B300 vs L40: Specs, VRAM & Price
Short answer: choose the B300 when 288 GB of VRAM is the gating requirement. Choose the L40 when 48 GB is enough for the workload and the lower hourly rate points that way.
The B300 has 288 GB of VRAM; the L40 has 48 GB.
You can rent both by the hour, no purchase required. B300 from $9.89/GPU/hr; L40 from $0.759/GPU/hr. The L40 is 92% cheaper at the entry rate.
B300
Good for training and serving frontier-scale LLMs
L40
Good for fine-tuning mid-size models and high-throughput inference
Last updated: 2026-10-10 18:00:06 UTCRefreshes hourly
Specs side by side
| Spec | B300 | L40 |
|---|---|---|
| VRAM | 288 GB | 48 GB |
| Memory bandwidth | – | 864 GB/s |
| Architecture | Blackwell Ultra | Ada Lovelace |
| Compute capability | 10.3 | 8.9 |
| Precisions in hardware | FP32, FP16, BF16, FP8, INT4 | FP32, FP16, BF16, FP8, INT4 |
| Peak FP16/BF16 tensor TFLOPS | – | 181.05 TFLOPS |
| Peak FP8 tensor TFLOPS | – | 362 TFLOPS |
| Max board power (TDP) | – | 300W |
| Regions | 1 | 3 |
On-demand price
| Rate | B300 | L40 |
|---|---|---|
| Lowest per GPU/hr | $9.89/GPU/hr | $0.850/GPU/hr |
Which is better for AI: B300 or L40?
Choose the B300 when 288 GB of VRAM is the gating requirement. Choose the L40 when 48 GB is enough for the workload and the lower hourly rate points that way.
B300 vs L40: common questions
Is the B300 or the L40 better for AI?
Short answer: choose the B300 when 288 GB of VRAM is the gating requirement. Choose the L40 when 48 GB is enough for the workload and the lower hourly rate points that way.
Is the B300 or the L40 cheaper to rent?
On Aquanode's live marketplace the B300 starts at $9.89/GPU/hr (median $9.89/GPU/hr) and the L40 starts at $0.759/GPU/hr (median $0.902/GPU/hr). The L40 is the cheaper of the two at the entry rate, by 92%.
What is the difference between the B300 and the L40?
The B300 has 288 GB of VRAM against the L40's 48 GB; the B300 is a Blackwell Ultra part and the L40 is Ada Lovelace (compute capability 10.3 vs 8.9); both run BF16, FP8, INT4 workloads in hardware. On price, the B300 lists from $9.89/GPU/hr and the L40 from $0.759/GPU/hr.
How widely available are the B300 and the L40?
The B300 is listed in 1 region and the L40 in 3 regions right now.
How much does 1,000 GPU-hours cost on the B300 vs the L40?
At the lowest rates listed today, 1,000 GPU-hours costs $9,889 on the B300 and $759 on the L40, a difference of $9,130 for the same runtime. Rates are per GPU per hour and update hourly.
How much VRAM does the B300 have compared to the L40?
The B300 has 288 GB of VRAM; the L40 has 48 GB.
How this comparison is calculated
Every price is normalized to a per-GPU hourly rate using the same pipeline as every other pricing surface on Aquanode, and only the cheapest qualifying offer per provider is shown. A dash means that provider does not currently list that GPU.
Architecture, compute capability and supported precisions come from each vendor's published datasheet for that generation, not from the marketplace feed. This page regenerates at most once per hour.
Memory bandwidth, tensor TFLOPS, NVLink/interconnect and max board power are the same verified-from-datasheet figures shown on each model's own B300 page and L40 page, which link the primary vendor source for that model. A dash here means Aquanode hasn't verified that figure from a primary source yet, never a guess.
Go further
Check what fits in each card's memory with the B300 VRAM calculator and the L40 VRAM calculator, see vendor-published throughput in the GPU benchmarks and specs table, or browse every model by generation in the GPU index.
More B300 comparisons
More L40 comparisons
Related reading: H100 pricing and specs, H100 vs H200, A100 vs H100, H100 and H200: SXM vs NVL vs PCIe, and The best GPUs for AI, ranked.
