Daily Edition · Page 3

THE DAILY INDEX

Technology, AI, software, security, crypto and business

HEADLINES & LAUNCHES

TLDR AI

FIRST MLPERF BENCHMARKS FOR NVIDIA BLACKWELL, AMD, GOOGLE, UNTETHER AI

Nvidia's new Blackwell chip demonstrated top per GPU performance in MLPerf's LLM Q&A benchmark, showcasing significant advancements with its 4-bit floating-point precision. However, competitors like Untether AI and AMD also showed promising results, particularly in energy efficiency. Untether AI's speedAI240 chip, for instance, excelled in the edge-closed category, highlighting diverse strengths across new AI inference hardware.

CAN AI SCALING CONTINUE THROUGH 2030?

AI training is growing at an unprecedented 4x per year, outpacing mobile adoption and genome sequencing rates from past technological expansions. Research indicates scaling AI training could feasibly continue until 2030, constrained primarily by power availability and chip manufacturing capacity. Training runs up to 2e29 FLOP may become possible, marking substantial progress akin to the leap from GPT-2 to GPT-4, provided hundreds of billions are invested. Data scarcity and latency also pose challenges, but these may be surmountable through multimodal and synthetic data generation strategies and advanced network topologies.

MEET BOARDWALK ROBOTICS' ADDITION TO HUMANOID ROBOTS

Boardwalk Robotics has unveiled Alex, a humanoid upper torso robot designed for tasks in manufacturing, logistics, and maintenance. Developed independently while leveraging the pedigree of IHMC's bipedal robot expertise, Alex is legless, with its creators emphasizing manipulation over mobility to ensure safety and efficiency. Commercial partners are currently being selected for pilots, while researchers can purchase Alex immediately.

GPU UTILIZATION IS A MISLEADING METRIC

While commonly monitored, GPU utilization may not accurately reflect true GPU performance in ML workloads, as it doesn't account for whether the GPU's computational power is fully leveraged. Trainy discovered this while only achieving ~20% model FLOPS utilization (MFU) despite 100% GPU utilization during LLM training. It recommends tracking SM efficiency for a better performance indicator and achieving a 4x speedup in training time by optimizing with fused kernels and the right model parallelism level.

TLDR IndexPage 3