QCT QuantaGrid D74F-7U vs QCT QuantaGrid D75F-7U
Both are built on the NVIDIA HGX H200 baseboard, so their accelerator performance is identical rather than merely similar. The choice between them is chassis, cooling, expansion and supplier, not FLOPS.
What actually differs
| Specification | QCT QuantaGrid D74F-7U | QCT QuantaGrid D75F-7U |
|---|---|---|
| Vendor | QCT | QCT |
| Part number | D74F-7U | D75F-7U |
| Form factor | rack | rack |
| Released | — | — |
| Accelerator | 8x H200 | 8x H200 |
| Aggregate memory | 1.1 TB | 1.1 TB |
| Aggregate bandwidth | — | — |
| System power | — | — |
| Interconnect | One NVIDIA HGX H200 baseboard connecting the eight GPUs at 900 GB/s over NVLink; one BlueField-3 DPU, with ten OCP 3.0 TSFF PCIe Gen5 x16 slots in place of the D75F's internal slots | One NVIDIA HGX H200 baseboard connecting the eight GPUs at 900 GB/s over NVLink, which QCT notes is seven times the bandwidth of PCIe Gen 5; two BlueField-3 DPUs handle networking |
Rows shown in grey are identical between the two.
QCT QuantaGrid D74F-7U
The OCP-slot variant of QCT's 7U Hopper machine, carrying the same eight NVIDIA HGX H200 GPUs at 900 GB/s over NVLink and the same 1.1 TB of aggregate high-bandwidth memory as the D75F-7U. The differences are in I/O and packaging: one BlueField-3 DPU rather than two, ten OCP 3.0 TSFF PCIe Gen 5 slots, a shorter 886 mm chassis, and three-plus-three 4,000 W Titanium supplies. QCT quotes the same over 32 petaflops of FP8 figure, which is not shown here for the same reason as on the D75F: the basis is unstated and cannot be settled. Power figures are supply capacity, not draw.
Vendor product pageQCT QuantaGrid D75F-7U
QCT's 7U Hopper machine, with two fifth or fourth generation Intel Xeon Scalable processors and eight NVIDIA HGX H200 GPUs linked at 900 GB/s over NVLink. Storage is eighteen small-form-factor all-NVMe bays for GPU-Direct Storage plus a boot drive, and two NVIDIA BlueField-3 DPUs offload networking. QCT gives 1.1 TB of aggregate high-bandwidth memory. It also quotes over 32 petaflops of FP8, which is not shown here: QCT does not say whether that is dense or sparse, it is close to double the dense figure NVIDIA published for this baseboard, and NVIDIA no longer lists the H200 on its HGX page, so there is no way to settle which column it belongs in without guessing. Power is four-plus-two 4,000 W Titanium supplies, which is capacity rather than draw.
Vendor product page