Skip to content
CLOUD
GPUs
All GPUs →
H100
80 GB HBM3
H200
141 GB HBM3e
B200
180–192 GB HBM3e
B300
288 GB HBM3e
GB200 NVL72
72-GPU system
GB300 NVL72
72-GPU system
A100
40 / 80 GB HBM2e
L40S
48 GB GDDR6
RTX PRO 6000 Server
96 GB GDDR7
GH200
96 / 144 GB HBM3 / HBM3e
RTX 5090
32 GB GDDR7
RTX 4090
24 GB GDDR6X
RTX 6000 Ada
48 GB GDDR6 ECC
Features
On-demand
Reserved Capacity
GPU clusters
Infrastructure
Resources
Documentation
GPU Advisor
Training Calculator
About Arvica
Careers
Support
Pricing
中文
Contact sales
Go to platform
Menu
GPUs
All GPUs →
H100
80 GB HBM3
H200
141 GB HBM3e
B200
180–192 GB HBM3e
B300
288 GB HBM3e
GB200 NVL72
72-GPU system
GB300 NVL72
72-GPU system
A100
40 / 80 GB HBM2e
L40S
48 GB GDDR6
RTX PRO 6000 Server
96 GB GDDR7
GH200
96 / 144 GB HBM3 / HBM3e
RTX 5090
32 GB GDDR7
RTX 4090
24 GB GDDR6X
RTX 6000 Ada
48 GB GDDR6 ECC
Features
On-demand
Reserved Capacity
GPU clusters
Infrastructure
Resources
Documentation
GPU Advisor
Training Calculator
About Arvica
Careers
Support
Pricing
Contact sales
Sign In
ARVICA TOOLS
Training cost calculator
Plan a budget with explicit assumptions. Validate actual performance with your workload.
GPU
H100
H200
B200
B300
A100
L40S
RTX PRO 6000 Server
GH200
RTX 5090
RTX 4090
RTX 6000 Ada
Model parameters (billions)
Dataset tokens (billions)
Epochs
GPU count
Assumed effective TFLOPS per GPU
Hourly rate per GPU (AUD cents)
Cookie settings