AI·ClickHouse scaled PgBouncer throughput 4x by running a fleet of processes sized to cores, using so_reuseport to balance connections on one port and inter-process peering for query cancellation.
Search
What are you looking for?
19 results for “performance”
AI·llama.cpp build b9966 fixes regex recompilations for -sm tensor mode, caching patterns per tensor to reduce CPU overhead on decode threads.
AI·Community discusses upgrades from Qwen3.6-27B, recommending options like DeepSeek V4 Flash or GLM-5.2 requiring 100-250GB VRAM for noticeable performance gains.
AI·MASTE: A Multi-Agent Pipeline for Zero-Shot Aspect Sentiment Triplet Extraction
AI·SQuaD-SQL uses LLM-guided synthetic data and LoRA fine-tuning to train 1.5B-parameter SLMs to reach 86.9% execution accuracy on WikiSQL, matching large models while requiring only one consumer GPU and delivering faster, lower-memory inference.
AI·Achieves 50-54 tok/s for Qwen3 30B A3B float8 on 16GB RTX 5060 Ti using custom CUDA/C++.
AI·User compares 5090 at 600W full compute and prompt processing vs shunt-modded 6000 PRO MaxQ water-cooled at 300-600W.
AI·Holographic Neural PCFG induces latent constituency trees from raw text using holographic memory and neural rule scoring for unsupervised parsing.
AI·TypeProbe recovers type representations from hidden states of pretrained code models using parallel Java and Python datasets.
AI·Reports EPYC 9374F CCD benchmarks using ik_llama.cpp; finds limited decoding gains vs older 9135 at low thread counts.
AI·Qwen3.6 35B-A3B Q8_0 model (no KV quant) generates full HTML flight simulator in single prompt and is praised as stronger than Q4_K_M on GPU due to better CPU performance.
More intelligence from every token, stronger performance per dollar, and more capability on demand for your hardest work.
Modern Video Object Segmentation (VOS) involves tracking and segmenting user-specified targets. While recent approaches have achieved remarkable performance in single-target scenarios, extending them to multi-target settings typically involves replicating the single-target processing for each individual object, resulting in reduced frame rates (FPS) with unbounded latency as target count increases. Built upon Segment Anything 2 (SAM2), we propose SAM-MT, which addresses this by transforming the…
Linear attention models allow a fixed state size and a fixed amount of compute per token. However, due to their limited state size, linear attention models fall behind in long-context recall compared to softmax-attention-based transformer architectures. Increasing the state size of linear attention improves recall performance but at the cost of higher FLOPs. In this work, we introduce Sparse Delta Memory (SDM), an architecture that scales the hidden state of gated linear RNNs to orders of magni…
Introducing GeneBench-Pro, a new benchmark testing AI performance in genomics, biology, and scientific research using complex, real-world datasets.
OpenAI and Broadcom introduce Jalapeño, a custom AI chip built for LLM inference to improve performance, efficiency, and scale across AI systems.
The growing demand for image-to-video creation on mobile devices has increasingly focused on cinematic motion effects like bullet time, dolly zoom, slow motion, etc. While Diffusion Transformers (DiTs) exhibit strong performance in video generation, their large parameter sizes and multi-step iterative denoising processes lead to substantial computational overhead, making efficient generation on mobile devices challenging. We propose CineMobile to bridge the gap. In particular, CineMobile adopts…
Pixel-wise Earth-observation (EO) foundation models are now achieving state-of-the-art performance via generated spatial embeddings. However, how these models scale and how best to spend a pretraining budget remain poorly understood. We present the largest controlled scaling study for EO to date: 395 training runs on 1,024 GH200 superchips within a fixed pixel-wise Barlow Twins family, each evaluated on 15 downstream tasks. We find that pretraining loss barely predicts downstream performance (|…
The inherent complexity of video understanding makes it difficult to determine whether Video-LLM benchmark performance stems from visual perception, linguistic reasoning, or knowledge priors. While many benchmarks have emerged to assess high-level reasoning, shared criteria for evaluating video understanding remain largely overlooked. Instead of introducing yet another benchmark, we take a step back to re-examine the criteria for evaluating video understanding. In this work, we introduce Video-…