Can I deploy NVIDIA H100 or NVIDIA GH200 on Runcrate today?

NVIDIA H100 is on demand in 6 regions from $3.55/hr, billed by the second on most GPUs. NVIDIA GH200 is not in stock on Runcrate right now; leave your email on its pricing page to register interest.

How much VRAM does each GPU have?

NVIDIA H100 has 80GB HBM3. NVIDIA GH200 has 96GB HBM3.

H100 or GH200: which holds a 70B model in 8-bit weights?

It needs 70 GB of 8-bit weights. On the H100: 1 GPU, 10 GB spare. On the GH200: 1 GPU, 26 GB spare. Pick the GH200: same GPU count, up to 19% more memory bandwidth.

GPU comparison

H100 vs GH200.

Pick the GH200 for memory: 96 GB vs 80 GB, with up to 19% more memory bandwidth. FP16 speed is close (up to 989 TFLOPS vs 990 TFLOPS), so for models under 80 GB pick by price.

Launch H100 Reserve a cluster

Rent either on Runcrate

H100from $3.55/hr
$0.044 per GB/hrper TFLOPS: varies by card
Launch H100 H100 pricing
GH200Not in stock right now
Register interest GH200 pricing

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.

JobH100GH200Pick

Hold a 70B model in 8-bit weights
70 GB of 8-bit weights
H100: 1 GPU, 10 GB spare
GH200: 1 GPU, 26 GB spare
GH200. Same GPU count, up to 19% more memory bandwidth.
Serve an 8B model in 16-bit weights
16 GB of 16-bit weights
H100: 1 GPU, 64 GB spare
GH200: 1 GPU, 80 GB spare
GH200. Same GPU count, up to 19% more memory bandwidth.
Fine-tune a 70B model with QLoRA
About 46 GB with 4-bit weights
H100: 1 GPU, 34 GB spare
GH200: 1 GPU, 50 GB spare
Either. Fits on one GPU of either.
Train on a full node
FP16 tensor throughput per node
H100: 8x, up to 7,912 TFLOPS
GH200: 1x, 990 TFLOPS
H100. 8 GPUs per node, not 1.
Throughput per watt
FP16 TFLOPS per 100 W of TDP
H100: Varies by card
GH200: Superchip power, not a GPU TDP
Either. Not comparable.

Compare another pair

Hopper

NVIDIA H100

from$3.55/hr

Available now · 6 regions

Memory: 80 GB HBM3
FP16 tensor: up to 989 TFLOPS
Bandwidth: up to 3.4 TB/s

Launch H100 Reserve a cluster

Full pricing details

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

At a glance

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

Memory

NVIDIA H10080 GB

NVIDIA GH20096 GB

Memory bandwidth

NVIDIA H100up to 3.4 TB/s

NVIDIA GH200up to 4.0 TB/s

FP16 tensor

NVIDIA H100up to 989 TFLOPS

NVIDIA GH200990 TFLOPS

FP32

NVIDIA H100up to 67 TFLOPS

NVIDIA GH20067 TFLOPS

Full specs

The complete spec sheet.

Spec	NVIDIA H100	NVIDIA GH200
Architecture	Hopper	Hopper
Memory	80 GB	96 GB
Memory type	HBM3	HBM3
Memory bandwidth	up to 3.4 TB/s	up to 4.0 TB/s
FP32	up to 67 TFLOPS	67 TFLOPS
FP16 tensor	up to 989 TFLOPS	990 TFLOPS
INT8	up to 1979 TOPS	1979 TOPS
Power	up to 700 W per GPU (TDP)	up to 1000 W per superchip (CPU, GPU and memory)
GPUs per node	8	1
Hourly price (live)	from $3.55/hr	Not in stock right now

Questions

H100 or GH200. What buyers ask first.

Other comparisons