Field-by-Field Decode

The complete wire schema of the Megascale fleet metadata, recovered from the protodesc_cold FileDescriptorProto pool inside libtpu.so. All messages are package xla.megascale.runtime unless noted; the topology-args family is package tpu.

SliceInfo

platforms/xla/megascale/common/runtime.proto — the per-slice fleet entry.

message SliceInfo {
  int32                    slice_id          = 1;   // fleet-wide slice index
  // field 2 retired (gap)
  tpu.TpuTopologyArgsProto tpu_topology_args = 3;   // slice 3D shape
}

This is the canonical SliceInfo: slice index plus shape, with host endpoints held separately in the NetworkAddressMapping list.

MultiSliceTopologyAndLocationProto

runtime.proto — the serialized FLEET object (the wire form of the MultiSliceTopologyAndLocation class).

message MultiSliceTopologyAndLocationProto {
  int32              local_slice_id = 1;   // THIS process's slice
  int32              local_host_id  = 2;   // THIS process's host (in slice)
  repeated SliceInfo slice_info     = 3;   // every slice in the fleet
  int64              incarnation_id = 4;   // fleet generation token
}

Accepted by Communicator::Create, MegaScaleMultiSliceConfig::Create, and AotMegaScaleMultiSliceConfig::Create. The local_* fields make it self-locating.

MultiSliceTopologyInfo

platforms/xla/megascale/runtime/communication/transport.proto — the wire payload the coordinator serializes into the bootstrap response.

message MultiSliceTopologyInfo {
  repeated SliceInfo             slice_info       = 1;   // all slices
  repeated NetworkAddressMapping address_mappings = 2;   // all host endpoints
  int64                          incarnation_id   = 3;
}

Pairs the slice inventory with the full endpoint table. Serialized into GetMultiSliceTopologyResponse.serialized_topology_info.

NetworkAddressMapping / HostNetworkAddress / EndpointAddresses

third_party/tensorflow/compiler/xla/megascale/addresses.proto — the DCN reachability layer.

message HostNetworkAddress {
  string address                 = 1;   // ip:port / host:port of a NIC
  string interface_name          = 2;   // NIC name (DCN port)
  int32  numa_node               = 4;   // NUMA affinity of this NIC
  string host_name_for_debugging = 3;   // human-readable host label
}
message NetworkAddressMapping {
  int32  slice_id                = 1;   // owning slice
  int32  host_id                 = 2;   // host within slice
  repeated HostNetworkAddress addresses = 3;   // one per DCN interface
}
message EndpointAddresses {
  repeated NetworkAddressMapping address_mappings = 1;   // whole fleet
}

(slice_id, host_id) is the universal host key. Note the non-contiguous field numbers in HostNetworkAddress (1, 2, 4, 3) — a retired field left a gap.

GetMultiSliceTopologyRequest / Response

transport.proto — the bootstrap RPC payloads.

message GetMultiSliceTopologyRequest {
  NetworkAddressMapping    address_mapping   = 1;   // THIS host's identity + endpoints
  tpu.TpuTopologyArgsProto tpu_topology_args = 2;   // THIS slice's shape
  int64                    incarnation_id    = 3;   // THIS process's generation
}
message GetMultiSliceTopologyResponse {
  bytes serialized_topology_info = 1;   // serialized MultiSliceTopologyInfo
}

The request's slice_id / host_id live inside address_mapping; the response is a single opaque bytes blob.

TpuTopologyArgsProto

learning/45eac/tpu/runtime/topology/tpu_topology_args.proto (package tpu, editions) — the slice-shape descriptor. Full decode on the Slice Shape page.

message TpuTopologyArgsProto {
  tpu.TpuVersionProto      version                  = 1;
  string                   variant                  = 2;
  tpu.TpuPlatformTypeProto platform_type            = 3;
  string                   chip_config_name         = 4;
  tpu.TpuDimensionsProto   chips_per_host_bounds    = 5;
  tpu.TpuDimensionsProto   host_bounds              = 6;
  tpu.TpuWrapProto         wrap                     = 7;
  bool                     twist                    = 8;
  bool                     enhanced_barrier_enabled = 9;
  bool                     use_continuations        = 11;
  tpu.TpuRoutingStrategyProto routing_strategy       = 12;
  SubSlice                 sub_slice                = 10;
  message SubSlice {
    tpu.TpuDimensionsProto chips_per_host_bounds    = 1;
    tpu.TpuDimensionsProto host_bounds              = 2;
  }
}

DCNTopology

third_party/tensorflow/compiler/xla/megascale/dcn_topology.proto — the cross-slice reduction plan.

message DCNTopology {
  oneof representation {                       // "representation"
    SymmetricTree symmetric_tree = 1;
    TreeNode      tree           = 2;
  }
  message SymmetricTree {
    repeated int32 branching_per_layer = 1;    // balanced tree fan-out per layer
  }
  message TreeNode {
    string            label             = 1;   // optional node label
    repeated TreeNode nodes             = 2;   // children
    repeated SliceRange slice_ranges    = 3;   // leaf slice ranges
    float             egress_constraint = 4;   // per-node egress cap
    bool              ring_transfers    = 5;   // ring vs tree at this node
    message SliceRange { int32 slice_id_start = 1; int32 slice_id_end = 2; }
  }
}

SymmetricTree is the compact form; TreeNode the explicit form whose leaves name contiguous [slice_id_start, slice_id_end] ranges.

TpuCoreLocationProto

platforms/xla/megascale/runtime/common/core_location.proto — the per-core handle (coupled with its slice by the MultiSliceTpuCoreLocation class).

message TpuCoreLocationProto {
  tpu.TpuCoreTypeProto core_type         = 1;   // TENSOR / BARNA / SPARSE
  int32                logical_device_id = 2;   // device id within its slice
}
enum tpu.TpuCoreTypeProto {
  TPU_CORE_TYPE_INVALID=0; TPU_CORE_TYPE_TENSOR_CORE=1;
  TPU_CORE_TYPE_BARNA_CORE=2; TPU_CORE_TYPE_SPARSE_CORE=3;
}

PerSliceTpuDimensionsProto / Topology

runtime.proto — a lightweight per-slice (x,y,z) view used where the full shape is unnecessary (e.g. the MEGASCALE_TOPOLOGY override path).

message PerSliceTpuDimensionsProto {
  int32 slice_id = 1; int32 x = 2; int32 y = 3; int32 z = 4;
}
message Topology { repeated PerSliceTpuDimensionsProto slices = 1; }

Supporting dimension and enum types

message TpuDimensionsProto   { int32 x=1; int32 y=2; int32 z=3; int32 w=4; }
message TpuWrapProto         { bool  x=1; bool  y=2; bool  z=3; }
message TpuDegradedAxesProto { bool  x=1; bool  y=2; bool  z=3; }

enum TpuRoutingStrategyProto { ROUTING_DEFAULT=0; ROUTING_MESH=1; ROUTING_NHOP=2; }
enum TpuVersionProto {
  TPU_VERSION_INVALID=0; TPU_VERSION_JELLYFISH=1; TPU_VERSION_DRAGONFISH=2;
  TPU_VERSION_PUFFERFISH=3; TPU_VERSION_VIPERFISH=4; TPU_VERSION_GHOSTLITE=5;
  TPU_VERSION_6acc60406=6;
}
enum TpuPlatformTypeProto {
  TPU_PLATFORM_TYPE_INVALID=0; TPU_PLATFORM_TYPE_HARDWARE=1;
  TPU_PLATFORM_TYPE_GRM=2; TPU_PLATFORM_TYPE_ISS=3;
}
message TpuConfiguredPropertiesProto {
  tpu.TpuDegradedAxesProto    degraded_axes           = 1;
  bool                        is_nhop_source_relative = 2;
  tpu.TpuRoutingStrategyProto routing_strategy        = 3;
}

Runtime options and collective slack (auxiliary)

runtime.proto carries two more fleet-scoped messages:

message MegascaleRuntimeOptions {                 // all fields optional (oneof presence)
  bool abort_on_errors      = 1;
  bool abort_on_hangs       = 2;
  bool debug_dump_on_error  = 3;
}
message CollectiveSlackInformation {
  repeated BudgetInfo     default_budgets         = 1;
  repeated CollectiveInfo collectives             = 2;
  CollectiveInfo          default_collective_info = 3;
  message BudgetInfo {
    repeated google.protobuf.Duration transmission_budget_per_level = 1;
    google.protobuf.Duration          safety_margin                 = 2;
    bool                              add_random_delay              = 3;
    repeated int32                    levels_with_delay             = 4;
    repeated google.protobuf.Duration explicit_delays               = 5;
    int64                             min_size_for_delay            = 6;
  }
  message CollectiveInfo {
    string                   key                  = 1;
    google.protobuf.Duration slack                = 2;
    oneof budget {
      BudgetInfo             transmission_budget   = 3;
      int32                  default_budget_index  = 4;
    }
  }
}

MegascaleRuntimeOptions mirrors the --megascale_*_abort_on_* flags; CollectiveSlackInformation tunes per-collective DCN transmission budgets (per-tree-level durations) and is passed alongside the fleet view into the HostCommandSchedulerFactory.

Verified presence in libtpu.so

Cross-References

• Fleet Metadata › Overview — the model these wire fields populate
• Topology Model — how the decoded coordinates assemble into the ToroidalTopology
• Global Addressing — the global_core_id / global_chip_id derivations decoded here
• Slice Shape — the per-slice bounds fields and their versioning gate
• Bootstrap Exchange — the rendezvous that carries this blob across DCN
• ICI › Topology Discovery — the in-slice producer of the per-slice tpu_topology_args.proto