NVIDIA CUDA Cores: 3072

Number of NVIDIA Tensor Cores: 384

NVIDIA RT Cores: 48

GPU Memory: 16 GB GDDR6

RTX-OPS: 62T

Raycasting: 8 Giga Rays/Sec

Maximum single precision (FP32) performance: 11.2 TFLOPS

Maximum single precision (FP16) performance: 22.3 TFLOPS

Highest integer arithmetic (INT8) performance: 178.4 TOPS

Deep Learning: TeraFLOPS1 89.2 TFLOPS

Video memory width: 256-bit

Video memory bandwidth: 448 GB/s

Maximum power consumption: 265 W

Bus: PCI Express 3.0 x16

Display Interface: DP 1.4 (4) + VirtualLink (1)

Dimensions: 4.4”H x 10.5”L Dual Slot

Weight: 972 g

Cooling scheme: Active

NVIDIA® 3D Vision® and 3D Vision Pro supported by 3 pin mini DIN

Frame Lock Compatible (with Quadro Sync II)

NVLink interconnect technology: 50 GB/s

Support platform

Microsoft Windows 10 (64-bit)

Microsoft Windows 8 and 8.1 (64-bit)

Microsoft Windows 7 (64-bit)

Linux® - Full OpenGL implementation, full NVIDIA and ARB extensions (64-bit)

3D Drawing Architecture

Scalable geometry architecture

Hardware tessellation engine

NVIDIA® GigaThreadTM engine and 5 asynchronous replication engines

Shader Model 5.1 (OpenGL 4.5 and DirectX 12)

Up to 32K x 32K texture and shading

Transparent Multisampling and Supersampling

16x angle-independent anisotropic filtering technology

32-bit texture filtering and color mixing technology with floating point capability

64x Panoramic Anti-Aliasing (FSAA) technology / SLI mode up to 128x

Hardware decoding acceleration supports MPEG-2, MPEG-4 Part 2 Advanced Simple Profile, H.264, HEVC, MVC, VC1, DivX (version 3.11 and later), and Flash (10.1 and later)

Proprietary H.264 and HEVC encoders2

Blu-ray dual-stream hardware acceleration (supports HD picture-in-picture playback)

NVIDIA GPU Boost (automatically increases GPU engine throughput to maximize application performance)

NVIDIA CUDA Parallel Processing Architecture

New RT (Ray Tracing) cores per SM

Turing SM Architecture (Streaming Multiprocessor Design for Higher Processing Efficiency)

Dynamic parallelism (the GPU dynamically spawns new threads without returning to the CPU)

Mixed precision (1, 4, 8, 16, 32 and 64 bit) calculations

API support includes: CUDA C, CUDA C++, DirectCompute 5.0, OpenCL, Java, Python, and Fortran

Error Correction Code (ECC) on graphics memory

Configurable up to 96 KB of memory (dedicated shared memory size per SM)

Display properties

Supports any combination of four monitors

Four DisplayPort 1.4 outputs (supports resolutions such as 3840 x 2160 @ 120 Hz, 5120x2880 @ 60Hz and 7680 x 4320 @ 60Hz)

DisplayPort to VGA, DisplayPort to DVI (single-link and dual-link) and DisplayPort to HDMI Flat Cables (supported resolutions depend on adapter specifications)

DisplayPort 1.4 (SMPTE 2084/2086, BT. 2020) (4K @ 60Hz 10b/12b HEVC decode, 4K @ 60Hz 10b HEVC encode) Support HDR

DisplayPort and HDMI connectors support HDCP 2.2

12-bit internal display pipeline (with supported panels, applications and connections, hardware supports 12-bit scanout)

Support for NVIDIA® 3D Vision™ technology, 3D DLP, interlacing, and other stereoscopic 3D specifications

Supports full OpenGL quad-buffer stereo technology

Underscan/Overscan Compensation and Hardware Scaling

NVIDIA® nView® Multi-Monitor Technology

Leverage the NVIDIA® SVS platform, including NVIDIA® Mosaic, NVIDIA® Sync, and NVIDIA® Warp/Blend technologies, to support large-scale and ultra-high-resolution visualizations

DisplayPort and HDMI digital sound

The following sound modes are supported:- Dolby Digital (AC3), DTS 5.1, Multichannel (7.1) LPCM, Dolby Digital Plus (DD+), and MPEG-2/MPEG-4 AAC

Supports DisplayPort data transfer rate 48 KHz

Supports 16-bit, 20-bit, and 24-bit size character groups

The Mini-DisplayPort connector is designed with a custom bayonet retention mechanism to securely hold the monitor cable.