NVIDIA B300 rental: early access

NVIDIA B300 specifications

Architecture

Blackwell

Announced

2025

Memory

270 GBUsable memory per NVIDIA. The chip’s maximum memory is 288 GB.

Memory type

HBM3e

Memory bandwidth

7.7 TB/s

FP32

75 TFLOPS

FP16/BF16, dense

2,250 TFLOPS

FP8, dense

4,500 TFLOPS

FP4, dense

14,000 TFLOPS

FP64

1.2 TFLOPS

FP64 Tensor Core

1.2 TFLOPS

RT cores

no

NVENC/NVDEC

NVENC 0, NVDEC 7

MIG

up to 7 instances

GPU interconnect

NVLink 1.8 TB/s

Power (TDP/TGP)

up to 1,100 W

Form factor

SXM (being confirmed)

Memory is specified per GPU. Multiple GPUs have separate memory: using a larger model across them requires compatible software and model partitioning. NVLink does not make them one GPU.

The B300 form factor is still being confirmed. Specs from NVIDIA’s official materials, checked 26 September 2026.

The B300 by workload

• Inference: LLMs up to 123B in FP8 on one GPU
• Fine-tuning: QLoRA up to 235B, LoRA up to 70B on one GPU
• Training from scratch: models up to 8B on one GPU
• Rendering: no RT cores
• FP64 computing: not suited to this GPU

An estimate from open formulas, not a benchmark or an invoice.

Servers with the NVIDIA B300

Server configuration (CPU, RAM, disk, network) is being confirmed.

Sizing assumption: up to 8 GPUs per server. We’ll confirm the actual configuration in our reply.

Sources

• NVIDIA Blackwell Ultra Datasheet (opens in a new tab)
• NVIDIA Technical Blog: Inside NVIDIA Blackwell Ultra (opens in a new tab)
• NVIDIA DGX B300 User Guide (opens in a new tab)
• NVIDIA Video Encode and Decode Support Matrix (opens in a new tab)
• NVIDIA Omniverse Developer Guide: Technical Requirements (opens in a new tab)
• NVIDIA Newsroom, March 18, 2025 (opens in a new tab)

Similar GPUs

• NVIDIA B200

  180 GB HBM3e · 7.7 TB/s · NVLink

  B300 leads on 2 of 8 metrics, B200 on 1. B300 has more FP4 (14,000 vs 9,000 TFLOPS) and memory (270 vs 180 GB). But B200 has more FP64 (37 vs 1.2 TFLOPS). Equal: memory bandwidth, FP32, FP16/BF16, FP8, NVLink.

• NVIDIA H200

  141 GB HBM3e · 4.8 TB/s · NVLink

  B300 leads on 6 of 7 metrics, H200 on 1. B300 has more FP16/BF16 (2,250 vs 989.4 TFLOPS) and FP8 (4,500 vs 1,978.9 TFLOPS). But H200 has more FP64 Tensor Core throughput (67 vs 1.2 TFLOPS).

• NVIDIA RTX PRO 6000 Blackwell

  96 GB GDDR7 · 1.6 TB/s · PCIe 5

  B300 leads on 5 of 6 metrics, RTX PRO 6000 on 1. B300 has more FP4 (14,000 vs 1,871.2 TFLOPS) and memory bandwidth (7.7 vs 1.6 TB/⁠s). But RTX PRO 6000 has more FP32 (120 vs 75 TFLOPS).

Dense FP16/BF16 and FP8 from the H100 SXM5 column of NVIDIA’s whitepaper: for the H200, NVIDIA lists only with-sparsity figures, the same as the H100 SXM’s.

The RTX PRO 6000’s FP16, FP8 and FP4 are derived figures. Derived: NVIDIA’s whitepaper per-SM Tensor Core rate × the 2,430 MHz boost clock in the product brief. For the Server Edition, NVIDIA gives FP16/FP8/FP4 only as rounded 1/2/4 PFLOPS peaks, the level of sparse figures; it publishes no dense ones.