Data Center Census Near me

GPU comparison

B200 vs B300: specs and rental prices

NVIDIA B200 against NVIDIA B300: 180 GB (derived) against 270 GB of memory, lowest listed on-demand price $6.00 against $7.40 per GPU-hour. Specs from NVIDIA, prices as of Oct 25, 2026.

In short

  • The B300 lists 270 GB of GPU memory against 180 GB (derived) for the B200: 1.5× higher.
  • Memory type is HBM3e on the B200 and HBM3E on the B300.
  • The B200 lists 8,000 GB/s (derived) of memory bandwidth against 7,700 GB/s for the B300: 1.04× higher.
  • Not every figure can be compared: at least one vendor page does not publish: FP8 tensor throughput; FP16 / BF16 tensor throughput; TF32 tensor throughput; FP64 throughput; Maximum power.
  • The lowest listed on-demand price is $6.00 per GPU-hour for the B200 (Hyperstack) and $7.40 for the B300 (Hyperstack).
  • Per GB of GPU memory that is $0.033 per hour for the B200 and $0.027 for the B300.

This summary is generated from the tables below. Prices were last checked Oct 25, 2026. A listing not re-checked for 7 days is greyed and left out of every lowest-price figure; after 14 days it is removed from the page.

Specifications side by side

SpecificationB200, per GPUB300, per GPU (HGX B300)Difference
GPU memory180 GB (derived)270 GBB300 1.5× higher
Memory typeHBM3eHBM3EDifferent
Memory bandwidth8,000 GB/s (derived)7,700 GB/sB200 1.04× higher
FP8 tensor throughputNot published9,000 TFLOPS (with sparsity)—
FP16 / BF16 tensor throughputNot published4,500 TFLOPS (with sparsity)—
TF32 tensor throughputNot published2,200 TFLOPS (with sparsity)—
FP64 throughputNot published1.2 TFLOPS—
Maximum powerNot published1,100 W—
GPU-to-GPU interconnect1,800 GB/s1,800 GB/sSame

Compared versions: B200, per GPU and B300, per GPU (HGX B300). The figure in the stronger cell is shaded; power is shown without a winner. Two throughput figures are compared only when both vendors state them on the same basis (both with sparsity, or both dense), and “derived” means a total the vendor publishes divided by the GPU count it publishes. Full tables with sources: B200 specifications · B300 specifications.

Whole 8-GPU board

SpecificationHGX B200 board (8 GPUs)HGX B300 board (8 GPUs)Difference
GPU memory1,400 GB2,100 GBB300 1.5× higher
FP8 tensor throughput, with sparsity72,000 TFLOPS72,000 TFLOPSSame
FP16 / BF16 tensor throughput, with sparsity36,000 TFLOPS36,000 TFLOPSSame
TF32 tensor throughput, with sparsity18,000 TFLOPS18,000 TFLOPSSame
FP64 throughput296 TFLOPS10 TFLOPSB200 29.6× higher
GPU-to-GPU interconnect1,800 GB/s1,800 GB/sSame

Totals NVIDIA publishes for the whole HGX board, not per GPU.

Rental prices

Rental priceB200B300Difference
Lowest listed on-demand price$6.00 per GPU-hour · Hyperstack · checked Oct 25, 2026$7.40 per GPU-hour · Hyperstack · checked Oct 25, 2026B300 1.23× higher
Lowest listed spot price$0.99 per GPU-hour · Nebius · checked Oct 25, 2026$0.99 per GPU-hour · Nebius · checked Oct 25, 2026Same
Price per GB of memory$0.033 per GB of GPU memory per hour$0.027 per GB of GPU memory per hourB200 1.22× higher
Fresh listings16 from 9 providers13 from 10 providers—

Prices cover every version and instance size the providers list, so they are not tied to the versions compared above; the price per GB of memory divides the lowest fresh on-demand price by the memory of the compared version. Listed prices are not like for like across providers. See every listing: B200 prices · B300 prices.

More on these GPUs

Cite this page

Last updated Oct 25, 2026. Figures may change as records are updated — include the date you retrieved them.

APAData Center Census. (2026). B200 vs B300. Retrieved from https://datacentercensus.com/gpu-prices/compare/b200-vs-b300
MLA“B200 vs B300.” Data Center Census, Oct 25, 2026, https://datacentercensus.com/gpu-prices/compare/b200-vs-b300.
Linkhttps://datacentercensus.com/gpu-prices/compare/b200-vs-b300

Republishing a figure or table? Attribute it to Data Center Census with a link to this page. The underlying data is downloadable as JSON and CSV; the press kit has standing figures and contact details.