|
gptkbp:instance_of
|
gptkb:Graphics_Processing_Unit
|
|
gptkbp:api
|
gptkb:Open_GL
gptkb:Vulkan
gptkb:Direct_X
|
|
gptkbp:application
|
gptkb:Deep_Learning
gptkb:Data_Analytics
Scientific Computing
AI Inference
AI Training
|
|
gptkbp:architecture
|
gptkb:GCN_5.0
|
|
gptkbp:compatibility
|
gptkb:Linux
gptkb:Windows
|
|
gptkbp:connects
|
gptkb:Infinity_Fabric
|
|
gptkbp:cooling_system
|
Passive
|
|
gptkbp:die_size
|
495 mm²
|
|
gptkbp:double_precision_performance
|
6.1 TFLOPS
|
|
gptkbp:form_factor
|
gptkb:PCIe
Full-height, Full-length
|
|
gptkbp:gpu
|
gptkb:Vega_10
|
|
gptkbp:has_ability
|
FP16
INT8
FP32
FP64
BFloat16
|
|
gptkbp:has_units
|
gptkb:64
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
AMD Radeon Instinct MI series
|
|
gptkbp:launch_event
|
gptkb:SIGGRAPH_2017
|
|
gptkbp:launch_price
|
$1999
|
|
gptkbp:manufacturer
|
gptkb:AMD
|
|
gptkbp:market_segment
|
gptkb:machine_learning
High-Performance Computing
|
|
gptkbp:memory_type
|
gptkb:HBM2
2048-bit
|
|
gptkbp:performance
|
INT8
FP32
FP64
12.3 TFLOPS
|
|
gptkbp:power_consumption
|
300 W
|
|
gptkbp:ram
|
32 GB
1 TB/s
|
|
gptkbp:release_date
|
2017-08-30
|
|
gptkbp:release_year
|
gptkb:2017
|
|
gptkbp:successor
|
AMD Radeon Instinct MI series 2
|
|
gptkbp:supports
|
gptkb:ROCm
gptkb:Open_CL
|
|
gptkbp:target_market
|
gptkb:research_center
|
|
gptkbp:tensor_performance
|
24.6 TOPS
|
|
gptkbp:transistor_count
|
12.5 billion
|
|
gptkbp:video_output
|
gptkb:None
|
|
gptkbp:bfsParent
|
gptkb:NVIDIA_T4_GPUs
|
|
gptkbp:bfsLayer
|
6
|