[Feature Request] Add Gated Delta Net (GDN) support

Is your feature request related to a problem? Please describe.

Megatron Core already has a Gated Delta Net (GDN) layer, but the current implementation depends on external Triton-based fla gated-delta-rule kernels and causal_conv1d. This means GDN does not benefit from native TE kernel performance, and users need extra dependencies.

The current training kernel path is also not performing well in practice, and the gap is noticeable on GB200.

In addition, the current path still has functional gaps: inference is not supported and packed sequences are not supported.

Describe the solution you'd like

Add native TE kernels for Gated Delta Net (GDN), with Megatron Core integration.

Ideally this would include:

• an optimized kernel path for the chunked gated delta rule
• substantially better training performance, especially on GB200
• compatibility with TE mixed-precision flows, especially BF16 and FP8

Describe alternatives you've considered

Using the current Triton-based fla.ops.gated_delta_rule + causal_conv1d implementation in Megatron Core.

Start by locating the existing GDN layer and its fla.ops.gated_delta_rule and causal_conv1d path. Define the native TE kernel scope for chunked training, inference, packed sequences, BF16, and FP8, then benchmark training performance on GB200. Done means Megatron Core uses the native path without those external dependencies and covers the requested functionality.