--- title: "JetStream vs TransformerEngine" type: "comparison" canonical_url: "https://www.graphcanon.com/compare/ai-hypercomputer-jetstream-vs-nvidia-transformerengine" tools: ["ai-hypercomputer-jetstream", "nvidia-transformerengine"] --- # JetStream vs TransformerEngine *GraphCanon updated Aug 25, 2026* ## Verdict Pick JetStream if jetStream optimises throughput and memory for LLM inference on XLA devices like TPUs, with potential GPU support in future; pick TransformerEngine if transformerEngine optimizes Transformer model performance with FP8/FP4 precision on NVIDIA GPUs like Hopper, Ada, and Blackwell, boosting throughput and reducing memory usage. [JetStream](https://github.com/AI-Hypercomputer/JetStream) reports 455 GitHub stars, 67 forks, and 26 open issues, last pushed Jan 5, 2026. [TransformerEngine](https://docs.nvidia.com/deeplearning/transformer-engine/user-guide/index.html) has 3.5k stars, 795 forks, and 310 open issues, last pushed Aug 7, 2026. Figures are from public GitHub metadata via [JetStream's repository](https://github.com/AI-Hypercomputer/JetStream) and [TransformerEngine's repository](https://github.com/NVIDIA/TransformerEngine). | | [JetStream](/tools/ai-hypercomputer-jetstream.md) | [TransformerEngine](/tools/nvidia-transformerengine.md) | | --- | --- | --- | | Tagline | Throughput and memory optimized engine for LLM inference on XLA devices | A library for accelerating Transformer models on NVIDIA GPUs using low precision formats like FP8 and FP4. | | Stars | 455 | 3,479 | | Forks | 67 | 795 | | Open issues | 26 | 310 | | Language | Python | Python | | Adopt for | JetStream optimises throughput and memory for LLM inference on XLA devices like TPUs, with potential GPU support in future. | TransformerEngine optimizes Transformer model performance with FP8/FP4 precision on NVIDIA GPUs like Hopper, Ada, and Blackwell, boosting throughput and reducing memory usage. | | Persona | - | - | | Runtime | - | - | | License | Apache-2.0 | Apache-2.0 | | Categories | Inference & Serving | Inference & Serving, Model Training | ## Trust and health _Sourced signals - not a safety guarantee. No winner column._ | | [JetStream](/tools/ai-hypercomputer-jetstream.md) | [TransformerEngine](/tools/nvidia-transformerengine.md) | | --- | --- | --- | | Maintenance | Slowing (36%) | Very active (96%) | | Days since push | 231d | 0d | | Open issues (now) | 26 | 310 | | Stars delta | +4 (30d) | Unknown | | Open issues delta | +1 (30d) | Unknown | | Full report | [trust report](/tools/ai-hypercomputer-jetstream/trust.md) | [trust report](/tools/nvidia-transformerengine/trust.md) | ## Decision facts: JetStream - **Adopt for:** JetStream optimises throughput and memory for LLM inference on XLA devices like TPUs, with potential GPU support in future. ## Decision facts: TransformerEngine - **Adopt for:** TransformerEngine optimizes Transformer model performance with FP8/FP4 precision on NVIDIA GPUs like Hopper, Ada, and Blackwell, boosting throughput and reducing memory usage. ## Choose when ### Choose JetStream if… - Tags unique to JetStream: gemma, gpt, inference, large language models. - * You are working with large language models (LLMs) that require efficient inference on hardware supported by XLA, particularly TPUs. - Leaner open-issue backlog (26). ### Choose TransformerEngine if… - Tags unique to TransformerEngine: cuda, deep-learning, fp4, fp8. - Also covers Model Training. - If you need high-throughput training or inference of Transformer models specifically on compatible NVIDIA GPUs (Hopper, Ada, Blackwell). ## When NOT to use JetStream - * If your primary compute platform is not an XLA-compatible device such as TPU; JetStream's current focus is on systems that are supported by XLA. - * When you need immediate support for GPUs, since GPU functionality is marked as a future potential enhancement. ## When NOT to use TransformerEngine - Avoid if your project is not running on NVIDIA's Hopper, Ada, or Blackwell GPUs. - If memory usage isn't a critical concern and you prefer higher precision over speed optimization. ## Common questions ### What is the difference between JetStream and TransformerEngine? JetStream: Throughput and memory optimized engine for LLM inference on XLA devices. TransformerEngine: A library for accelerating Transformer models on NVIDIA GPUs using low precision formats like FP8 and FP4.. See the comparison table for live GitHub stats and shared categories. ### When should I choose JetStream over TransformerEngine? Choose JetStream over TransformerEngine when Tags unique to JetStream: gemma, gpt, inference, large language models; * You are working with large language models (LLMs) that require efficient inference on hardware supported by XLA, particularly TPUs; Leaner open-issue backlog (26). ### When should I choose TransformerEngine over JetStream? Choose TransformerEngine over JetStream when Tags unique to TransformerEngine: cuda, deep-learning, fp4, fp8; Also covers Model Training; If you need high-throughput training or inference of Transformer models specifically on compatible NVIDIA GPUs (Hopper, Ada, Blackwell). ### When should I avoid JetStream? * If your primary compute platform is not an XLA-compatible device such as TPU; JetStream's current focus is on systems that are supported by XLA. * When you need immediate support for GPUs, since GPU functionality is marked as a future potential enhancement. ### When should I avoid TransformerEngine? Avoid if your project is not running on NVIDIA's Hopper, Ada, or Blackwell GPUs. If memory usage isn't a critical concern and you prefer higher precision over speed optimization. ### Is JetStream or TransformerEngine more popular on GitHub? TransformerEngine has more GitHub stars (3,479 vs 455). Stars measure visibility, not whether either tool fits your constraints. ### Are JetStream and TransformerEngine open source? Yes - both are open-source projects on GitHub (JetStream: Apache-2.0, TransformerEngine: Apache-2.0). ### Where can I find alternatives to JetStream or TransformerEngine? GraphCanon lists graph-backed alternatives at [JetStream alternatives](/tools/ai-hypercomputer-jetstream/alternatives) and [TransformerEngine alternatives](/tools/nvidia-transformerengine/alternatives) ([JetStream markdown twin](/tools/ai-hypercomputer-jetstream/alternatives.md), [TransformerEngine markdown twin](/tools/nvidia-transformerengine/alternatives.md)), ranked by typed relationship edges rather than popularity votes. ### Is there a machine-readable version of this comparison? Yes. The markdown twin at [this comparison](/compare/ai-hypercomputer-jetstream-vs-nvidia-transformerengine.md) mirrors this page for agents and LLM crawlers, with the same stats table and FAQ answers. ### Which is better maintained, JetStream or TransformerEngine? JetStream: Slowing. TransformerEngine: Very active. Compare maintenance labels, days since push, and release cadence in the trust section below - stars alone do not measure maintenance. ### Where are the full trust reports for JetStream and TransformerEngine? GraphCanon publishes per-repo trust reports with dated maintenance, provenance, and scan summaries: [JetStream trust report](/tools/ai-hypercomputer-jetstream/trust); [TransformerEngine trust report](/tools/nvidia-transformerengine/trust). --- **Machine-readable endpoints** - JSON: [`/api/graphcanon/graph?tool=ai-hypercomputer-jetstream`](/api/graphcanon/graph?tool=ai-hypercomputer-jetstream) - LLM index: [/llms.txt](/llms.txt) - Full corpus: [/llms-full.txt](/llms-full.txt) _GraphCanon - The knowledge graph for AI development. https://www.graphcanon.com/_