Live DB — analyst tables
Test
Amazon 131072 P5en-H100 NVL8 Nitro-v5 Hyperscaler EFA Cluster Configuration
AWS-P5en-H100 NVL8-EFA-WB
Amazon · H100 · Nitro-v5
Rack type
P5en NVL8
Cluster / location
P5en-H100 NVL8 Clusters
Buyer profile
Hyperscaler
Networking type
EFA
Number of XPUs
131,072
NIC used
Nitro-v5
NICs / XPU
2
Scale-out BW / XPU
400G
XPU / server
8
Scale-up network
1 rows
Assigned blockDefault
H100 - NVL8
Scale-UpXPU: H100 NVL81 lineshared by 17 SKUs
Scale-up Units are based on the scale-up world size (8 XPUs).
XPU World Size
2
WS| Scale-Up Network Layer | Part / Model | Units# | Unit BWGbit/s uni-d | Port Count# | Port BWGbit/s uni-d |
|---|---|---|---|---|---|
| XPUs per Node | · | 8 | 3,600 | · | · |
| Scale-Up World Size | · | 8 | 28,800 | · | · |
Switches
5
SW| Scale-Up Network Layer | Part / Model | Units# | Unit BWGbit/s uni-d | Port Count# | Port BWGbit/s uni-d |
|---|---|---|---|---|---|
| Scale-Up Switch A | Nvidia NVSwitch 3.0 (NVLink 4) | 1 | 8,000 | 20 | 400 |
| Scale-Up Switch B | Nvidia NVSwitch 3.0 (NVLink 4) | 1 | 6,400 | 16 | 400 |
| Scale-Up Switch C | Nvidia NVSwitch 3.0 (NVLink 4) | 1 | 6,400 | 16 | 400 |
| Scale-Up Switch D | Nvidia NVSwitch 3.0 (NVLink 4) | 1 | 8,000 | 20 | 400 |
| Scale-Up Switch Total | Nvidia NVSwitch 3.0 (NVLink 4) | 4 | 28,800 | 72 | · |
Scale-Up Parts Summary
1
| Item | Qty/Cluster | Qty/Server | Attach/XPU |
|---|---|---|---|
| Nvidia NVSwitch 3.0 (NVLink 4) | 65,536 | 4.00 | 0.50 |
Backend network
5 rows
Assigned blockDefault
H100 - P5en NVL8 - EFAv3 - 3L - ETH - TH3 - AWS
BackendXPU: H100 P5en NVL85 linesshared by 1 SKU
Backend Units are based on the cluster size (131,072 XPUs).
Switches
5
SW Network Layer | Switch and Server Spec | Aggregate Port Utilization | Network Bandwidth Aggregate | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
SwitchPart Name / Model | Units# | Unit BWGbit/s | Port BWGbit/s | Downlink Ports / SW# | Uplink Ports / SW# | Max Ports# | Downlink Ports# | Uplink Ports# | Total Ports# | Downlink BWTbit/s | Uplink BWTbit/s | OversubRatio | |
| XPUs | · | 131,072 | 400 | 400 | 0 | 131,072 | 131,072 | 0 | 131,072 | 131,072 | 0 | 52,428.8 | · |
| NICs per Node | EFAv3 Nitro-v5 200G NIC | 16 | 200 | 400 | · | · | 8 | · | · | · | · | · | · |
| ToR | WhiteBox Tomahawk 3 12.8T | 8,192 | 12,800 | 400 | 16 | 16 | 32 | 131,072 | 131,072 | 262,144 | 52,428.8 | 52,428.8 | 1 |
| Leaf | WhiteBox Tomahawk 3 12.8T | 8,192 | 12,800 | 400 | 16 | 16 | 32 | 131,072 | 131,072 | 262,144 | 52,428.8 | 52,428.8 | 1 |
| Spine | WhiteBox Tomahawk 3 12.8T | 4,096 | 12,800 | 400 | 32 | 0 | 32 | 131,072 | 0 | 131,072 | 52,428.8 | 0 | · |
Transceivers
8
CN Network Layer | Transceiver / CablePart Name / Model | Total Units# | Units per SW/GPU# | Unit BWGbit/s | Lanes# | BW per LaneGbit/s | Agg. BWTbit/s |
|---|---|---|---|---|---|---|---|
| GPU Nodes | None - AECs are Used -> | 131,072 | 131,072 | 400 | 4 | 100 | 52,428.8 |
| CPU Nodes | None - AECs are Used -> | 32,768 | · | · | · | · | · |
| Leaf | |||||||
| Downlink | Generic 400G DR4 | 131,072 | 16 | 400 | 4 | 100 | 52,428.8 |
| Uplink | Generic 400G DR4 | 131,072 | #ERR | 400 | #ERR | 100 | 0 |
| Spine | |||||||
| Downlink | Generic 400G DR4 | 131,072 | #ERR | 400 | #ERR | 100 | 0 |
| Uplink | · | · | · | · | · | · | · |
| Core | |||||||
| Downlink | · | · | · | · | · | · | · |
| Uplink | · | · | · | · | · | · | · |
| Super-Core | |||||||
| ToR | |||||||
| Downlink | None - AECs are Used -> | 131,072 | 16 | 400 | 4 | 100 | 52,428.8 |
| Uplink | Generic 400G DR4 | 131,072 | 16 | 400 | 4 | 100 | 52,428.8 |
Cables
8
CB Network Layer | Cable/ FiberType | Total Units# | Units per Transceiver# | Unit BWGbit/s | Lane Count# | SerDes BWGbit/s |
|---|---|---|---|---|---|---|
| GPU Nodes | Generic 400G Y Cable AEC | 131,072 | 1 | 400 | 4 | 100 |
| CPU Nodes | · | · | · | · | · | · |
| Leaf | ||||||
| Downlink | Already counted above | · | · | · | · | · |
| Uplink | Generic 400G Single Mode | 131,072 | 1 | 400 | 4 | 100 |
| Spine | ||||||
| Downlink | · | · | · | · | · | · |
| Uplink | · | · | · | · | · | · |
| Core | ||||||
| Downlink | · | · | · | · | · | · |
| Uplink | · | · | · | · | · | · |
| Super-Core | ||||||
| ToR | ||||||
| Downlink | Already counted above | · | · | · | · | · |
| Uplink | Generic 400G Single Mode | 131,072 | 1 | 400 | 4 | 100 |
Backend Parts Summary
5
| Item | Qty/Cluster | Qty/Server | Attach/XPU |
|---|---|---|---|
| Generic 400G Y Cable AEC | 131,072 | 8.00 | 1.00 |
| EFAv3 Nitro-v5 200G NIC | 262,144 | 16.00 | 2.00 |
| Generic 400G DR4 | 524,288 | 32.00 | 4.00 |
| Generic 400G Single Mode | 262,144 | 16.00 | 2.00 |
| WhiteBox Tomahawk 3 12.8T | 20,480 | 1.25 | 0.16 |
Frontend network
No converged frontend + storage fabric data for this cluster.
Frontend Parts Summary
0
| Item | Qty/Cluster | Qty/Server | Attach/XPU |
|---|
Compute Tray Networking Content
2 rows
Networking components inside the compute tray — accelerators, board-to-board connectors, retimers, and their per-tray bandwidth.
Assigned blockDefault
H100 - P5en NVL8 - AWS
Compute TrayXPU: H100 P5en NVL82 linesshared by 1 SKU
| Tray layer | Part / Model | Units per Tray | Unit BWGb/s | Lanes | Lane BWGb/s | Per switch |
|---|---|---|---|---|---|---|
| XPUs per Tray | Blackwell 200 | 8 | · | · | · | · |
| Compute Tray | Broadcom PEX89144 Gen 5.0 144L | 4 | · | · | · | · |
| GPU to CPU Tray | Generic PCIe Gen5 x16 2-meter DAC | 8 | · | · | · | · |
Compute Tray Parts Summary
2
| Item | Qty/Cluster | Qty/Server | Attach/XPU |
|---|---|---|---|
| Broadcom PEX89144 Gen 5.0 144L | 65,536 | 4.00 | 0.50 |
| Generic PCIe Gen5 x16 2-meter DAC | 131,072 | 8.00 | 1.00 |