|
gptkbp:instanceOf
|
gptkb:microprocessor
|
|
gptkbp:announced
|
2019
|
|
gptkbp:architecture
|
gptkb:POWER_ISA_v3.1
|
|
gptkbp:coreCount
|
up to 15 per chip
|
|
gptkbp:designedBy
|
gptkb:IBM
|
|
gptkbp:firstSystemRelease
|
gptkb:IBM_Power_E1080
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Power10
|
|
gptkbp:instructionSet
|
gptkb:POWER10_ISA
|
|
gptkbp:L2Cache
|
120 MB
8 MB per core
|
|
gptkbp:manufacturer
|
gptkb:Samsung_Electronics
|
|
gptkbp:maxThreadsPerCore
|
8
|
|
gptkbp:predecessor
|
Power9
|
|
gptkbp:processNode
|
7 nm
|
|
gptkbp:releaseDate
|
2021
|
|
gptkbp:size
|
602 mm²
|
|
gptkbp:supports
|
gptkb:Red_Hat_Enterprise_Linux
gptkb:AIX
gptkb:SUSE_Linux_Enterprise_Server
gptkb:CAPI_2.0
gptkb:IBM_i
gptkb:OpenCAPI
gptkb:PowerVM_virtualization
gptkb:DDR5
gptkb:NVLink_4.0
gptkb:PCIe_5.0
gptkb:SMT4
gptkb:SMT8
gptkb:Secure_Boot
gptkb:DDR4
gptkb:Simultaneous_Multithreading
AI workloads
Big Endian mode
Little Endian mode
Trusted Platform Module
RAS features
Transactional Memory
Dynamic execution
Speculative execution
hardware memory encryption
matrix math acceleration
Open Memory Interface (OMI)
PowerAXON interconnect
Accelerated AI inference
CAPI 3.0
CAPI 4.0
CAPI 5.0
SMT1
SMT2
|
|
gptkbp:transistorCount
|
18 billion
|
|
gptkbp:usedIn
|
gptkb:IBM_Power_Systems_servers
|
|
gptkbp:bfsParent
|
gptkb:Power_microprocessors
|
|
gptkbp:bfsLayer
|
7
|