Tesla T4 GPU
The T4 is a low-power data center GPU optimized for AI inference, offering mixed-precision performance in a compact design. It's commonly used in cloud deployments for cost-efficient scaling of NLP and recommendation systems.
Starting Price
$0.15/hr
Available on 1 cloud providers
Key Specifications
💾Memory
16GB VRAM
🏗️Architecture
Turing
⚙️Compute Units
N/A
🧮Tensor Cores
320
Technical Specifications
Hardware Details
ManufacturerNVIDIA
ArchitectureTuring
CUDA Cores2560
Tensor Cores320
RT Cores40
Compute UnitsN/A
GenerationN/A
Memory & Performance
VRAM16GB
Memory Interface256-bit
Memory Bandwidth320 GB/s
FP32 Performance8.1 TFLOPS
FP16 Performance65 TFLOPS
INT8 Performance130 TOPS
Performance
Computing Power
CUDA Cores2,560
Tensor Cores320
RT Cores40
Computational Performance
FP32 (TFLOPS)8.1
FP16 (TFLOPS)65
INT8 (TOPS)130
Common Use Cases
Inference, video transcoding
Machine Learning & AI
- Training large language models and transformers
- Computer vision and image processing
- Deep learning model development
- High-performance inference workloads
Graphics & Compute
- 3D rendering and visualization
- Scientific simulations
- Data center graphics virtualization
- High-performance computing (HPC)
Market Context
The Tesla T4 sits within NVIDIA's Turing architecture lineup, positioned in the entry performance tier.
Cloud Availability
Available across 1 cloud providers with prices ranging from $0.15/hr. Pricing and availability may vary by region and provider.
Market Position
Released in 2018, this GPU is positioned for professional workloads.