|
gptkbp:instance_of
|
gptkb:Graphics_Processing_Unit
|
|
gptkbp:aiinference_performance
|
Up to 185 TOPS
|
|
gptkbp:application
|
gptkb:AI_Research
gptkb:machine_learning
gptkb:Data_Analytics
Scientific Computing
High Performance Computing
|
|
gptkbp:architecture
|
CDNA
|
|
gptkbp:cooling_system
|
Active Cooling
|
|
gptkbp:cudacompatibility
|
No
|
|
gptkbp:direct_xsupport
|
No
|
|
gptkbp:features
|
High Performance Computing
|
|
gptkbp:form_factor
|
PCIe Card
Full Height, Full Length
|
|
gptkbp:gpu
|
gptkb:Yes
|
|
gptkbp:gpuarchitecture
|
CDNA
|
|
gptkbp:has_ability
|
CDNA 1.0
|
|
gptkbp:has_units
|
120
7680
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
AMD MI100
|
|
gptkbp:launch_price
|
$6,999
|
|
gptkbp:manufacturer
|
gptkb:AMD
|
|
gptkbp:manufacturing_process
|
7 nm
|
|
gptkbp:market_position
|
High-End
|
|
gptkbp:market_segment
|
gptkb:enterprise_solutions
|
|
gptkbp:memory_type
|
gptkb:HBM2
4096-bit
|
|
gptkbp:open_glsupport
|
gptkb:Yes
|
|
gptkbp:opened
|
gptkb:Yes
|
|
gptkbp:peak_fp32_performance
|
11.5 TFLOPS
|
|
gptkbp:peak_fp64_performance
|
11.5 TFLOPS
|
|
gptkbp:performance
|
11.5 TFLOPS
23 TFLOPS
|
|
gptkbp:power_connector
|
8-pin and 6-pin
|
|
gptkbp:ram
|
32 GB
1.2 TB/s
|
|
gptkbp:ray_tracing_support
|
No
|
|
gptkbp:release_date
|
November 2020
|
|
gptkbp:released
|
MI Series
|
|
gptkbp:resolution
|
7680 x 4320
|
|
gptkbp:slisupport
|
gptkb:Yes
|
|
gptkbp:successor
|
AMD MI200
|
|
gptkbp:supports
|
gptkb:machine_learning
PCIe 4.0
AI Workloads
|
|
gptkbp:target_market
|
Data Centers
|
|
gptkbp:tdp
|
300 W
|
|
gptkbp:thermal_design_power
|
300 W
|
|
gptkbp:video_output
|
1 x HDMI 2.1
4 x Display Port 1.4
|
|
gptkbp:virtualization_support
|
gptkb:Yes
|
|
gptkbp:bfsParent
|
gptkb:NVIDIA_A10_GPUs
|
|
gptkbp:bfsLayer
|
6
|