NVIDIA GeForce RTX 4070 Max-Q Full Specs
4,608 Shaders 1.23GHz |
8GB GDDR6256GB/s |
·· 11.34 TFLOPS |
Form Factor Soldered |
TDP 35W |
Clock Speed ··· |
Peak OPS 362.7 TOPSINT4 Tensor Sparse |
Tensor FP8-16 90.69 TFLOPSFP8-16 Tensor Sparse 181.4 TFLOPSTensor FP8-32 45.34 TFLOPSFP8-32 Tensor Sparse 90.69 TFLOPS |
Tensor FP16-16 45.34 TFLOPSFP16-16 Tensor Sparse 90.69 TFLOPSTensor FP16-32 22.67 TFLOPSFP16-32 Tensor Sparse 45.34 TFLOPS |
BF16 11.34 TFLOPSTensor BF16 22.67 TFLOPSBF16 Tensor Sparse 45.34 TFLOPS |
Tensor TF32 11.34 TFLOPS |
FP32 11.34 TFLOPS |
FP64 177.1 GFLOPS |
Tensor INT4 181.4 TOPS |
Tensor INT8 90.69 TOPS |
Ray 26.2 TOPS |
Pixel Rate 59 GPixel/s |
Texture Rate 177.1 GTexel/s |
Shaders 4,608 Shaders |
TMUs 144 TMUs |
ROPs 48 ROPs |
Tensor Cores 144 T-Cores |
RT Cores 36 RT-Cores |
SMs 36 SMs |
Base Clock 735MHz |
Boost Clock 1.23GHz |
L2 Cache 32.8MB shared |
8GB GDDR6 |
Memory Bus 128-bit |
Memory Speed 16GT/s |
Memory Bandwidth 256GB/s |
ECC No |
TDP 35W |
Multi-Monitor 4 |
HDCP HDCP 2.3 |
Encoder Model NVENC 8 |
Decoder Model NVDEC 5 |
Form Factor Soldered |
Manufacturer |
Chip Designer |
Architecture |
Family |
Branding  |
Codename NV186 |
Chip Variant AD106-350-A1 |
Market Segment Laptop |
Release Date Feb 22, 2023 |
Foundry TSMC |
Fabrication Node 4N |
Die Size 188mm² |
Transistor Count 22.9 Billion |
Transistor Density 121.8 MTr/mm² |
No images available
