N9500-128QC, 128x400G QSFP112 Ethernet L3 4U Managed Switch, 51.2Tbps, Broadcom Tomahawk 5, Support RoCEv2, for AI/ML/Cloud Data Center/HPC

#102323
Model: N9500-128QC
Sold: 0
In Stock: Available
$ 37399.00

Item Spotlights

  • 128-port 400G (QSFP112), maximum support 256-port 200G
  • Switching capacity 102.4T, packet forwarding rate 21000Mpps
  • Support RoCE configuration one-click deployment
  • Level 2 networking can support up to 8K 400G ports, and level 3 networking can support up to 32K 400G ports
  • Support dual-port access networking, GR perfect restart, BFD fast forwarding detection and other mechanisms
  • AI ECN automatic parameter adjustment, RoCE network key indicators real-time telemetry
  • Support IPv4/IPv6 dual-stack protocols: static routing, equal-cost routing, policy routing, RIPng, OSPFv3, BGP4+, etc.
  • Support SNMPv1/v2c/v3, support general network management platform
  • 2+2 power supply redundancy, 7+1 fan redundancy
N9500-128QC, 128x400G QSFP112, Broadcom TH 5, 51.2Tbps, Managed AI Data Center Switch
N9500-128QC, 128x400G QSFP112 Ethernet L3 4U Managed Switch, 51.2Tbps, Broadcom Tomahawk 5, Support RoCEv2, for AI/ML/Cloud Data Center/HPC
Specifications
Product Highlights
Questions & Answers
Resources
Specifications
Product Highlights
Questions & Answers
Resources
Description

N9500-128QC, 128-Port Ethernet L3 4U Data Center Switch, 128x 400Gb QSFP112, Support RoCEv2, Broadcom Chip

The N9500-128QC switch is a new generation of high-performance, high-density box switches launched by NADDOD for high-end data centers and AIGC intelligent computing scenarios. The N9500-128QC adopts advanced hardware architecture design and provides 128 400GE ports. All ports support line-speed forwarding and redundant pluggable power supplies and fans. It supports AI-Fabric solution RALB and AILB load balancing technology to improve traffic bandwidth in AIGC intelligent computing scenarios and shorten AI training time. A typical secondary network can support up to 8K 400G ports, and a typical third-level network can support up to 32K 400G ports. It supports one-click deployment of RoCE configuration, and can complete the complex configuration import of RDMA networks such as ECN and PFC with one click to achieve rapid deployment.

Specifications
Ports
128x400G QSFP112
Switch Chip
Tomahawk 5 / BCM78900 Series
Switching capacity
102.4T
Packet forwarding rate
21000Mpps
Power consumption
Maximum: 3145W
Typical: 2167W
RoCE
Supported
Management port
1xMGMT, 1xConsole, 1xUSB
CPU
2.0 GHz quad-core
Memory
16 GB, DDR4 (compatible with 32 GB)
Swap Cache
165.2MB
Airflow
power-to-connector
Flash memory
NOR flash: 64 MB (for Boot ROM)
SSD
240 GB
Power Redundancy
2 + 2 (3 + 1 for 100–127 V AC)
Fan Redundancy
7 + 1
Dimensions (W × D × H)
442mm × 760mm × 175mm(4RU)
Operating temperature
0℃~40℃
Weight
37.3KG
Humidity
Operating humidly: 10% RH to 90% RH (noncondensing)
Storage humidity: 5% RH to 95% RH (non-condensing)
Product Highlights
4U Chassis with 128x 400G Ports for AI and Data Center Networks

Optimized for high-density AI computing and next-gen data centers, the N9500-128QC provides 128x 400G ports in a compact 4U design, ensuring efficient, high-performance networking.

Broadcom Tomahawk 5 Chip Unleashes Exceptional Efficiency for Data Center

NADDOD 800G/400G switches are integrated with Broadcom Tomahawk 5 chip, supporting RoCEv2 lossless networking protocols and providing full L2 and L3 forwarding. They deliver ultra-high performance, low-latency, and high-reliability solutions, especially for data centers, AI/ML clusters, high-performance computing, and distributed storage applications.

High-Performance AIGC Network

Supports large GPU clusters with 200/400/800G bandwidth and 97% utilization. Load balancing and copper interconnects reduce failure risks and costs, enabling efficient, energy-saving data centers.

RoCE Lossless Technology for AI Networks

RDMA (Remote Direct Memory Access) technology based on Ethernet protocol enables low-latency, lossless communication, ensuring efficient data transfer and optimized AI workload performance.

RALB Ensures High Bandwidth Utilization

Maximize bandwidth efficiency with RALB technology, optimizing load distribution and ensuring consistent, high-performance data transfer.

AILB Technology for Dynamic Path Switching

Automatically reroutes paths in response to congestion and conflicts, ensuring optimal data flow and network efficiency.

Questions & Answers
Ask a Question
Q:
What is the throughput of the N9500-128QC switch?
A:
The N9500-128QC switch delivers up to 102.4 Tbps of bi-directional throughput with a forwarding capacity of 21,000 Mpps.
Q:
How large scale of networking can N9500-128QC switch support?
A:
Layer 2 networking can support up to 8,000-card-sized GPU clusters, and Layer 3 can support up to 32,000-card-sized GPU clusters.
Q:
What is the port density and speed of the N9500-128QC switch?
A:
It supports up to 128*400GbE or 256*200GbE ports.
View More
Quality Certification
ISO140012015
ISO 90012015
ISO450012018
FDA
FCC
CE
RoHS
TUV-Mark
UL
WEEE
What We Supply