Loading page
vram
80 GB+ VRAM GPUs Cloud Pricing
Train models with hundreds of billions of parameters. Serve 70B+ models at full precision. Run multi-node distributed training with NVLink interconnects. This tier — A100 80 GB, H100, H200, MI300X, and Blackwell-generation GPUs — uses high-bandwidth memory (HBM2e/HBM3/HBM3e) that delivers 2–5x the bandwidth of GDDR, directly impacting training throughput and large-batch inference speed.
GPUs 21
Providers 60
From $0.482/hr
80 GB+ VRAM GPUs Available in the Cloud
A100 SXM
80 GBAmpereNVIDIA
$0.500/hr37 providers
B100
Server192 GBBlackwellNVIDIA
No recent pricing
B200
180 GBBlackwellNVIDIA
$2.80/hr28 providers
Gaudi 2
Server96 GBGaudiIntel
$1.10/hr5 providers
Gaudi 3
Server128 GBGaudiIntel
No recent pricing
GB200
Server384 GBBlackwellNVIDIA
$8.00/hr2 providers
GB300
Server576 GBBlackwellNVIDIA
$4.31/hr2 providers
GH200
96 GBHopperNVIDIA
$2.29/hr5 providers
H100 NVL
Server94 GBHopperNVIDIA
$1.82/hr6 providers
H100 PCIe
Server80 GBHopperNVIDIA
$0.950/hr5 providers
H100 SXM
Server80 GBHopperNVIDIA
$1.10/hr49 providers
H200
141 GBHopperNVIDIA
$1.39/hr39 providers
HGX B300
Server288 GBBlackwellNVIDIA
$3.83/hr21 providers
Max 1550
Server128 GBXe HPCIntel
No recent pricing
MI250
Server128 GBCDNA 2AMD
No recent pricing
MI250X
Server128 GBCDNA 2AMD
No recent pricing
MI300A
Server128 GBCDNA 3AMD
No recent pricing
MI300X
192 GBCDNA 3AMD
$1.85/hr9 providers
MI325X
Server256 GBCDNA 3AMD
$2.00/hr2 providers
MI355X
Server288 GBCDNA 4AMD
$2.59/hr3 providers
RTX PRO 6000
Server96 GBBlackwellNVIDIA
$0.482/hr27 providers
Sample 80 GB+ VRAM GPUs Pricing
| Provider | Price / hr |
|---|---|
$1.57/hr 1× | |
$1.59/hr12mo 1× | |
$1.73/hr 4× | |
$2.30/hr 4× | |
$2.63/hr 4× | |
$2.74/hr 1× | |
$3.75/hr 1× | |
$4.54/hr 1× | |
$6.89/hr 2× |
Direct from providerVia marketplace
Showing 9 of 642 price points. Visit individual GPU pages above for full pricing.
Frequently Asked Questions
Do I need 80 GB+ VRAM?
If you're training models larger than 13B parameters at full precision, running inference on 70B+ models, or need large batch sizes for production throughput, 80 GB+ is recommended. For smaller workloads, 24–48 GB GPUs are more cost-effective.
What is the difference between HBM and GDDR memory?
HBM (High Bandwidth Memory) provides 2–5x the bandwidth of GDDR6X, which directly impacts training throughput and large-model inference speed. HBM is standard on datacenter GPUs (A100, H100, MI300X) while consumer GPUs use GDDR.