Match hardware, precision, sequence length, concurrency and output quality. Report p50/p95 latency and total accepted-task cost, not only a peak throughput number.
Compute & efficiency · primary_technical_publication
Teams building AI applications need a repeatable path from a model configuration to a working inference endpoint. NVIDIA Dynamo is a distributed inference...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · preprint
Modern diffusion models are trained on massive datasets that often include copyrighted or private images. Membership inference attacks (MIAs) address this by testing whether a
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · preprint
Building a language model of one's own no longer requires a data centre. This tutorial trains one from scratch, tokenizer included, on one consumer GPU,
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · preprint
Looping in transformer models reuses weights by repeatedly applying the same transformer, or parts of it, making the model deeper without increasing its parameter count.
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · preprint
Prefill-decode (PD) disaggregation isolates the two phases of large language model serving, but request-level path selection must account for prefill work, cache reuse, and decode
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
Many AI workloads need only a fraction of a GPU. Giving each one an entire device can leave capacity idle; sharing GPUs lets platform teams
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
AI Factories are some of the most technologically complex operational environments ever created. Hundreds of thousands of GPUs, NVIDIA Spectrum-X Ethernet...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
Training large-scale AI models on thousands of GPUs is a high-stakes operation where monitoring and automation must be built-in to optimize cluster resiliency...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
Deploying a trained neural network involves more than translating each operator into a GPU kernel. It is a compilation problem with many legitimate answers. For...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
GPU applications are processing ever-growing datasets with increasingly powerful accelerators. According to the 2025 Stanford AI Index, compute for notable AI...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
AI developers aren’t just running a single model anymore—they’re orchestrating multiple models in parallel, chaining inference calls, and experimenting...
Automatically screened · not peer reviewed by this siteRead analysis note ↗Compute & efficiency · primary_technical_publication
Protein design is a search problem. Choose amino acids for an eight-residue protein sequence and more than 25 billion combinations are possible. Choose amino...
Automatically screened · not peer reviewed by this siteRead analysis note ↗