NVIDIA Thor T4000 Blackwell Full Specs

Summary

1,536 Shaders 1.53GHz

Shared Memory273GB/sLPDDR5X

Peak·FP32·FP16 4.7 TFLOPS

Form Factor iGPU

TDP Shared

Highlights

Peak OPS

Peak OPS FP32 FP16 Tensor FP6 FP6 Tensor Sparse Tensor NVFP4 NVFP4 Tensor Sparse Tensor FP4 FP4 Tensor Sparse Tensor FP8-32 FP8-32 Tensor Sparse Tensor FP16-32 FP16-32 Tensor Sparse FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Memory Bandwidth

Thor T4000 Blackwell1.2 PFLOPSNVFP4 Tensor Sparse

x1

Top Score3.35 PFLOPS FP4 Tensor Sparse

x2.79

Desktop1xGeForce RTX 5090

Theoretical Performance

Clock Speed1530MHz UC·Boost·OC

Peak OPS 1.2 PFLOPSNVFP4 Tensor Sparse

Tensor FP4 601.6 TFLOPS FP4 Tensor Sparse 1.2 PFLOPS NVFP4 Tensor 601.6 TFLOPS NVFP4 Tensor Sparse 1.2 PFLOPS

FP6 Tensor 300.8 TFLOPS FP6 Tensor Sparse 601.6 TFLOPS

Tensor FP8-32 300.8 TFLOPS FP8-32 Tensor Sparse 601.6 TFLOPS

FP16 9.4 TFLOPS Tensor FP16-32 150.4 TFLOPS FP16-32 Tensor Sparse 300.8 TFLOPS

BF16 9.4 TFLOPS Tensor BF16 150.4 TFLOPS BF16 Tensor Sparse 300.8 TFLOPS

Tensor TF32 75.2 TFLOPS

FP32 4.7 TFLOPS

FP64 73.44 GFLOPS

Tensor INT8 300.8 TOPS

Texture Rate 73.4 GTexel/s

Core Config

Shaders 1,536 Shaders

TMUs 48 TMUs

Tensor Cores 48 T-Cores

SMs 12 SMs

Clocks

Boost Clock 1.53GHz

Tensor Clock 1.53GHz

Cache

L1 192KB/SM Tex

L2 Cache 24MB shared

Memory

Shared MemoryLPDDR5X

Memory Bus 256-bit

Memory Speed 8.5GT/s

Memory Bandwidth 273GB/s

Power & Thermals

TDP Shared

Max Temp 115°C Max

API Support

OpenGL 4.6 Vulkan 1.3

CUDA 11

Encoder

Encoder NVENC 9.1

Codec AVC (H.264) HEVC (H.265)

Decoder

Decoder NVDEC 6.1

Codec MPEG-1 MPEG-2 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265) AV1

Card

Form Factor iGPU

Chip

Manufacturer NVIDIA

Chip Designer NVIDIA

Architecture Blackwell

Family Blackwell

Branding

Codename GB10B

Market Segment Embedded

Release Date Jan 5, 2026

Fabrication

Foundry TSMC

Fabrication Node 4N

Die Size 402mm²

Images