News and insights on edge AI
Releases and notes from the team.
LLMs & GenAIThe cost of running frontier AI modelsUnderstanding the high costs of developing and running frontier AI models, from training expenses to rising inference costs, and how solutions like Embedl SDK can help mitigate them.
Embedl··6 min read
LLMs & GenAIAgentic AI is Accelerating – But Can It Reach the Edge?Explore how agentic AI is evolving beyond generative AI and the challenges it faces in achieving edge deployment for real-time, autonomous applications.
Embedl··6 min read
LLMs & GenAILlama 4 on the Edge: Overcoming Limitations of Deploying Mixture-of-Experts on Edge DevicesExploring how Llama 4’s Mixture-of-Experts architecture can be optimized to run efficiently on edge devices, overcoming significant computational and memory challenges.
Embedl··7 min read
LLMs & GenAIGoogle vs. OpenAI: New Breakthroughs in the AI Race – But at What Cost?Google's new Gemini 2.5 Pro surpasses OpenAI's ChatGPT in benchmarks, but the energy cost of advancing AI raises sustainability concerns. Discover insights and solutions for efficient AI models.
Embedl··4 min read
LLMs & GenAIBattle of the SLMs: Gemma vs LLamaGemma 3 models challenge Llama 3.2 in SLM performance, excelling in math and coding while Llama leads in general knowledge, as tested by Embidl engineers.
Devdatt Dubhashi··4 min read
LLMs & GenAIDeep Learning’s Sputnik Moment?Chinese startup DeepSeek shakes up AI industry with breakthrough models, offering efficient, affordable solutions, and sparking a "Sputnik moment" for the West.
Devdatt Dubhashi··4 min read
LLMs & GenAIChatGPT in your lap: Bringing LLMs from the Cloud to the EdgeExplore how LLM technology is making its way to edge devices, revolutionizing AI applications. Discover the challenges and solutions in bringing LLMs to the masses for a more efficient and accessible future.
Devdatt Dubhashi··5 min read
LLMs & GenAITowards 1 Bit LLMs!Exciting advancements in LLM technology enable 1-bit models for affordable deployment, thanks to innovative compression methods like HQQ and QLoRA. Discover the future of accessible AI innovation!
Devdatt Dubhashi··4 min read
LLMs & GenAIVision Transformers vs CNNs at the EdgeVision Transformers vs CNNs: Discover the transformative power of Vision Transformers in AI and edge computing. Learn how they outperform CNNs and the potential for uniform solutions in vision tasks. Explore the hardware and optimization challenges, and get a…
Devdatt Dubhashi··5 min read
LLMs & GenAIThe Big Story of AI in 2023: LLMsDiscover the groundbreaking advancements in Large Language Models (LLMs) in 2023, from their transformative capabilities to their widespread accessibility. Explore how LLMs have revolutionized AI and the potential they hold for the future.
Devdatt Dubhashi··10 min read
LLMs & GenAIWhat do Vision Transformers See?Discover how Vision Transformers are revolutionizing computer vision with their general-purpose architecture and global self-attention. Learn about their advantages over CNNs and their potential for multi-modal inputs. Join our webinar to optimize ViTs for…
Devdatt Dubhashi··2 min read
LLMs & GenAIThe Future of LLMs: from Pieter Abbeel and John SchulmanThe Future of LLMs in robotics and reinforcement learning is discussed by Pieter Abbeel and John Schulman, exploring the capabilities and limitations of models like ChatGPT. They also delve into the role of academia versus industry in advancing research and…
Devdatt Dubhashi··2 min read
LLMs & GenAIDemocratizing LLMs: Efficiency is the Key!Democratizing LLMs: Learn how compression techniques are making large language models more efficient and accessible, opening doors for wider AI innovation.
Devdatt Dubhashi··2 min read
LLMs & GenAIA Picture is Worth … 16X16 Words: The Vision Transformer!Discover how the Vision Transformer revolutionizes image recognition by treating image patches as words. With impressive results, this blog explores the power of general-purpose methods in AI.
Devdatt Dubhashi··2 min read
LLMs & GenAIMaking Transformers EfficientLearn how to make Transformer models more efficient and reduce their size for low-resource hardware. Explore different methods like attention mechanism optimization, pruning attention heads, and assessing the benefits of model pruning.
Devdatt Dubhashi··2 min read
LLMs & GenAIEnter the Transformer!The Transformer: Revolutionizing AI with its attention-based architecture, transforming natural language processing and paving the way for generative language models. Stay updated with our newsletter.
Devdatt Dubhashi··3 min read