|
gptkbp:instance_of
|
gptkb:NVIDIA
|
|
gptkbp:apisupport
|
gptkb:Direct_X_11
CUDA 5.0
Open GL 4.2
|
|
gptkbp:architecture
|
Unified Shader Architecture
|
|
gptkbp:architecture_release
|
gptkb:Kepler_architecture
|
|
gptkbp:clock_speed
|
6000 MHz
|
|
gptkbp:colorway
|
up to 2880
|
|
gptkbp:developed_by
|
gptkb:NVIDIA
|
|
gptkbp:die_size
|
294 mm²
|
|
gptkbp:end_of_life
|
gptkb:2016
|
|
gptkbp:features
|
Dynamic Parallelism
GPU Boost
Hyper-Q
|
|
gptkbp:fuel_economy
|
Improved power efficiency
|
|
gptkbp:has_ability
|
gptkb:3.0
|
|
gptkbp:has_units
|
up to 48
up to 240
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
NVIDIA Kepler
|
|
gptkbp:launch_event
|
gptkb:NVIDIA_GPU_Technology_Conference_2012
|
|
gptkbp:launch_price
|
$499
|
|
gptkbp:legacy
|
Backward compatible with Fermi
|
|
gptkbp:manufacturing_process
|
28 nm
|
|
gptkbp:market_position
|
High-end GPU
|
|
gptkbp:memory_type
|
256-bit
GDDR5
|
|
gptkbp:notable_products
|
gptkb:Quadro_K6000
gptkb:Ge_Force_GTX_770
gptkb:Ge_Force_GTX_780
gptkb:Ge_Force_GTX_680
gptkb:Ge_Force_GTX_Titan
gptkb:Tesla_K20
gptkb:Tesla_K40
|
|
gptkbp:performance
|
4.5 TFLOPS
20% over Fermi
|
|
gptkbp:population_density
|
48 GPixel/s
192 GTexel/s
|
|
gptkbp:power_consumption
|
up to 250 W
|
|
gptkbp:predecessor
|
gptkb:NVIDIA_Fermi
|
|
gptkbp:provides_support_for
|
NVIDIA drivers
|
|
gptkbp:ram
|
4 GB
192 GB/s
|
|
gptkbp:release_date
|
gptkb:2012
|
|
gptkbp:successor
|
gptkb:NVIDIA_Pascal
gptkb:NVIDIA_Maxwell
NVIDIA Pascal in 2016
|
|
gptkbp:target_market
|
Gaming and professional graphics
|
|
gptkbp:thermal_design_power
|
up to 300 W
|
|
gptkbp:transistor_count
|
over 3 billion
|
|
gptkbp:used_in
|
gptkb:Quadro_K_series
gptkb:Ge_Force_GTX_600_series
gptkb:Ge_Force_GTX_700_series
gptkb:Tesla_K_series
|
|
gptkbp:bfsParent
|
gptkb:NVIDIA_K20_X_GPUs
|
|
gptkbp:bfsLayer
|
5
|