In 2019, OpenAI announced GPT-2 with this post: https://
openai.com/index/better-l
anguage-models/
… Today (~5 years later) you can train your own for ~$672, running on one 8XH100 GPU node for 24 hours. Our latest llm.c post gives the walkthrough in some detail: https://
github.com/karpathy/llm.c
/discussions/677
… Incredibly, the
AI HARDWARE
-
GPT-2 Training Now Affordable: $672 on Single GPU Node
By
–
-
Cerebras Leaders Win Women in AI Awards Recognition
By
–
Julie Choi and Elif Albuz represented Cerebras at the VentureBeat Transform Women in AI Awards. Congrats to all the nominees and winners!
-
Groq at Imagine AI Live: Fast AI Inference Demos and Workshops
By
–
See you at Imagine AI Live tomorrow. Stop by our table to see some demos of fast AI inference, and don't miss a presentation by @lifebypixels and a workshop by Hatice Ozen, Sr. Customer Apps Engineer. https://
hubs.la/Q02G9BhZ0 -
Galaxy S24 Ultra: Restore Old Photos with Integrated AI
By
–
Revive tus viejos negativos usando el Galaxy S24 Ultra y dales una nueva vida con el poder de la inteligencia artificial integrada en el teléfono 😍
— Juan Merodio (@juanmerodio) 11 juillet 2024
Tuto vía: hussseinlucky (Hussein alDulaimi) pic.twitter.com/zEFFCVQ8nLRevive tus viejos negativos usando el Galaxy S24 Ultra y dales una nueva vida con el poder de la inteligencia artificial integrada en el teléfono Tuto vía: hussseinlucky (Hussein alDulaimi)
-
Neuromorphic Computing: Balancing Opportunities and Technical Challenges
By
–
Well said. My work aims to capture the dual nature of neuromorphic computing, highlighting both its opportunities and challenges. Thank you.
Antonio. -
Quantization Methods Enhance Metis AIPU Efficiency Accuracy
By
–
Read our latest blog from Bram Verhoef and Martino Dazzi about how our quantization methods make the Metis AIPU highly efficient and accurate. https://
tinyurl.com/25ddyy8r
#ArtificialIntelligence #AI -
RAND Should Focus on Global Compute Acceleration
By
–
Looks like I have a new adversary. Sorry to say, I have been a huge fan of RAND for years. But they should be writing about global compute acceleration.
-
Untether AI speedAI architecture excels in AI inference performance
By
–
A @MoorInsStrat report highlights our at-memory compute architecture for #AI #inference. Our speedAI® family excels in performance, latency & energy-efficiency. See why this makes MI&S "bullish" about the prospects of Untether AI in our latest blog post. https://
untether.ai/untether-ai-wh
ere-performance-intersects-with-efficiency-by-moor-insights-strategy/
… -
Groq Developer Meetup: AMA, Demos, and Networking Success
By
–
Another Groq Developer Meetup is in the books! @GavinSherry
, VP of GroqCloud, & Andrew Ling, VP of Compiler Software, held an informative AMA, 5 developers showcased demos powered by Groq, and of course there was networking, & . Thanks to everyone who joined. See you in Aug! -
AI Trade Reckoning: Nvidia, Tesla, and Micron Analysis
By
–
The TL Podcast for July 8th: Reckoning with the AI trade Worrying about Nvidia’s upside, questioning Tesla’s turnaround, debating Micron Technology’s spending, and picking software winners https://
thetechnologyletter.com/the-posts/the-
tl-podcast-for-july-8th-reckoning-with-the-ai-trade
… Or catch it on Apple Podcasts: http://
bit.ly/40HBLJg $NVDA