Can I deploy NVIDIA GH200 or NVIDIA GB300 on Runcrate today?

NVIDIA GH200 is not in stock on Runcrate right now; leave your email on its pricing page to register interest. NVIDIA GB300 is not in stock on Runcrate right now; reserve a cluster, and once we've verified your company and your need we take your exact spec to GPU owners that fit.

How much VRAM does each GPU have?

NVIDIA GH200 has 96GB HBM3. NVIDIA GB300 has 288GB HBM3e.

GH200 or GB300: which holds a 70B model in 8-bit weights?

It needs 70 GB of 8-bit weights. On the GH200: 1 GPU, 26 GB spare. On the GB300: 1 GPU, 218 GB spare. Pick the GB300: same GPU count, up to 2.0x the memory bandwidth.

GPU comparison

GH200 vs GB300.

The GB300 is ahead on memory (288 GB vs 96 GB) and speed (2.5x the FP16 throughput and up to 2.0x the memory bandwidth). The GH200 makes sense when your model fits in 96 GB and its hourly price is lower.

Rent either on Runcrate

GH200Not in stock right now
Register interest GH200 pricing
GB300Not in stock right now
Reserve a cluster GB300 pricing

Which is better for

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

JobGH200GB300Pick

Hold a 70B model in 8-bit weights
70 GB of 8-bit weights
GH200: 1 GPU, 26 GB spare
GB300: 1 GPU, 218 GB spare
GB300. Same GPU count, up to 2.0x the memory bandwidth.
Serve an 8B model in 16-bit weights
16 GB of 16-bit weights
GH200: 1 GPU, 80 GB spare
GB300: 1 GPU, 272 GB spare
GB300. Same GPU count, up to 2.0x the memory bandwidth.
Fine-tune a 70B model with QLoRA
About 46 GB with 4-bit weights
GH200: 1 GPU, 50 GB spare
GB300: 1 GPU, 242 GB spare
GB300. Same GPU count, 2.5x the FP16 throughput.
Train on a full node
FP16 tensor throughput per node
GH200: 1x, 990 TFLOPS
GB300: 8x, 20,000 TFLOPS, NVLink
GB300. 8 GPUs per node, not 1.
Throughput per watt
FP16 TFLOPS per 100 W of TDP
GH200: Superchip power, not a GPU TDP
GB300: Varies by card
Either. Not comparable.

Compare another pair

Hopper

NVIDIA GH200

Not in stock right now

Memory: 96 GB HBM3
FP16 tensor: 990 TFLOPS
Bandwidth: up to 4.0 TB/s

Full pricing details

Blackwell

NVIDIA GB300

Not in stock right now

Memory: 288 GB HBM3e
FP16 tensor: 2500 TFLOPS
Bandwidth: up to 8.0 TB/s

Reserve a cluster Deploy B200 instead

Full pricing details

At a glance

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

Memory

NVIDIA GH20096 GB

NVIDIA GB300288 GB

Memory bandwidth

NVIDIA GH200up to 4.0 TB/s

NVIDIA GB300up to 8.0 TB/s

FP16 tensor

NVIDIA GH200990 TFLOPS

NVIDIA GB3002500 TFLOPS

FP32

NVIDIA GH20067 TFLOPS

NVIDIA GB30083 TFLOPS

Full specs

The complete spec sheet.

Spec	NVIDIA GH200	NVIDIA GB300
Architecture	Hopper	Blackwell
Memory	96 GB	288 GB
Memory type	HBM3	HBM3e
Memory bandwidth	up to 4.0 TB/s	up to 8.0 TB/s
FP32	67 TFLOPS	83 TFLOPS
FP16 tensor	990 TFLOPS	2500 TFLOPS
INT8	1979 TOPS	167 TOPS
Power	up to 1000 W per superchip (CPU, GPU and memory)	up to 1400 W per GPU (TDP)
Form factor	Superchip	Superchip
NVLink	One GPU	Yes
GPUs per node	1	8
Hourly price (live)	Not in stock right now	Not in stock right now

Questions

GH200 or GB300. What buyers ask first.

Other comparisons