GPU

A dedicated NVIDIA graphics card (GPU) is strongly recommended for training and applying deep learning models. Model training and applying performance on a GPU can be up to 20x faster than on a CPU, and can shorten training time from weeks to hours.

Since applying often takes seconds, compared to hours for training, a CPU can more reasonably be used if necessary. Please refer to the general System Requirements for more information on CPU requirements and recommendations.

Recommendations

Below is a table of recommended GPU options.

Specs

  • Architectures: (later is better): Ada Lovelace > Ampere > Turing > Volta > Pascal
  • RAM: The more the better. More important than core count. HBM2 > GDDR6X > GDDR6 > GDDR5X > GDDR5 in terms of memory bandwidth
  • Cores: For a given architecture, more often leads to faster training and application times

Use Cases

  • Deep Learning: Training and running deep learning models (requires 8+ GB RAM)
  • Spotlight: Using Spotlight and Snap tools (requires 12+ GB RAM)
Model Architecture RAM Cores Deep Learning Spotlight
NVIDIA RTX 6000 Ada Ada Lovelace 48 GB GDDR6 18176 ✅✅✅✅ ✅✅✅✅
GeForce RTX 4090 Ada Lovelace 24 GB GDDR6X 16384 ✅✅✅✅ ✅✅✅✅
NVIDIA RTX 4500 Ada Ada Lovelace 24 GB GDDR6 7680 ✅✅✅✅ ✅✅✅✅
NVIDIA RTX 4000 Ada Ada Lovelace 20 GB GDDR6 6144 ✅✅✅⚠️ ✅✅✅⚠️
GeForce RTX 4080 Super Ada Lovelace 16 GB GDDR6X 10240 ✅✅✅⚠️ ✅✅✅⚠️
GeForce RTX 4080 Ada Lovelace 16 GB GDDR6X 9728 ✅✅✅⚠️ ✅✅✅⚠️
NVIDIA RTX 5000 Ada Ada Lovelace 16 GB GDDR6 9728 ✅✅✅⚠️ ✅✅✅⚠️
GeForce RTX 4070 Ti Super 16GB Ada Lovelace 16 GB GDDR6X 8448 ✅✅✅⚠️ ✅✅✅⚠️
GeForce RTX 4070 Ti Ada Lovelace 12 GB GDDR6X 7680 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 4070 Super Ada Lovelace 12 GB GDDR6X 7168 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 4070 Ada Lovelace 12 GB GDDR6X 5888 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 4060 Ti 16GB Ada Lovelace 16 GB GDDR6 4352 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 4060 Ti 8GB Ada Lovelace 8 GB GDDR6 4352 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
GeForce RTX 4060 Ada Lovelace 8 GB GDDR6 3072 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
NVIDIA RTX 2000 Ada Ada Lovelace 8 GB GDDR6 3072 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
NVIDIA RTX A4500 Ampere 20 GB GDDR6 7168 ✅✅✅⚠️ ✅✅✅⚠️
NVIDIA RTX A4000 Ampere 16 GB GDDR6 6144 ✅✅✅⚠️ ✅✅✅⚠️
GeForce RTX 3080 Ti Ampere 12 GB GDDR6X 10240 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 3080 12GB Ampere 12 GB GDDR6X 8960 ✅✅✅⚠️ ✅✅⚠️⚠️
GeForce RTX 3080 10GB Ampere 10 GB GDDR6X 8704 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
GeForce RTX 3070 Ti Ampere 8 GB GDDR6X 6144 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
GeForce RTX 3070 Ampere 8 GB GDDR6 5888 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
GeForce RTX 3060 Ampere 12 GB GDDR6 3584 ✅✅✅⚠️ ✅✅⚠️⚠️
Quadro RTX 6000 Turing 24 GB GDDR6 4608 ✅✅✅✅ ✅✅✅✅
Titan RTX Turing 24 GB GDDR6 4608 ✅✅✅✅ ✅✅✅✅
Quadro RTX 5000 Turing 16 GB GDDR6 3072 ✅✅✅⚠️ ✅✅✅⚠️
Quadro RTX 4000 Turing 8 GB GDDR6 2304 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️
Tesla V100 32GB Volta 32 GB HBM2 5120 ✅✅✅✅ ✅✅✅✅
Tesla V100 16GB Volta 16 GB HBM2 5120 ✅✅✅⚠️ ✅✅✅⚠️
Titan V Volta 12 GB HBM2 5120 ✅✅⚠️⚠️ ✅✅⚠️⚠️
Titan Xp Pascal 12 GB GDDR5X 3840 ✅✅⚠️⚠️ ✅✅⚠️⚠️
Quadro P4000 Pascal 8 GB GDDR5 1792 ✅✅⚠️⚠️ ✅⚠️⚠️⚠️

Need more help with this?
Chat with an expert now ››

Was this helpful?

Yes No
You indicated this topic was not helpful to you ...
Could you please leave a comment telling us why? Thank you!
Thanks for your feedback.