*   GPU
    *   GPU acceleration now supports quantized models by default
*   `DynamicBuffer::AddJoinedString()` will now add a separator if the first
    string to be joined is empty.
*   Adds support for cumulative sum (cumsum), both as builtin op and MLIR
    conversion.

### `TensorRT`

*   Issues a warning when the `session_config` parameter for the TF1 converter

"crypto/internal/fips140" "crypto/internal/fips140deps/byteorder" "errors" "hash" ) const ( // size512 is the size, in bytes, of a SHA-512 checksum. size512 = 64 // size224 is the size, in bytes, of a SHA-512/224 checksum. size224 = 28 // size256 is the size, in bytes, of a SHA-512/256 checksum. size256 = 32 // size384 is the size, in bytes, of a SHA-384 checksum. size384 = 48 // blockSize is the block size, in bytes, of the SHA-512/224, // SHA-512/256, SHA-384 and SHA-512 hash functions. blockSize...

import ( "crypto/internal/fips140" "crypto/internal/fips140deps/byteorder" "errors" ) const ( // size512 is the size, in bytes, of a SHA-512 checksum. size512 = 64 // size224 is the size, in bytes, of a SHA-512/224 checksum. size224 = 28 // size256 is the size, in bytes, of a SHA-512/256 checksum. size256 = 32 // size384 is the size, in bytes, of a SHA-384 checksum. size384 = 48 // blockSize is the block size, in bytes, of the SHA-512/224, // SHA-512/256, SHA-384 and SHA-512 hash functions. blockSize...

  network plugins which was first widely noticed in 1.16 (though some users saw it
  earlier than that, possibly only with specific network plugins). If you were previously
  using ethtool to disable checksum offload on your primary network interface, you should
  now be able to stop doing that. ([#92035](https://github.com/kubernetes/kubernetes/pull/92035), [@danwinship](https://github.com/danwinship)) [SIG Network and Node]