, and dequantize them back to BF16 before the actual computation. This technique is both very fast and