Need Help?
+(65) 6018 4212 Singapore
+1 (888) 998 7997 United States
+1 (855) 706 2447 Canada
200G QSFP56 to QSFP56 AOC In Stock
Item Spotlights
N9500-128QC, 128-Port Ethernet L3 4U Data Center Switch, 128x 400Gb QSFP112, Preloaded with SONiC OS, Support RoCEv2, Broadcom Chip
N9500-128QC is a core or high-speed, high-density aggregation switch for a new generation of high-performance data centers. N9500-128QC provides 128 QSFP112 400GE high-density ports and L2/L3 line-speed switching capabilities, and is backward compatible with 100GE/200GE interfaces. It supports smooth network expansion, and the basic Layer 2 network can support 8K 400G ports to calculate the network scale. The simplified network topology makes it easier to build a large-capacity, low-latency, non-blocking switching network, reduce the cost of network construction equipment and reduce energy consumption per Gbit, and further optimize and improve the performance of the entire network.
Adopting 400G RoCE solution to achieve low-latency and lossless network communication to meet the core network of large data centers and DCI interconnection scenarios of ultra-large-scale data centers.
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.
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.
RDMA (Remote Direct Memory Access) technology based on Ethernet protocol enables low-latency, lossless communication, ensuring efficient data transfer and optimized AI workload performance.
Global load balancing, reactive path rebalancing, hardware link failover and drop congestion notification included in Cognitive Routing significantly enhance network capacity utilization and the job completion times for AI/ML training clusters.
