Final fresh8group model/ELF and2assertion nominal-clock checks. No runtime or GUI measurements.

verify_model.py
{
  "scope": "Array arithmetic and offline inspection of actual linked ELF; no DMA run or cycle measurement",
  "words": 256,
  "bytes_per_copy": 1024,
  "half_transfer_bytes": 512,
  "expected_sum32": 1460172672,
  "foreground_LCG1024_expected": 805413889,
  "MEM2MEM_with_Circular_allowed": false,
  "checks_passed": [
    "word/byte count",
    "half transfer size",
    "copy content expectation",
    "little-endian word layout",
    "uint32 cycle wrap",
    "foreground workload expected result",
    "actual ELF arrays in aligned SRAM notCCM",
    "actual ELF source/destinations do not overlap"
  ],
  "ELF": {
    "gebrabit_lesson14_dma_v2/Debug/gebrabit_lesson14_dma_v2.elf": {
      "cpu_destination": {
        "address": "0x2000042c",
        "bytes": 1024,
        "alignment4": true,
        "in_main_SRAM_not_CCM": true
      },
      "dma_destination": {
        "address": "0x2000082c",
        "bytes": 1024,
        "alignment4": true,
        "in_main_SRAM_not_CCM": true
      },
      "dma_source": {
        "address": "0x2000002c",
        "bytes": 1024,
        "alignment4": true,
        "in_main_SRAM_not_CCM": true
      }
    }
  },
  "measured_cycles": null
}

calculate_clock.py
{
  "scope": "Nominal clock/element arithmetic, not runtime timing evidence.",
  "configured_HCLK_hz": 72000000,
  "nominal_cycle_ns": 13.88888888888889,
  "words": 256,
  "bytes_per_word": 4,
  "transfer_bytes": 1024,
  "half_transfer_words": 128,
  "half_transfer_bytes": 512,
  "measured_DMA_cycles": null,
  "measured_memcpy_cycles": null
}
