EVGA RTX 3070 FTW3 ULTRA LHR

Graphics Card EVGA RTX 3070 FTW3 ULTRA LHR — Specifications, Tests & Benchmarks
NVIDIA

EVGA RTX 3070 FTW3 ULTRA LHR

Nvidia GeForce 30 GA104 01.09.2020 8 GB GDDR6 256 bit 220W
🏆 Top #1850
Performance
Synthetic Score
26,014
PassMark
22,139
Efficiency
high
0 перегляди
0 порівняння
0 гри
0 користувачі
0 коментарі
0 закладки
The EVGA RTX 3070 FTW3 ULTRA LHR (Ampere) model joined the NVIDIA lineup of graphics accelerators on 01.09.2020. The graphics processor is manufactured using a modern process. The 392 mm² die contains approximately 17.4 billion transistorsistors transistors. The model's computational arsenal includes 5888 universal processors. Also available are 46 specialized cores for ray rendering and 184 cores for accelerating artificial intelligence. The GDDR6 video memory capacity is 8 GB. Thanks to the 256 bit-bit bus, bandwidth reaches 448.00 GB/s, eliminating delays when working with demanding textures. In standard mode, the core operates at 1500 MHz, but auto-overclocking technology allows it to reach 1815 MHz under load. The card's power consumption (TDP) is stated at 220 W. The manufacturer recommends using a power supply unit with a wattage of at least 550 W.

General Information

Release Date 01.09.2020
Usage Class Hardcore Gaming
Market Segment Desktop
Series Nvidia GeForce 30
GPU Variant GA104-300-A1

GPU

Architecture Ampere
Codename GA104
Process Node 8 nm (Samsung)
Transistor Count 17,400,000.0 million
Die Size 392 mm²
Transistor Density 44,400,000 million / mm²

Rendering Configuration

Shading Units (ALU) 5888
Texture Mapping Units (TMU) 184
Render Output Units (ROP) 96
Tensor Cores 184
RT Cores (Ray Tracing) 46
Compute Units (SM/CU) 46

Clock Speeds

Base Clock 1500 MHz
Boost Clock 1815 MHz
Memory Clock (Actual) 1750 MHz
Memory Clock (Effective) 14000 MHz

Memory

Memory Capacity 8 GB
Memory Type GDDR6
Memory Bus 256 bit bit
Memory Bandwidth 448.00 GB/s

Cache Memory

L1 Cache 5888 KB
L2 Cache 4 MB

Board Design & Power

Interface PCIe 4.0 x16
Slot Width 3
Dimensions (L x W x H)
Power Consumption (TDP) 220 W
Recommended PSU 550 W
Power Connectors 2x 8-pin
Board Number PG142 SKU 10

API Support

DirectX 12 Ultimate (12_2)
OpenGL 4.6
OpenCL
Vulkan 1.4
CUDA
Shader Model

Theoretical Performance

Pixel Fill Rate 174.2 GPixel/s
Texture Fill Rate 334 GTexel/s
FP16 Performance (Half) 21.37 TFLOPS (1:1)
FP32 Performance (Float) 21.37 TFLOPS
FP64 Performance (Double) 334 GFLOPS (1:64)

Synthetic Tests & Benchmarks

PassMark (G3D Mark) 22 139
PassMark (G2D Mark) 1 004
TPU Index Score 32.54%
Performance Tier high

How this graphics card performs in games

Всі тести
Тестів поки немає.

Comments

0

Please log in to your account to leave a comment.

Log In or Register

No comments yet. Be the first to comment!

Compare with another graphics card

Швидка навігація

We Use Cookies 🍪

This site uses cookies to enhance your user experience. By continuing to browse, you agree to our Privacy Policy.
By continuing to use this site, you agree to our Terms of Service and Privacy Policy.