Huawei NetEngine: six routers, and a cage speed list that is the whole product #1184

Merged
jaylouisw merged 1 commit from feat/huawei-netengine into main 2026-11-09 03:41:05 +00:00
Owner

Clears the fixed-configuration NetEngine requests: 8000 M1A, M1C, M1D and the AR8140 / AR8140-T / AR8700 SD-WAN routers.

model fixed ports switching
8000 M1A 4 × 10GE SFP+ + 12 × GE 352 Gbps, 72 Mpps
8000 M1C same ports as the M1A 344 Gbps, 78 Mpps
8000 M1D 4 × 100GE/50GE QSFP28 + 16 × 25GE/10GE/GE 1.76 Tbps, 398 Mpps
AR8140-12G10XG 10 × 10GE SFP+ + 8 × GE Combo + 4 × GE + 4 SIC 25 Gbps NAT+ACL+QoS
AR8140-T-12G10XG identical ports, TPM variant same
AR8700-8 1 × 40GE QSFP+ + 8 × 10GE SFP+ + 8 × GE Combo + 8 SIC 30 Gbps

Three things recorded rather than averaged away

  1. The M1D's cage speeds are the product. Each 100GE cage falls back to 50/40/25/10G and each 25GE cage to 10G or 1G — the same 400G access router becomes a 16-port 10G router on optics alone, with no licence or part-number change.
  2. M1A and M1C have the same fixed interface (410GE + 12GE); the difference is capacity and power (352 vs 344 Gbps, 74.8 vs 89 W). Swapping one for the other changes nothing about the links.
  3. AR8140-T is a TPM, not a port layout. Huawei's specification column covers both AR8140 variants together — separate entries are right because the part numbers differ, and the note says why the ports don't.

Deliberately still open

The modular chassis (X4/X8/X16, M4/M6/M8/M14, F8), the M1D-B (configurable port modes: 2x50GE+2x25GE+2x10GE+14xGE or 2x50GE+6x10GE+14xGE) and the F1A (port groups that split one 40/50GE cage into four 10GE/25GE). Their port counts depend on a configuration, not on a datasheet line, and Huawei's support pages that answer it are behind a login wall — that's the blocker, not the research.

Catalogue 1,968 → 1,974 models (Huawei 64 → 70). Suite: 114 suites, 5334 assertions, 0 failed.

Clears the fixed-configuration NetEngine requests: 8000 M1A, M1C, M1D and the AR8140 / AR8140-T / AR8700 SD-WAN routers. | model | fixed ports | switching | |---|---|---| | 8000 M1A | 4 × 10GE SFP+ + 12 × GE | 352 Gbps, 72 Mpps | | 8000 M1C | **same ports as the M1A** | 344 Gbps, 78 Mpps | | 8000 M1D | 4 × 100GE/50GE QSFP28 + 16 × 25GE/10GE/GE | 1.76 Tbps, 398 Mpps | | AR8140-12G10XG | 10 × 10GE SFP+ + 8 × GE Combo + 4 × GE + 4 SIC | 25 Gbps NAT+ACL+QoS | | AR8140-T-12G10XG | **identical ports**, TPM variant | same | | AR8700-8 | 1 × 40GE QSFP+ + 8 × 10GE SFP+ + 8 × GE Combo + 8 SIC | 30 Gbps | ### Three things recorded rather than averaged away 1. **The M1D's cage speeds are the product.** Each 100GE cage falls back to 50/40/25/10G and each 25GE cage to 10G or 1G — the same 400G access router becomes a 16-port 10G router on optics alone, with no licence or part-number change. 2. **M1A and M1C have the same fixed interface** (4*10GE + 12*GE); the difference is capacity and power (352 vs 344 Gbps, 74.8 vs 89 W). Swapping one for the other changes nothing about the links. 3. **AR8140-T is a TPM, not a port layout.** Huawei's specification column covers both AR8140 variants together — separate entries are right because the part numbers differ, and the note says why the ports don't. ### Deliberately still open The **modular chassis** (X4/X8/X16, M4/M6/M8/M14, F8), the **M1D-B** (configurable port modes: `2x50GE+2x25GE+2x10GE+14xGE` or `2x50GE+6x10GE+14xGE`) and the **F1A** (port groups that split one 40/50GE cage into four 10GE/25GE). Their port counts depend on a configuration, not on a datasheet line, and Huawei's support pages that answer it are behind a login wall — that's the blocker, not the research. Catalogue 1,968 → **1,974 models** (Huawei 64 → 70). Suite: 114 suites, 5334 assertions, 0 failed.
Huawei NetEngine: six routers, and a cage speed list that is the whole product
Some checks failed
CI / Test (Node 18) (pull_request) Blocked by required conditions
CI / Test (Node 20) (pull_request) Blocked by required conditions
CI / Test (Node 22) (pull_request) Blocked by required conditions
CI / Server smoke test (pull_request) Blocked by required conditions
CI / Docker build (pull_request) Blocked by required conditions
CI / Classify the change (pull_request) Has started running
PR rebase / Rebase onto main and arm auto-merge (pull_request) Failing after 19s
cc12793ad1
Clears the fixed-configuration NetEngine requests: 8000 M1A, M1C, M1D and the AR8140 /
AR8140-T / AR8700 SD-WAN routers.

| model | fixed ports | switching |
|---|---|---|
| 8000 M1A | 4 x 10GE SFP+ + 12 x GE | 352 Gbps, 72 Mpps |
| 8000 M1C | **same ports as the M1A** | 344 Gbps, 78 Mpps |
| 8000 M1D | 4 x 100GE/50GE QSFP28 + 16 x 25GE/10GE/GE | 1.76 Tbps, 398 Mpps |
| AR8140-12G10XG | 10 x 10GE SFP+ + 8 x GE Combo + 4 x GE + 4 SIC | 25 Gbps NAT+ACL+QoS |
| AR8140-T-12G10XG | **identical ports**, TPM variant | same |
| AR8700-8 | 1 x 40GE QSFP+ + 8 x 10GE SFP+ + 8 x GE Combo + 8 SIC | 30 Gbps |

Three things the entries record rather than average away:

- **The M1D's cage speeds are the product.** Each 100GE cage falls back to 50/40/25/10G and each
  25GE cage to 10G or 1G, so the same 400G access router becomes a 16-port 10G router on optics
  alone - no licence change, no different part number.
- **M1A and M1C have the same fixed interface** (4*10GE + 12*GE). The difference is capacity and
  power (352 vs 344 Gbps, 74.8 vs 89 W), so swapping one for the other in a diagram changes
  nothing about the links.
- **AR8140-T is a TPM, not a port layout.** Huawei's own specification column covers both
  AR8140 variants together; a separate entry is right because the part numbers differ, and the
  note says exactly why the ports do not.

Still open, deliberately: the modular chassis (X4/X8/X16, M4/M6/M8/M14, F8), the **M1D-B**
(configurable port modes: `2x50GE+2x25GE+2x10GE+14xGE` or `2x50GE+6x10GE+14xGE`) and the **F1A**
(port groups that split one 40/50GE cage into four 10GE/25GE) - their port counts depend on a
configuration, not on a datasheet line. Huawei's support pages answer those and are behind a login
wall, which is the blocker, not the research.

Catalogue 1,968 -> 1,974 models (Huawei 64 -> 70). Suite: 114 suites, 5334 assertions, 0 failed.
Author
Owner

The auto-rebase job failed. This is a problem with the automation, not with this change.

the rebase job itself failed before it could report a reason - check the runner.

The run is at the Actions tab. Nothing was pushed, so the branch is exactly as you wrote it.

**The auto-rebase job failed.** This is a problem with the automation, not with this change. ``` the rebase job itself failed before it could report a reason - check the runner. ``` The run is at the Actions tab. Nothing was pushed, so the branch is exactly as you wrote it.
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
jaylouisw/packetforge!1184
No description provided.