NVIDIA GeForce RTX 4070 Ti Super vs NVIDIA H200 SXM Full Specs - HMC

8,448 Shaders 2.61GHz	16,896 Shaders 1.98GHz
16GB GDDR6X716.8GB/s	141GB HBM3e4.81TB/s
·· 44.1 TFLOPS	·· 66.91 TFLOPS
Form Factor PCIe Card	Form Factor SXM5
TDP 285W	TDP 700W
Power Connectors 1x 16-Pin 12VHPWR	Power Connectors -

GeForce RTX 4070 Ti Super

GeForce RTX 4070 Ti Super1.41 POPSINT4 Tensor Sparse

x1

H200 SXM

H200 SXM3.96 POPSINT8 Tensor Sparse

x2.80

Clock Speed ···	Clock Speed ···
Peak OPS 1.41 POPSINT4 Tensor Sparse	Peak OPS 3.96 POPSINT8 Tensor Sparse
Tensor FP8-16 352.8 TFLOPS FP8-16 Tensor Sparse 705.6 TFLOPS Tensor FP8-32 176.4 TFLOPS FP8-32 Tensor Sparse 352.8 TFLOPS	Tensor FP8-16 1.98 PFLOPS FP8-16 Tensor Sparse 3.96 PFLOPS Tensor FP8-32 1.98 PFLOPS FP8-32 Tensor Sparse 3.96 PFLOPS
Tensor FP16-16 176.4 TFLOPS FP16-16 Tensor Sparse 352.8 TFLOPS Tensor FP16-32 88.2 TFLOPS FP16-32 Tensor Sparse 176.4 TFLOPS	Tensor FP16-16 989.4 TFLOPS FP16-16 Tensor Sparse 1.98 PFLOPS Tensor FP16-32 989.4 TFLOPS FP16-32 Tensor Sparse 1.98 PFLOPS
BF16 44.1 TFLOPS Tensor BF16 88.2 TFLOPS BF16 Tensor Sparse 176.4 TFLOPS	BF16 133.8 TFLOPS Tensor BF16 989.4 TFLOPS BF16 Tensor Sparse 1.98 PFLOPS
Tensor TF32 44.1 TFLOPS	Tensor TF32 494.7 TFLOPS
FP32 44.1 TFLOPS	FP32 66.91 TFLOPS
FP64 689 GFLOPS Tensor FP64 -	FP64 33.45 TFLOPS Tensor FP64 61.84 TFLOPS
Tensor INT4 705.6 TOPS	Tensor INT4 -
Tensor INT8 352.8 TOPS	Tensor INT8 1.98 POPS
Ray 101.9 TOPS	Ray -
Pixel Rate 292.3 GPixel/s	Pixel Rate 47.5 GPixel/s
Texture Rate 689 GTexel/s	Texture Rate 1.04 TTexel/s

Shaders 8,448 Shaders	Shaders 16,896 Shaders
TMUs 264 TMUs	TMUs 528 TMUs
ROPs 112 ROPs	ROPs 24 ROPs
Tensor Cores 264 T-Cores	Tensor Cores 528 T-Cores
RT Cores 66 RT-Cores	RT Cores -
SMs 66 SMs	SMs 132 SMs

Base Clock 2.34GHz	Base Clock 1.06GHz
Boost Clock 2.61GHz	Boost Clock 1.98GHz
Tensor Clock -	Tensor Clock 1.83GHz

L1 64KB/SM Tex	L1 64KB/SM Tex
L2 Cache 64MB shared	L2 Cache 50MB shared

16GB GDDR6X	141GB HBM3e
Memory Bus 256-bit	Memory Bus 6144-bit
Memory Speed 22.4GT/s	Memory Speed 6.3GT/s
Memory Bandwidth 716.8GB/s	Memory Bandwidth 4.81TB/s
ECC No	ECC No

TDP 285W	TDP 700W
Max Temp 90°C Max	Max Temp -

Max Resolution 7680x4320	Max Resolution -
Refresh Rate Calculator 7680x4320 Resolution8K UHD Refresh Rate 60Hz	Refresh Rate Calculator -
Multi-Monitor 4	Multi-Monitor -
Adaptive Sync G-Sync FreeSync	Adaptive Sync G-Sync FreeSync
DSC Supported	DSC Not Supported
HDCP HDCP 2.3	HDCP -

3x DisplayPort 1.4

1x HDMI 2.1

-

Shader Model 6.6	Shader Model -
DirectX DirectX 12 Direct3D 12_3	-
OpenGL 4.6 OpenCL 3 Vulkan 1.3	OpenGL - OpenCL 3 Vulkan -
CUDA 8.9 PureVideo HD VP12 VDPAU Feature Set L	CUDA 9 PureVideo HD VP11 VDPAU Feature Set K

Encoder 2x NVENC 8	-
Codec AVC (H.264) HEVC (H.265) AV1	Codec -

Decoder NVDEC 5	Decoder 7x NVDEC 5
Codec MPEG-1 MPEG-2 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265) AV1	Codec MPEG-1 MPEG-2 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265) AV1

Form Factor PCIe Card	Form Factor SXM5
PCIe PCIe 4.0 x16 3-Slots	PCIe PCIe 5.0 x16 -
Height 137mm (5.39") Width 304mm (11.97") Depth 61mm (2.4")	-
Cooling Open-Air 2x Fans	Cooling Passive -
Power Connectors 1x 16-Pin 12VHPWR	Power Connectors -
Multi-GPU -	Multi-GPU Multi-GPU Support Supported Multi-GPU Type NVLink

Manufacturer NVIDIA	Manufacturer NVIDIA
Chip Designer NVIDIA	Chip Designer NVIDIA
Architecture Ada Lovelace	Architecture Hopper
Family GeForce 40	Family Server
Branding	Branding
Codename NV183	Codename NV180
Chip Variant AD103-275-A1	Chip Variant -
Market Segment Desktop	Market Segment Server
Release Date Jan 8, 2024	Release Date Nov 13, 2023

Foundry TSMC	Foundry TSMC
Fabrication Node 4N	Fabrication Node 4N
Die Size 379mm²	Die Size 814mm²
Transistor Count 45.9B	Transistor Count 80B
Transistor Density 121.2 MTr/mm²	Transistor Density 98.28 MTr/mm²