@TheAhmadOsman: How am I connecting the DGX Spark cluster > Mikrotik CRS804-4DDQ 1.6Tbps switch > (4) 400G QSFP-DD to 2x 200G QSFP56 Ea…
Summary
Ahmad 分享了如何通过 Mikrotik CRS804-4DDQ 1.6Tbps 交换机，将 4 条 400G QSFP-DD 拆分为 2x200G QSFP56，连接 8 台配备 ConnectX-7 的 DGX Spark，实现 1.6Tbps 全速互联的集群方案。
View Cached Full Text
Cached at: 10/18/26, 03:06 AM
How am I connecting the DGX Spark cluster > Mikrotik CRS804-4DDQ 1.6Tbps switch > (4) 400G QSFP-DD to 2x 200G QSFP56 Each DGX Spark has a ConnectX-7 supporting 200Gbps Each cable out of the switch goes into 2 DGX Sparks This allows 8x DGX Sparks cluster at the full 1.6Tbps x.com/TheAhmadOsman/…
Ahmad (@TheAhmadOsman): How am I connecting the DGX Spark cluster
> Mikrotik CRS804-4DDQ 1.6Tbps switch
> (4) 400G QSFP-DD to 2x 200G QSFP56
Each DGX Spark has a ConnectX-7 supporting 200Gbps
Each cable out of the switch goes into 2 DGX Sparks
This allows 8x DGX Sparks cluster at the full 1.6Tbps
Similar Articles
@MiaAI_lab: https://x.com/MiaAI_lab/status/2084629608062673379
A practical guide to selecting MikroTik switches (CRS504, CRS804, CRS812) for building DGX Spark clusters, including pricing and real user examples of 8x DGX Spark setups.
Setting up of a 16xGB10 (DGX Spark) cluster
A user is setting up a 16-node DGX Spark cluster to run frontier open models locally, discussing networking tradeoffs between 200 and 100 Gbit/s.
Cheap Desktop 400GbE Switch MikroTik CRS804-4DDQ-HRM Review
The article reviews the MikroTik CRS804-4DDQ-HRM, a cost-effective desktop 400GbE switch with four QSFP56-DD ports and 1.6Tbps switching capacity, emphasizing its value for high-performance networking environments.
@ivanfioravanti: Best way to connect 4 x DGX Sparks without a switch? Asking for a friend that is evaluating... Thanks
A short tweet asking for recommendations on the best way to connect four NVIDIA DGX Spark units without using a network switch, likely for an evaluation setup.
@onusoz: 16x parallel Gemma-4-26B-A4B-NVFP4 runs 18 output tokens/s, aggregate 300 tok/s 1 DGX Spark with 128 GB unified memo…
@onusoz demonstrates running 16 parallel instances of NVIDIA's quantized Gemma-4-26B-A4B-NVFP4 model on a single DGX Spark with 128GB unified memory, achieving 300 tok/s aggregate, showcasing high concurrency without flashinfer.