What's the price difference between NVIDIA B200 and NVIDIA B300?

On Runcrate today, NVIDIA B200 starts at $7.67/hr and NVIDIA B300 starts at $14.30/hr per GPU on an 8-GPU machine; NVIDIA B200 is cheaper by 46%.

Can I deploy NVIDIA B200 or NVIDIA B300 on Runcrate today?

NVIDIA B200 is in stock as a reserved cluster: we verify your company and your need, then take your exact spec to GPU owners that fit, and you see their offers side by side. NVIDIA B300 is on demand in 2 regions from $14.30/hr per GPU on an 8-GPU machine, billed by the second on most GPUs.

How much VRAM does each GPU have?

NVIDIA B200 has 180GB HBM3e. NVIDIA B300 has 288GB HBM3e.

B200 or B300: which holds a 70B model in 8-bit weights?

It needs 70 GB of 8-bit weights. On the B200: 1 GPU, 110 GB spare. On the B300: 1 GPU, 218 GB spare. Fits on one GPU of either.

GPU comparison

B200 vs B300.

Pick the B300 for memory: 288 GB vs 180 GB. FP16 speed is the same (2,250 TFLOPS), so for models under 180 GB pick by price.

Launch B300 Reserve a cluster

Rent either on Runcrate

B200Cheaperfrom $7.67/hr
$0.043 per GB/hr$0.34 per 100 TFLOPS/hr
Reserve a cluster
B300from $14.30/hr per GPU
on an 8-GPU machine
$0.050 per GB/hr$0.64 per 100 TFLOPS/hr
Launch B300

Live hourly price, then per GB of memory and per 100 FP16 TFLOPS.

Which is better for

Which to pick for each job. Memory fit and throughput, from the datasheets.

JobB200B300Pick

Hold a 70B model in 8-bit weights
70 GB of 8-bit weights
B200: 1 GPU, 110 GB spare
B300: 1 GPU, 218 GB spare
Either. Fits on one GPU of either.
Serve an 8B model in 16-bit weights
16 GB of 16-bit weights
B200: 1 GPU, 164 GB spare
B300: 1 GPU, 272 GB spare
Either. Fits on one GPU of either.
Fine-tune a 70B model with QLoRA
About 46 GB with 4-bit weights
B200: 1 GPU, 134 GB spare
B300: 1 GPU, 242 GB spare
Either. Fits on one GPU of either.
Train on a full node
FP16 tensor throughput per node
B200: 8x, 18,000 TFLOPS, NVLink
B300: 8x, 18,000 TFLOPS, NVLink
Either. Close per node.
Throughput per watt
FP16 TFLOPS per 100 W of TDP
B200: 225 TFLOPS per 100 W
B300: Varies by card
Either. Varies by card.

Compare another pair

Blackwell

NVIDIA B200

from$7.67/hrCheaper

Reserved clusters

Memory: 180 GB HBM3e
FP16 tensor: 2250 TFLOPS
Bandwidth: 8.0 TB/s

Reserve a cluster Deploy B200 now

Full pricing details

Blackwell

NVIDIA B300

from$14.30/hr per GPU

on an 8-GPU machine

Available now · 2 regions

Memory: 288 GB HBM3e
FP16 tensor: 2250 TFLOPS
Bandwidth: up to 8.0 TB/s

Launch B300 Reserve a cluster

Full pricing details

At a glance

Spec by spec. The longer bar is the better number.

Memory

NVIDIA B200180 GB

NVIDIA B300288 GB

Memory bandwidth

NVIDIA B2008.0 TB/s

NVIDIA B300up to 8.0 TB/s

FP16 tensor

NVIDIA B2002250 TFLOPS

NVIDIA B3002250 TFLOPS

Hourly price, from (live)

NVIDIA B200$7.67/hr

NVIDIA B300$14.30/hr

Full specs

The complete spec sheet.

Spec	NVIDIA B200	NVIDIA B300
Architecture	Blackwell	Blackwell
Memory	180 GB	288 GB
Memory type	HBM3e	HBM3e
Memory bandwidth	8.0 TB/s	up to 8.0 TB/s
FP32	75 TFLOPS	75 TFLOPS
FP16 tensor	2250 TFLOPS	2250 TFLOPS
INT8	4500 TOPS	187.5 TOPS
TDP	1000 W	up to 1400 W
Form factor	SXM	SXM
NVLink	Yes	Yes
GPUs per node	8	8
Hourly price (live)	from $7.67/hr	from $14.30/hr per GPU on an 8-GPU machine

Questions

B200 or B300. What buyers ask first.

Other comparisons