[#8542][feat] AutoDeploy: add DeepSeek-R1 FP8 perf-sanity test on 8x B200 post merge

[MrGeva]

Mirrors super_ad_blackwell-super_ad_ws4_1k1k but for DeepSeek-R1 0528 FP8 on a full DGX B200 (8 GPUs).

Changes:

• tests/scripts/perf-sanity/aggregated/deepseek_r1_fp8_ad_blackwell.yaml: new perf-sanity config with one server config r1_fp8_ad_ws8_1k1k using the _autodeploy backend, world_size: 8, and the existing examples/auto_deploy/model_registry/configs/deepseek-r1.yaml (MLA + trtllm_mla cached attention + fuse_rope_into_trtllm_mla + multi-stream MoE + AllReduce-residual-RMSNorm fusion + sharding). Client config matches the reference: concurrency 64, 10 iterations, ISL=OSL=1024, openai backend.
• tests/integration/test_lists/test-db/l0_dgx_b200.yml: enroll the new test in the existing 8-GPU stage: post_merge / backend: autodeploy block alongside the existing DeepSeek-R1-0528 accuracy registry test.

The deepseek_r1_0528_fp8 -> DeepSeek-R1/DeepSeek-R1-0528/ mapping already exists in test_perf_sanity.py::MODEL_PATH_DICT, so no edit to the test driver is required.

Test ID:
perf/test_perf_sanity.py::test_e2e[aggr_upload-deepseek_r1_fp8_ad_blackwell-r1_fp8_ad_ws8_1k1k]

@coderabbitai summary

Description

Test Coverage

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[MrGeva]

/bot run --extra-stage "H100_PCIe-AutoDeploy-1" --disable-fail-fast

[MrGeva]

/bot run --extra-stage "DGX_B200-8_GPUs-AutoDeploy-Post-Merge-1" --disable-fail-fast

[tensorrt-cicd]

PR_Github #47935 [ run ] triggered by Bot. Commit: 1320deb Link to invocation

[tensorrt-cicd]

PR_Github #47937 [ run ] triggered by Bot. Commit: 1320deb Link to invocation