11. Release Notes

Software Manifest

TI_MCU_2.1.1

  • TI_MCU_2.1.1 User’s Guide

  • Supported devices

    • F28P55x and C28x core based devices

    • F29H85x and C29x core based devices

    • AM13x and Arm Cortex-M33 core based devices

    • AM26x and Arm Cortex-R5 core based devices

    • MSPM0G5187 and Arm Cortex-M0+ core based devices

    • MSPM33C321A and Arm Cortex-M33 core based devices

    • CC2745 device and Arm Cortex-M33 core based devices

  • Added a model memory usage summary and layer offloading summary to compiler output.

  • Optimized NPU setup time for the F28P55x device, resulting in improved overall performance.

  • Updated the “skip_normalize” option to extract input normalization scale values as floats instead of integers.

  • Resolved defects

    • CODEGEN-14807: TI MCU NNC crashes when NHWC only padding optimization is applied to NCHW data layout

    • CODEGEN-14838: TI MCU NNC crashes when a non-existent attribute is accessed on a node

    • CODEGEN-14981: TVM ONNX frontend fails with ONNX 1.20

TI_MCU_2.1.0

  • TI_MCU_2.1.0 User’s Guide

  • Supported devices

  • Supported all layer configs with 8-bit activations and 8-/4-/2-bit weights that can be offloaded to TI-NPU

  • Supported all layer configs with 8-bit activations and 8-bit weights that can be accelerated using the M33 Custom Datapath Extension (CDE).

  • Improved quantized (8-bit) model inference performance on C29x core. Note that floating-point models typically run faster than their quantized counterparts on C29x core, despite requiring a larger memory footprint.

TI_MCU_2.0.0

  • TI_MCU_2.0.0 User’s Guide

  • Supported devices

  • Expanded layer configs that can be supported on TI-NPU

  • Added an option to compress TI-NPU layer data

  • Optimized floating point inference on C29x core

  • Supported quantized models in QDQ format with integer inference code on CPU

TI_MCU_1.3.0

  • TI_MCU_1.3.0 User’s Guide

  • Windows and Linux support

  • Multiple model support in the same application

  • Generic padding support

  • Optimized performance for motor fault classification models and arc fault detection models

    • With TI-NPU acceleration, these models saw a 1.2x to 1.5x speedup from TI_MCU_1.2.0 release

    • With software-only execution on C28, these models saw a 3.9x to 4.8x speedup from TI_MCU_1.2.0 release

  • Added example and documentation on how to apply TI-NPU quantization to an existing PyTorch training script

  • Simplified command line options

TI_MCU_1.2.0