Hopper
NVIDIA H100 PCIe 80GB
wins
4
The H100 PCIe is the more affordable form factor of the H100.
Turing
NVIDIA T4
wins
1
The NVIDIA T4 is the most widely deployed inference GPU in the cloud.
Performance comparison
Visual bars — winner highlighted
NVIDIA H100 PCIe 80GB
Metric
NVIDIA T4
FP16 TFLOPS
Bandwidth
VRAM
TDP (lower=better)
TFLOPS/watt
GB/s per watt
Full specification table
| Spec | NVIDIA H100 PCIe 80GB | NVIDIA T4 |
|---|---|---|
| Architecture | Hopper | Turing |
| VRAM | 80 GB HBM2e | 16 GB GDDR6 |
| VRAM type | HBM2e | GDDR6 |
| Memory bandwidth | 2,000 GB/s | 320 GB/s |
| FP16 TFLOPS | 756 TFLOPS | 65 TFLOPS |
| TDP | 350 W | 70 W |
| TFLOPS/watt | 2.16 T/W | 0.93 T/W |
| GB/s per watt | 5.71 GB/s/W | 4.57 GB/s/W |
| NVLink | No | No |
| Release year | 2022 | 2018 |
Live cloud pricing
On-demand pricing across providers — updated in real time
Power efficiency analysis
TFLOPS/watt and GB/s/watt — critical for data center TCO
NVIDIA H100 PCIe 80GB
NVIDIA T4
TFLOPS/watt measures compute efficiency — how much AI throughput you get per watt of power consumed. For data centers with PUE of 1.2–1.5, a 10% improvement in TFLOPS/watt translates directly to lower electricity costs and cooling requirements. GB/s/watt measures memory bandwidth efficiency, which is the binding constraint for memory-bound LLM inference workloads.
When to choose each GPU
Choose NVIDIA H100 PCIe 80GB for:
- Budget Hopper inference
- PCIe server deployments
- Cost-sensitive LLM serving
Weaknesses / Limitations
- Lower FP16 TFLOPS (756) and bandwidth (2,000 GB/s) than H100 SXM5
- No NVLink — cannot scale across cards for very large models
- Higher cost than A100 with only moderate performance advantage
Choose NVIDIA T4 for:
- Low-cost inference
- Edge AI
- Multi-tenant serving
Weaknesses / Limitations
- 16 GB VRAM limits model size to ~13B at INT4
- Low FP16 TFLOPS (65) — not suitable for throughput-sensitive serving
- Older Turing architecture — no Ada or Hopper efficiency improvements
Popular GPU comparisons