pchardware.org
Categories

NVIDIA · 2014

NVIDIA K80 GPU accelerator

The NVIDIA K80 GPU accelerator is a dual-GPU computing card released in 2014, based on the Kepler architecture with 24 GB GDDR5 memory.

The NVIDIA K80 GPU accelerator is a dual-GPU computing card released on November 17, 2014, with a launch MSRP of $5,999. It is based on the Kepler architecture and uses two GK210 chips fabricated on a 28 nm process. Each GPU features 2496 shading units, totaling 4992 across both GPUs. The core operates at a base clock of 562 MHz and a boost clock of 824 MHz. The card is equipped with 24 GB of GDDR5 memory across a 768-bit bus, delivering a memory bandwidth of 480 GB/s. It has a board power (TDP) of 300 W and connects via a PCIe 3.0 x16 interface. The K80 is designed for high-performance computing workloads and is typically used in servers and workstations alongside compatible CPUs and motherboards.

Specifications

Architecture unverified Kepler
GPU chip unverified GK210
Process node unverified 28 nm
Shading units unverified 4992
Base clock unverified 562 MHz
Boost clock unverified 824 MHz
Memory size unverified 24 GB
Memory type unverified GDDR5
Memory bus unverified 768 bit
Memory bandwidth unverified 480 GB/s
Board power (TDP) unverified 300 W
Bus interface unverified PCIe 3.0 x16
Release date unverified 2014-11-17
Release year unverified 2014
Launch MSRP unverified 5999 USD