on Page 332 27% 19% 8/23 4/25 % of San Francisco Gross Bookings 34% 0% East Asia & Pacific Sub-Saharan Africa South Asia North America Middle East & North Africa Latin America & Caribbean Europe Years, per Epoch AI Note: In AI language models, tokens represent basic units of text (e.g., words or sub-words) used during training. Training dataset sizes are often measured in total tokens processed. A burn. *Cost-per-token = The expense incurred for processing or generating a single token (a word, sub-word, or character) during the operation of a language model. It is a key metric used to evaluate

Public mailing lists and IRC channel ● Internal Jira project restricted to Linaro members ● Three sub-projects: ○ Arm Compute Library ○ Arm NN ○ Android NN Driver ● Arm Compute Library has been integrated

Tensorization 下 CUDA CodeGen *。IR passes to automatically transform sub-tree to TensorCore Intrinsics Pattern Matching 计算平台事业部 shared/global

runtime for Relay ● Dynamic codegen (WIP) ○ Kernel dispatch for a single op ○ Graph dispatch for a (sub-)graph In collaboration with Jared Roesch, Zhi Chen, Wei Chen© 2019, Amazon Web Services, Inc. or