# Ember changes data placement when workloads exceed GPU memory

> Researchers at the Technical University of Munich developed Ember, a memory manager that tracks which data should remain on a GPU when its memory fills. In the paper’s data-intensive workloads, it reported higher speeds than standard Unified Virtual Memory on three NVIDIA GPUs.

- Canonical URL: https://newsverum.com/en/news/ember-unified-memory-oversubscribed-gpus-2026
- Publisher: Newsverum
- Section: Technology
- Country: Germany
- Published: 2026-10-13T01:40:12+00:00
- Updated: 2026-10-13T01:40:12+00:00
- Source review period: 2026-10-09
- Other editions: [Türkçe](https://newsverum.com/tr/news/ember-unified-memory-oversubscribed-gpus-2026.md), [العربية](https://newsverum.com/ar/news/ember-unified-memory-oversubscribed-gpus-2026.md), [Français](https://newsverum.com/fr/news/ember-unified-memory-oversubscribed-gpus-2026.md), [Español](https://newsverum.com/es/news/ember-unified-memory-oversubscribed-gpus-2026.md), [Deutsch](https://newsverum.com/de/news/ember-unified-memory-oversubscribed-gpus-2026.md), [Português](https://newsverum.com/pt/news/ember-unified-memory-oversubscribed-gpus-2026.md)

*The source does not state the event date; the period shown is the source review period.*

Applications that exceed GPU memory must move data between host and device memory. Standard page migration can slow down when it repeatedly moves data that is not useful for the current workload. The adaptive Ember variant maps regions to host memory by default, promotes frequently accessed regions using GPU hardware counters, and reclaims colder regions before device memory fills.

For the static Ember variant at about 1.5× memory oversubscription, the authors report geometric-mean speedups over stock UVM across six data-intensive applications of 9.12× on RTX 5080, 6.83× on A40 and 10.62× on H200. They also tested Qwen2.5 14B and Qwen3 32B generation with GPUs capped at 16 GiB. These are results for the named hardware and workloads, not a general production-performance guarantee. Some baseline UVM runs hit the 1,800-second cap, so speedups in affected comparisons are lower bounds.

## Data

### GPU memory experiment results

Speedups apply only to the GPUs and workloads specified in the paper.

Methodology: Values are geometric-mean speedups of the static Ember variant over stock UVM across six applications, measured on three GPUs at about 1.5× memory oversubscription. Some baseline UVM runs hit the 1,800-second cap, so speedups in affected comparisons are lower bounds.

| GPU count | Data-intensive workload count | Reported result |
|---|---|---|
| 3 | 6 | Geometric-mean speedup: static Ember / stock UVM |

Sources: <https://arxiv.org/abs/2610.13408> · Download data: [JSON](https://newsverum.com/api/v1/articles/ember-unified-memory-oversubscribed-gpus-2026/data/gpu-memory-benchmarks?lang=en&format=json) · [CSV](https://newsverum.com/api/v1/articles/ember-unified-memory-oversubscribed-gpus-2026/data/gpu-memory-benchmarks?lang=en&format=csv)

## Sources

- [arXiv](<https://arxiv.org/abs/2610.13408>) (Source checked: 2026-10-13; Published: 2026-10-09)

---

Cite the canonical URL above. [Editorial standards and corrections](https://newsverum.com/en/editorial)