sccl: drive allreduce tests via torch.distributed; reorganize into tests/sccl/

Convert the multidevice allreduce correctness + latency/buffer-kind sweeps
to run through the real PyTorch-distributed path
(init_process_group(backend="ahbm") -> mp.spawn -> dist.all_reduce) instead
of direct ctx.launch, and reorganize the CCL/allreduce tests into a
tests/sccl/ package split one test per file.

Production change (required for the distributed path on non-square SIP grids):
- AhbmCCLBackend now reads explicit system.sips.w/h from the spec, with a
  square-only sqrt fallback that raises on ambiguity, instead of silently
  guessing round(sqrt(count)). This fixes the 2x3 / 3x2 torus + mesh cases,
  which previously resolved to a wrong 2x2 grid. Mirrors the test helper's
  _sip_topo_dims precedence (explicit w/h > square fallback > raise).

Test reorganization (tests/sccl/):
- _allreduce_helpers.py: shared plumbing (distributed driver, config writers,
  direct-launch run_allreduce parity reference, sweep/buffer-kind constants,
  plot aggregators, topology-diagram + FSIM-comparison emitters).
- test_allreduce_ring_torus_mesh.py: correctness across ring/torus/mesh.
- test_distributed_default_topology.py: full distributed path on topology.yaml.
- test_plot_latency_sweep.py / test_plot_buffer_kind_sweep.py: sweep rows.
- test_plot_topology_diagram.py / test_plot_comparison_fsim.py: plot emitters.
- test_intercube_root_center.py: moved in (ADR-0032 center-root latency guard).

Also:
- Move the FSIM comparison plot generator out of scripts/ into the sccl suite.
- Delete superseded test files (test_allreduce_multidevice,
  test_distributed_lrab_hierarchical_allreduce, test_allreduce_buffer_kind_sweep)
  and repoint conftest aggregators + the ipcq buffer-kind importers.
- Regenerate the allreduce_latency_plots derived artifacts from the full sweep.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

tests/test_ipcq_buffer_kind_latency.py

@@ -20,7 +20,7 @@ Reference (Phase 2 will edit these):
   - ccl.yaml                                    — algorithm.buffer_kind
 
 The tests reuse the existing config-driven allreduce app
-(``run_allreduce`` in tests/test_allreduce_multidevice.py) with a 2-SIP
+(``run_allreduce`` in tests/sccl/_allreduce_helpers.py) with a 2-SIP
 ring topology and a SMALL n_elem so they finish fast (~3-5 s each).
 """
 from __future__ import annotations
@@ -37,7 +37,7 @@ from kernbench.topology.builder import resolve_topology
 
 # Reuse the test app's helpers so this micro-test file does not
 # duplicate the run-allreduce + write-temp-configs plumbing.
-from tests.test_allreduce_multidevice import (
+from tests.sccl._allreduce_helpers import (
     _write_temp_configs,
     run_allreduce,
 )