Statements (23)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:artificial_intelligence_accelerator
|
| gptkbp:application |
gptkb:machine_learning
deep learning |
| gptkbp:connects |
100 Pb/s fabric
|
| gptkbp:coreCount |
400,000
|
| gptkbp:countryOfOrigin |
gptkb:United_States
|
| gptkbp:firstReleased |
2019
|
| gptkbp:manufacturer |
gptkb:Cerebras_Systems
|
| gptkbp:notableFor |
largest chip ever built
single wafer design |
| gptkbp:onChipRAM |
18 GB SRAM
|
| gptkbp:performanceContext |
9 petaflops (FP16)
|
| gptkbp:powerSource |
15 kW
|
| gptkbp:size |
46,225 mm²
|
| gptkbp:successor |
gptkb:Cerebras_Wafer_Scale_Engine_2
|
| gptkbp:technology |
16nm
|
| gptkbp:transistorCount |
1.2 trillion
|
| gptkbp:usedIn |
Cerebras CS-1 system
Cerebras CS-2 system |
| gptkbp:waferDiameter |
300 mm
|
| gptkbp:bfsParent |
gptkb:CS-1_supercomputer
|
| gptkbp:bfsLayer |
8
|
| https://www.w3.org/2000/01/rdf-schema#label |
Cerebras Wafer Scale Engine
|