NVIDIA GPU architectures

E760427

NVIDIA GPU architectures are the underlying hardware designs for NVIDIA’s graphics processing units, providing massively parallel compute capabilities that power high-performance graphics, AI, and scientific computing workloads.

Try in SPARQL Jump to: Surface forms Statements Referenced by

All labels observed (1)

Label Occurrences
NVIDIA GPU architectures canonical 1

Statements (50)

Predicate Object
instanceOf GPU microarchitecture family
developedBy NVIDIA NERFINISHED
enables high‑performance simulation
inference acceleration
large‑scale neural network training
real‑time rendering
hasComponent CUDA cores
L2 cache
NVLink interface NERFINISHED
PCI Express interface
RT cores
memory controllers
streaming multiprocessors
tensor cores
includesArchitecture Ada Lovelace architecture NERFINISHED
Ampere architecture NERFINISHED
Blackwell architecture NERFINISHED
Fermi architecture NERFINISHED
Hopper architecture NERFINISHED
Kepler architecture NERFINISHED
Maxwell architecture NERFINISHED
Pascal architecture NERFINISHED
Tesla architecture
Turing architecture NERFINISHED
Volta architecture NERFINISHED
optimizedFor energy efficiency
parallel throughput
supports CUDA programming model NERFINISHED
DirectX NERFINISHED
GPGPU computing
GPU acceleration
NVIDIA CUDA Toolkit NERFINISHED
OpenCL NERFINISHED
OpenGL NERFINISHED
Vulkan NERFINISHED
deep learning acceleration
double‑precision floating point
half‑precision floating point
massively parallel processing
mixed‑precision compute
ray tracing acceleration
single‑precision floating point
targetMarket automotive computing
consumer graphics
data center computing
professional visualization
usedFor artificial intelligence workloads
graphics processing
high‑performance computing
scientific computing

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

cuDNN optimizedFor NVIDIA GPU architectures