AMD Instinct MI355X Platform vs Ingrasys GB2281A
Accelerator configuration, aggregate memory, bandwidth and power, side by side.
Specifications
| Specification | AMD Instinct MI355X Platform | Ingrasys GB2281A |
|---|---|---|
| Vendor | AMD | Ingrasys |
| Part number | — | GB2281A |
| Form factor | baseboard | rack |
| Released | 2025 | — |
| Accelerator | 8x Instinct MI355X | 8x Instinct MI355X |
| Aggregate memory | 2.3 TB | — |
| Aggregate bandwidth | 64.0 TB/s | — |
| System power | — | — |
| Interconnect | 4th Gen AMD Infinity Fabric, eight fully connected OAM modules on an OCP UBB design | Eight MI355X OAM modules on a UBB 2.0 universal baseboard, fully connected by AMD Infinity Fabric mesh; eight PCIe Gen5 x16 low-profile or half-length slots for networking |
Rows shown in grey are identical between the two.
AMD Instinct MI355X Platform
AMD's eight-GPU building block, integrating eight fully connected MI355X OAM modules on an industry-standard OCP universal baseboard over fourth-generation Infinity Fabric. AMD publishes the platform totals of 2.3 TB of HBM3E and 64 TB/s of aggregate memory bandwidth, both of which are exactly eight times the per-GPU figures, which is why the precision values here are also eight times AMD's published per-GPU specification rather than a separate platform table, since AMD provides none. Sparse figures appear only for the four datatypes where AMD publishes a structured-sparsity number, which are FP8, FP16, BF16 and INT8; AMD gives no sparsity figure for MXFP4 or MXFP6, so neither does this row. Supported in both air-cooled UBB servers and direct liquid cooled platforms. AMD publishes no platform power figure.
Vendor product pageIngrasys GB2281A
Ingrasys, the data centre arm of Foxconn, builds this as its AMD accelerator node: eight Instinct MI355X modules on a UBB 2.0 universal baseboard connected by Infinity Fabric mesh. The dimensions give away who it is for. At 524.3 mm wide and 93.5 mm high it is an Open Compute 21 inch node at two OpenU, not a 19 inch rack unit, which is how these reach hyperscale buyers rather than enterprise racks. Storage is sixteen E1.S NVMe bays, expansion eight PCIe Gen5 x16 slots, management an AST2600, and cooling is direct liquid to the accelerators with four 80 mm fans in N+1 for everything else. Ingrasys publishes no power figure and no throughput for the node. It does quote 2.3 TB of HBM3E and 64 TB/s of aggregate bandwidth, but for the AMD platform rather than for this product, so those are on the linked platform entry where AMD publishes them.
Vendor product page