Models & Labs
Thinking Machines releases Inkling model
Fireshipmedium confidence
Why it matters
- →Inkling's open-weights model allows for greater customization by developers.
- →The 975 billion parameters make it one of the largest models available.
- →This release could democratize access to advanced AI capabilities.
Thinking Machines has released Inkling, a large language model with 975 billion parameters. The model is open-weights, allowing developers to access and customize it freely. Despite being described as 'deliberately mid,' Inkling represents a significant addition to the landscape of large AI models. This release could enhance the accessibility of advanced AI tools for developers, promoting further innovation in the field.
Read originalMore in Models & Labs
Models & Labsmodels
llama.cpp b10069 Release Enhances OpenCL Support
The b10069 release of llama.cpp brings notable improvements to OpenCL support, particularly targeting Adreno GPUs. By enabling broadcast for Adreno MUL_MAT and respecting view offsets, this update aims to boost performance for multi-stream operations on llama-server. The release also extends general GEMM/GEMV support for broadcast, which could optimize operations across different hardware setups. Although there are no revolutionary new features, these updates represent a consistent enhancement in compatibility and performance, especially for developers working with a range of hardware configurations.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10075 release expands platform support
The b10075 release of llama.cpp marks a significant step in enhancing its compatibility across diverse hardware setups. With the addition of ROCm 7.2 support on Ubuntu, AMD GPU users can now enjoy improved performance. Windows users benefit from the inclusion of CUDA 13.3, ensuring better integration with NVIDIA GPUs. The update also brings Vulkan support, which optimizes GPU utilization for developers. Although no new model architectures are introduced, this release reinforces llama.cpp's role as a flexible and adaptable inference runtime for developers working in varied environments.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google Developing New AI Chip for Gemini Models
Google is reportedly working on a new AI chip, dubbed 'Frozen v2', aimed at significantly enhancing the efficiency of its Gemini models. Expected to be released by 2028, this chip could be six to ten times more efficient than current AI chips, potentially transforming Google's AI capabilities. This move aligns with a broader industry trend where tech giants are developing custom chips to reduce reliance on Nvidia and address AI computing capacity shortages. The anticipation of this chip has already positively impacted Google's stock, reflecting investor confidence in the company's strategic direction.
TechCrunch AI