A30 vs A100 PCIe
Manufacturer specifications of the A30 and the A100 PCIe, side by side, per GPU. A value the manufacturer does not publish for the exact model is marked as such, never estimated.
NVIDIA A30 24GB
Ampere
- GPU memory
- 24 GB
- Memory bandwidth
- 933 GB/s
Live price on EverXCloud · No offer right now
Suited workloads
NVIDIA A100 PCIe 80GB
Ampere
- GPU memory
- 80 GB
- Memory bandwidth
- 1,935 GB/s
Live price on EverXCloud · from $1.79 per GPU-hour
Suited workloads
Key differences
- The A100 PCIe has 3.3× the GPU memory of the A30 (80 GB vs 24 GB).
- The A100 PCIe has 2.1× the memory bandwidth of the A30 (1,935 vs 933 GB/s), which usually sets the pace of LLM token generation.
- The A100 PCIe has 1.9× the dense BF16 tensor throughput of the A30 (312 vs 165 TFLOPS), the precision of most training runs.
Specifications
| Specification | A30 | A100 PCIe |
|---|---|---|
| GPU memory | 24 GB | 80 GB (higher) |
| Memory type | HBM2 | HBM2e |
| Memory bandwidth | 933 GB/s | 1,935 GB/s (higher) |
| FP16 tensor | 165 TFLOPS | 312 TFLOPS (higher) |
| BF16 tensor | 165 TFLOPS | 312 TFLOPS (higher) |
| Max power | 165 W | 300 W |
| Architecture | Ampere | Ampere |
| Interconnect | NVLink bridge (200 GB/s) · PCIe 4.0 | NVLink bridge (600 GB/s) · PCIe 4.0 |
The higher value of each measured row is shown in bold. Manufacturer specifications, per GPU. Tensor throughput is shown without sparsity. Values the manufacturer does not publish for this exact model are left out.
What fits in memory, on one GPU
| Workload | A30 | A100 PCIe |
|---|---|---|
| LLM inference, 16-bit | ≈ 9B parameters | ≈ 30B parameters |
| LLM inference, 4-bit | ≈ 30B parameters | ≈ 100B parameters |
| Full training, mixed precision | ≈ 1B parameters | ≈ 4B parameters |
Cautious estimates counting 80% of memory for model weights; context length, batch size and framework change the real limit. Several GPUs together hold larger models.
A30 vs A100 PCIe: frequently asked questions
Which has more memory, the A30 or the A100 PCIe?
The A100 PCIe has 80 GB of GPU memory, against 24 GB for the A30.
Which is faster for LLM inference, the A30 or the A100 PCIe?
Generating tokens is usually bound by memory bandwidth, and the A100 PCIe has 2.1× the bandwidth of the A30. Memory size then decides which models fit. Measure your own model before a large run.
Can I rent the A30 and the A100 PCIe by the hour?
Yes, when sellers list them on the EverXCloud spot exchange: you buy GPU hours at the live price from a prepaid balance in euros, US dollars or pounds. Prices are shown excluding tax and EverXCloud is the merchant of record for every rental.
Be ready on launch day
Create your account now: you can top up your balance and rent the GPU that fits your workload as soon as the market opens.
Have GPU servers?
Businesses can join the EverXCloud Network and sell hours of their GPU servers on the spot exchange.