Lenovo ThinkSystem SR680a V4 vs QCT QuantaGrid D75H-10U
Both are built on the NVIDIA HGX B300 8-GPU 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 | Lenovo ThinkSystem SR680a V4 | QCT QuantaGrid D75H-10U |
|---|---|---|
| Vendor | Lenovo | QCT |
| Part number | 7DMK | D75H-10U |
| Form factor | rack | rack |
| Released | — | — |
| Accelerator | 8x B300 | 8x B300 |
| Aggregate memory | — | — |
| Aggregate bandwidth | — | — |
| System power | — | — |
| Interconnect | One NVIDIA HGX B300 SXM baseboard, which Lenovo pairs with fast interconnects between GPUs, CPUs and networking | One NVIDIA HGX B300 8-GPU baseboard; scale-out is eight ConnectX-8 OSFP ports carrying 800G Ethernet or InfiniBand at one port per GPU for GPU-Direct RDMA |
Rows shown in grey are identical between the two.
Lenovo ThinkSystem SR680a V4
Lenovo's Blackwell Ultra machine, an 8U server developed in-house around eight NVIDIA HGX B300 SXM GPUs and two Intel Xeon 6700P processors reaching 86 cores at up to 350 W. Notably it is designed for air cooling, where most B300 machines in this catalogue are liquid cooled or need direct-to-chip plumbing, and it carries that with 21 hot-swap fans, six at the front for CPUs and memory and fifteen at the rear for drives and the GPU subsystem. Power is six or eight hot-swap supplies, either 3,200 W AC or 3,800 W high-voltage AC or DC, in N+1 or N+N redundancy. The chassis is 447 by 351 by 924 mm at up to 124.7 kg, the heaviest of Lenovo's 8U accelerator servers. Lenovo publishes no system power draw, no accelerator memory total and no system throughput figure; its per-GPU B300 table matches NVIDIA's own specification for the baseboard, which is linked here.
Vendor product pageQCT QuantaGrid D75H-10U
QCT's 10U Blackwell Ultra server, built on the NVIDIA HGX B300 platform with two Intel Xeon 6 processors. Networking is eight ConnectX-8 OSFP ports at 800G, one per accelerator, for one-to-one GPU-Direct RDMA, and there is room for four full-height full-length PCIe Gen 5 slots. QCT quotes 72 PFLOPS of FP8 and 144 PFLOPS of FP4 without saying whether either is dense or sparse. Both match NVIDIA's sparse figures for this baseboard exactly, and neither matches the dense figures of 36 and 108, so they are recorded as sparse. Worth noting QCT calls the FP8 number a training figure, which would normally imply dense; the number itself settles the basis, not the label. Power is twelve 3,200 W supplies in N+N redundancy, which is capacity rather than draw, so no system power is shown.
Vendor product page