Offline batch inference is gaining popularity as a cost-effective solution for latency-insensitive tasks, such as model evaluation and data curation. As the latency objective is highly relaxed, maximizing throughput is the primary goal in offline inference. Previous studies focused solely on throughput optimization within a batch. However, the diverse resource demands (compute-intensive vs. memory-intensive) across a wide range of applications make these approaches less effective, as imbalanced resource demands between batches restrict optimization opportunities.

Datacenters often underutilize expensive AI accelerators (GPUs, TPUs, etc). A natural solution is disaggregation, where servers borrow network-attached accelerators on demand. However, current approaches to disaggregation suffer from a semantic translation gap: as computation descends the software stack, critical application knowledge—like model structure or execution phases—is lost. This forces an undesirable choice between low-level, general-purpose systems that are semantically-blind and inefficient, and high-level, single-workload systems that are efficient but not general.

Besides typical generative applications, like ChatGPT, GitHub Copilot, and Cursor, we observe an emerging trend that LLMs are increasingly used in traditional discriminative tasks, such as recommendation, credit verification, and data labeling. The key characteristic of these emerging use cases is that the LLM generates only a single output token, rather than an arbitrarily long sequence of tokens. We refer to this as a prefill-only workload. However, since existing LLM engines assume arbitrary output lengths, they fail to leverage the unique properties of prefill-only workloads. In this paper, we present PrefillOnly, the first LLM inference engine that improves the inference throughput and latency by fully embracing the properties of prefill-only workloads. First, since it generates only one token, PrefillOnly only needs to store the KV cache of only the last computed layer, rather than of all layers. This drastically reduces the GPU memory footprint of LLM inference and allows handling long inputs without using solutions that reduce throughput, such as cross-GPU KV cache parallelization. Second, because the output length is fixed, rather than arbitrary, PrefillOnly can precisely determine the job completion time (JCT) of each prefill-only request before it starts. This enables efficient JCT-aware scheduling policies such as shortest prefill first. PrefillOnly can process up to 4× larger queries per second without inflating the average and P99 latency.

Even with substantial endeavors to test and validate processors, computational errors may still arise post-installation. One particular category of CPU errors transpires discreetly, without crashing applications or triggering hardware warnings. These elusive errors pose a significant threat by undermining user data, and their detection is challenging. This paper introduces Orthrus, a solution for the timely detection of silent user data corruption caused by post-installation CPU errors. Orthrus safeguards user data in cloud applications by providing simple annotations and compiler support for users to identify data operators and validating these operators asynchronously across cores while maintaining a low overhead (2%–6%), making it practical for production deployment. Our evaluation, using carefully injected errors, demonstrates that Orthrus can detect 87% of data corruptions with just a single core dedicated to validation, increasing to 91% and 96% when two and four cores are used, respectively.

Resource disaggregation has gained much traction as an emerging datacenter architecture, as it improves resource utilization and simplifies hardware adoption. Under resource disaggregation, different types of resources (memory, CPUs, etc.) are disaggregated into dedicated servers connected by high-speed network fabrics. Memory disaggregation brings efficiency challenges to concurrent garbage collection (GC), which is widely used for latency-sensitive cloud applications, because GC and mutator threads simultaneously run and constantly compete for memory and swap resources.

Fault tolerance is one of the major design goals for HPC. The emergence of non-volatile memories (NVM) provides a solution to build fault tolerant HPC. Data in NVM-based main memory are not lost when the system crashes because of the non-volatility nature of NVM. However, because of volatile caches, data must be logged and explicitly flushed from caches into NVM to ensure consistence and correctness before crashes, which can cause large runtime overhead. In this paper, we introduce an algorithm-based method to establish crash consistence in NVM for HPC applications. We slightly extend application data structures or sparsely flush cache blocks, which introduce ignorable runtime overhead. Such extension or cache flushing allows us to use algorithm knowledge to reason data consistence or correct inconsistent data when the application crashes. We demonstrate the effectiveness of our method for three algorithms, including an iterative solver, dense matrix multiplication, and Monte-Carlo simulation. Based on comprehensive performance evaluation on a variety of test environments, we demonstrate that our approach has very small runtime overhead (at most 8.2% and less than 3% in most cases), much smaller than that of traditional checkpoint, while having the same or less recomputation cost.