The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.
Cisco Nexus 9364D-GX2A switches are qualified to connect Intel Gaudi 2 servers to build a scale-out network for running Large-Language Model (LLM) training, inference, a similar Artificial Intelligence (AI), Machine Learning (ML), or generative AI workload.
Intel Gaudi accelerator
A Gaudi 2 server provides 24 x 100 GbE scale-out or back-end interfaces exposed via six QSFP-DD ports for an inter-Gaudi 2 network, also known as a scale-out network or a back-end network. These ports (marked as QSFP-DD0 to QSFP-DD5 in Figure 2) are dedicated to interconnecting Gaudi 2 accelerators in other servers and are not connected to external networks.
A Gaudi 2 server also provides two NICs, each with two 100 GbE ports (a total of four) for external, storage, and management connectivity. In Figure 1, these NIC ports are below the Gaudi 2 scale-out ports.
Front view of a Intel Gaudi 2 Server
For more information, refer to https://www.intel.com/content/www/us/en/products/details/processors/ai-accelerators/gaudi2.html
Port-side view of a Cisco Nexus 9364D-GX2A switch
# Gaudi2 accelerators |
# HLS-Gaudi2 servers(1) |
# scale-out/ backend QSFP-DD ports(2) |
# scale-out/ backend 100 GbE interfaces(3) |
# Cisco Nexus 9364D-GX2A Leaf switches(4) |
# Cisco Nexus 9364D-GX2A Spine switches(4) |
# Total Cisco Nexus 9364D-GX2A switches |
# QDD- 400-AOCxM transceivers(5) |
# QDD- 400G-DR4-S transceivers(6) |
# of MPO cables(7) |
8 |
1 |
6 |
24 |
0 |
0 |
0 |
0 |
0 |
0 |
16 |
2 |
12 |
48 |
1 |
0 |
1 |
0 |
24 |
12 |
32 |
4 |
24 |
96 |
1 |
0 |
1 |
0 |
48 |
24 |
64 |
8 |
48 |
192 |
1 |
0 |
1 |
0 |
96 |
48 |
128 |
16 |
96 |
384 |
3 |
0 |
3 |
96 |
192 |
96 |
256 |
32 |
192 |
768 |
6 |
3 |
9 |
192 |
384 |
192 |
512 |
64 |
384 |
1536 |
12 |
6 |
18 |
384 |
768 |
384 |
1024 |
128 |
768 |
3072 |
24 |
12 |
36 |
768 |
1536 |
768 |
2048 |
256 |
1536 |
6144 |
48 |
24 |
72 |
1536 |
3072 |
1536 |
4096 |
512 |
3072 |
12288 |
96 |
48 |
144 |
3072 |
6144 |
3072 |