November 2022 Summaries
2 posts from Modular
Filter
Month:
Year:
Post Summaries
Back to Blog
In a two-part series, Modular details improvements made to the MLIR (Multi-Level Intermediate Representation) compilation stack, focusing on the development of large models. Key enhancements include the introduction of a "Resources" mechanism to manage constant data like weights more efficiently, separating them from MLIR attributes to improve memory management and development velocity. The team also developed a new binary encoding format for MLIR to replace the previous textual format, significantly increasing the speed of serialization and reducing memory usage during model compilation. These changes resulted in faster model compilation and serialization processes, improving both productivity and execution speed, particularly when compiling models from TensorFlow to runtime input formats. The advancements have been contributed to the LLVM/MLIR repository, benefiting the broader AI ecosystem, and Modular has become a Platinum Sponsor of LLVM, emphasizing its commitment to supporting the growth of the LLVM community.
Nov 10, 2022
1,658 words in the original blog post.
In November 2022, the economic landscape is marked by significant layoffs and the collapse of companies that overextended financially without solid foundations, leading to a call for businesses to revert to traditional principles of value creation and sustainable growth. Modular, a company born in a down economy, emphasizes the importance of building businesses on first-principle concepts rather than relying on hype, focusing on creating value for customers, employees, and investors. The company sees a massive opportunity in AI, where current software fails due to its development as a by-product of AI-first companies prioritizing internal needs over external customer satisfaction. Modular aims to address this by building a cohesive team of experts in AI infrastructure, emphasizing a modern work culture and offering competitive salaries to drive technology forward. The company is preparing to reveal advancements in AI pipelines, data, models, and developer tools in 2023, promising significant changes in hardware and accelerated computing, and invites interested individuals to stay informed through their newsletter and social media channels.
Nov 08, 2022
819 words in the original blog post.