ENTITY PROFILE / KNOWLEDGE GRAPH
FP8
Reflection BeamHAS_CHECKPOINT_VARIANTFP8 FP8DEPLOYED_ONml.p5.48xlarge FP8DEPLOYED_ONml.p5e.48xlarge FP8DEPLOYED_ONml.p5en.48xlarge FP8USES NVIDI a H100 GPUs FP8USES NVIDI a H200 GPUs The MPS backend supports the FP8 data type NVIDIA Nemotron 3.5 Transcribe is available in FP8 format· Entity summary generated by Graphiti
RELATIONS
8RelationsCURRENT
7CurrentEVENTS
3EventsREPORTS
3Source reportsNo public neighbors yet
Explore all published links of a node; previously discovered relations remain visible.
Relation timeline
BY VALID-FROM DATE- NVIDIA Nemotron 3.5 TranscribeAVAILABLE INFP8
NVIDIA Nemotron 3.5 Transcribe is available in FP8 format
- MPS backendSUPPORTSFP8
The MPS backend supports the FP8 data type
- FP8USESMagic Pointer
FP8USES NVIDI a H100 GPUs
- FP8DEPLOYED ONml.p5e.48xlarge
FP8DEPLOYED_ONml.p5e.48xlarge
- FP8USESStable Diffusion
FP8USES NVIDI a H200 GPUs
- Reflection BeamHAS CHECKPOINT VARIANTFP8
Reflection BeamHAS_CHECKPOINT_VARIANTFP8
- FP8DEPLOYED ONml.p5en.48xlarge
FP8DEPLOYED_ONml.p5en.48xlarge
- FP8DEPLOYED ONml.p5.48xlarge
FP8DEPLOYED_ONml.p5.48xlarge
Relations are facts Graphiti extracted from reports, not verbatim quotes. An end date means later reporting showed the relation no longer holds.
Events 3
Source reports 3
- Thin, Light and Built for Everything: NVIDIA RTX Spark Windows Laptops Gain New Models, Agent Tools and GamesNVIDIA Blog · first-party · 2026-10-28
- MPS in PyTorch 2.15Hugging Face Blog · first-party · 2026-10-20
- Introducing Reflection Beam: an open-weight model for agentic coding on AWSAWS Machine Learning Blog · first-party · 2026-10-16