CS Paper Radar

Beyond Bitwidth: Deployment-Constrained Mixed-Precision Quantization for Multilingual LLM Inference on NVIDIA Blackwell

Hyojung Han

cs.IR cs.CLcs.LG ·submitted 2026-11-16 22:58 UTC · v1

Abstract

Low-bit LLM quantization is useful only if a model preserves service-relevant accuracy, executes through supported GPU kernels, and meets measured memory and latency limits. We study Qwen3-8B on NVIDIA B200 with vLLM, including INT2/INT3, GPTQ, NVFP4, and mixed-precision configurations. We verify native packed execution and precision on all 144 fused linear modules and evaluate Korean, English, numerical, and reasoning quality under preregistered gates. Packed INT2 and INT3 execute, but all seven tested INT2 grids degenerate on generation. Isolated module-restoration effects are strongly non-additive; Korean and reasoning sensitivity rankings disagree, and a promising INT4/INT8 development candidate fails independent final evaluation. A configuration with 16-bit activations, 98 NVFP4 weight-only modules, 46 INT8 modules, and an INT8 output head passes all gates on the untouched FINAL-81 set (5,168 items): 5.917 GiB of weight-loading memory (limit 5.98), Korean teacher-correct retention 0.904 (gate 0.90), and TPOT ratio 0.939 relative to BF16. It improves paired accuracy over NVFP4 W4A4 by +1.4 percentage points (Bonferroni-adjusted interval [+0.2, +2.6]), compared with +3.4 points on the selection set. This comparison changes both activation precision and INT8 protection; exploratory analyses suggest that the activation change contributes substantially, but the isolated allocation effect is not established. Final checkpoints were not retained, and bitwise rebuild reproducibility remains unverified.

Comment
18 pages, 2 figures

Closest papers

Nearest neighbours by meaning across the whole index.

AlignQuant: Tile-Aligned Mixed-Precision Quantization for Efficient LLM Generation

cs.LG cs.AIcs.CL ·2026-10-05 ·Hanzhi Zhang, Qiao Zhang, Qinglei Cao, Heng Fan, Yan Huang +2 more
code/data similarity 0.72
Abstract

Fine-grained mixed-precision quantization promises efficient large language model inference, but local precision choices can conflict with regular GPU storage and computation units. This precision-boundary mismatch limits the translation of compression into practical acceleration. We introduce AlignQuant, a post-training quantization method that uses GPU-compatible two-dimensional weight tiles as the common unit of precision allocation, compact storage, and execution. This shared partition lets precision follow sensitivity within output channels. Joint prefill/decode calibration scores precision reductions using projection-output perturbations weighted by language-model loss gradients under quantized activations. Phase-normalized scores prioritize higher precision for tiles important to either phase under a model-wide weight-storage budget. Each tile stores one selected representation, while phase-specialized kernels reuse the packed model and expand lower-bit weights for INT8 computation with 8-bit activations. Across four LLMs spanning 3B to 14B parameters, AlignQuant achieves up to $2.50\times$ generation speedup over BF16 while preserving model quality. Evaluations further cover three GPUs and contexts up to 64K tokens. These results show that local precision flexibility and regular GPU execution can coexist through a shared tile unit. The implementation is available at https://github.com/HanzhiZhang-Ulrica/AlignQuant.

0.34

Compression Effects on Low-Resource Multilingual Translation

cs.CL ·2026-11-18 ·Abbas Sani
code/data similarity 0.68
Abstract · `Pilot study. 3 model families x 3 precisions x 6 directions on FLORES-200. Code and raw results: https://github.com/avvas200/efficient-mult…

We present a pilot study characterising the effect of post-training quantisation on multilingual machine translation quality and inference cost for three African languages - Hausa, Swahili, and Yoruba - across three model families: a small general-purpose LLM (Qwen2.5-1.5B), a larger multilingual instruction-tuned LLM (Aya-23-8B), and a dedicated multilingual MT model (NLLB-200-distilled-1.3B). Evaluating on the FLORES-200 devtest set (1,012 sentences per direction) under FP16, bitsandbytes INT8, and NF4 INT4 precisions on consumer-grade NVIDIA T4 GPUs, we find: (i) the specialised MT model outperforms both LLMs by 32.7-43.3 chrF on the Hausa and Swahili directions, and by 12.9-24.7 chrF on Yoruba, despite being roughly a sixth the size of the 8B instruction-tuned LLM; (ii) post-training quantisation preserves aggregate translation quality within approximately 1 chrF across all model families and language directions, with a single worst case of 2.19 chrF; (iii) on consumer GPUs, INT8 and INT4 quantisation provide substantial memory savings (49-61% at INT4) but do not reduce inference latency relative to FP16 - in fact, latency increases by 1.3x to 5.3x; (iv) a clear performance gradient tracking training-resource availability is visible across languages (Swahili > Hausa > Yoruba). These findings motivate a research agenda focused on hardware-aware compression and error-type analysis beyond aggregate metrics for low-resource deployment.

0.27

Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference

cs.LG cs.AIcs.CY ·2026-10-13 ·Petar Basta
similarity 0.67
Abstract

While it is known that quantization of large language models (LLMs) reduces memory usage via lower-bit weights, studies quantifying the resulting impact on energy usage and carbon intensity are scarce. We present the first unified evaluation of weight-only quantization strategies for LLM inference across varying input/output lengths and GPUs, highlighting differences in energy efficiency in addition to accuracy degradation and runtime. We study quantized inference on Llama 2 7B, Phi 3.5 Mini Instruct, OLMo 1B, Qwen 2.5 7B, and Qwen 2.5 14B across GLUE MNLI, MMLU, HumanEval, and GSM8K datasets, evaluating 10 post-training quantization (PTQ) strategies on NVIDIA H100, A6000, A100, and L4 GPUs. We identify a novel trend showing that quantization techniques tend to exhibit peak energy efficiency relative to full-precision baselines when inputs are sufficiently long and outputs are short. Furthermore, we show that quantization strategies become marginally more energy-efficient relative to full-precision models as batch sizes increase, though gains are modest. Notably, fused-kernel implementations such as EETQ int8 and Bitsandbytes int8 offer the highest energy savings, up to 4x compared to FP32 on short text generation, with negligible accuracy loss. Finally, we observe that energy usage closely tracks runtime on our evaluated benchmarks, indicating that, in practice, latency optimization can serve as a reliable proxy for reducing the environmental footprint of LLM services. We conclude with suggested directions for strategically selecting low-carbon quantization strategies based on specific inference requirements.

0.10

BitLLM: Efficient Low-Bit, Conversion-Hiding LLM Decoding on Mobile GPUs

cs.AR cs.LGcs.OScs.PF ·2026-10-31 ·Hangyu Zheng, Zhihao Shu, Kunxiong Zhu, Minghai Qin, Xiaolong Ma +2 more
top venue similarity 0.67
Abstract

The rapid scaling of large language models (LLMs) has unlocked remarkable capabilities across various domains. However, deploying LLMs on mobile devices remains challenging because of their high computational and memory requirements. Low-bit quantization (e.g., 4-bit weights, 8-bit activations) can reduce model size while preserving accuracy, but its practical efficacy on mobile GPUs has been limited so far. In this paper, we identify that the performance gap stems primarily from frequent data type conversions between quantized and floating-point values, and dynamic and inefficient memory access patterns caused by the growing Key-Value (KV) cache during autoregressive decoding. We introduce BitLLM, a mobile inference engine that addresses type conversion overhead and suboptimal memory accesses in quantized LLMs. Our approach features an adaptive layout and packing strategy for the KV cache that ensures efficient memory access throughout LLM decoding. This is coupled with an integer-native execution pipeline that defers and overlaps type conversion outside the compute hot path, keeping the critical MatMul stages in the integer domain. Implemented on off-the-shelf mobile GPUs, BitLLM outperforms leading engines, i.e., LLAMA.CPP, MNN-LLM, LiteRT-LM, and MLC-LLM, achieving average speedups of 1.78$\times$, 1.73$\times$, 1.58$\times$, and 1.47$\times$, respectively, across various models and quantization levels. Under long-context generation, BitLLM reduces average energy consumption (per token) by up to 88.9%.

0.57

StrataFlux: Memory-Hierarchy-Native Low-Bit LLM Inference

cs.LG cs.DC ·2026-11-05 ·Jake Hyun, Ziyu Zhao, Jae-sun Seo, Mohamed S. Abdelfattah
similarity 0.67
Abstract · 34 pages, 18 figures, 44 tables. Includes appendix

Quantization reduces the memory and bandwidth demands of large language model (LLM) inference but can degrade accuracy; offloading higher-precision data can instead stall generation. We introduce StrataFlux, a framework that challenges this tradeoff by recovering accuracy lost to low-bit weight and key-value (KV) cache quantization while retaining quantization's memory and speed advantages. StrataFlux co-designs refinement representation, selection, and delivery to make host RAM and solid-state drives useful sources of additional precision. Its selective precision streaming reduces transfer volume, overlaps delivery with computation, and reuses fetched precision across decoding steps so generation need not wait for every transfer. We evaluate 2-4-bit weights and 2-3-bit KV-cache on dedicated GPUs and unified-memory platforms. For 2-bit Llama-3.1-8B-Instruct weights, StrataFlux reduces WikiText-2 perplexity from 9.46 to 8.40 while maintaining 95.4% of the quantized base's throughput on RTX 5090. With 2-bit weights and KV, StrataFlux raises Qwen3-8B's AIME 2024/2025 accuracy from 11.0% to 47.1% at unchanged GPU memory use. At 128K context, Llama-3.1-8B-Instruct with 3-bit weights and 2-bit KV achieves comparable perplexity and speed to vLLM's 4-bit AWQ weights and 8-bit KV on RTX 5090, using approximately half the GPU memory. These results show that GPU-resident precision need not set the accuracy ceiling, nor offloading-tier latency dictate decoding speed.

0.20

SemKV: Semantic Mixed-Precision KV Cache Quantization Guided by the Quality Cliff for Long-Context LLM Inference

cs.LG cs.CLcs.IT ·2026-08-28 ·Daeha Lee, Do-Hyung Kim, Jae-Hong Kim
similarity 0.66
Abstract · 34 pages, 19 figures

The key-value (KV) cache is the dominant memory bottleneck of long-context large language model (LLM) inference, growing linearly with context length. We show that uniform KV quantization on a fractional-bit grid does not degrade gracefully: under a prespecified multi-seed statistical protocol, Llama-3.1-8B-Instruct with an affine quantizer is statistically indistinguishable from FP16 KV down to 2.322 code bits/value and collapses at 2.0 bits - a quality cliff in (2.0, 2.322] that reappears in generation-time quantization and multi-turn dialogue and transfers to Mistral-7B. The cliff reframes importance-aware mixed precision: above it, eight model-internal importance indicators are statistically interchangeable, so the benefit of mixing is grid interpolation, reaching average precisions uniform quantization cannot realize. SemKV preserves every token, ranks tokens by a model-internal score, and assigns two adjacent above-cliff precisions, achieving a measured 6.0x storage reduction with no statistically detectable quality difference from full KV (n=900, three seeds), and outperforming FP16 token pruning granted a 1.5x larger memory budget. Replacing the affine base with a distortion-optimized quantizer (TurboQuant-MSE) lowers the cliff in every protocol tested, raising the no-detectable-loss operating point to 7.9x. The recipe: measure the cliff for the target deployment setting, then interpolate above it.

0.22

Edge-Native Multimodal RAG: Hardware-Aware Quantization and Architectural Optimization for Consumer GPUs

cs.AI ·2026-11-11 ·Zahraa Selim, Menna Hamed, Wesam Ahmed, Sohyla Said, Sara Basheer +1 more
published similarity 0.66
Abstract · Accepted to the 14th International Japan-Africa Conference on Electronics, Communications and Computations (JAC-ECC 2026).6 pages, 3 figures…

Deployment of retrieval-augmented generation (RAG) for Edge-AI is constrained by memory and bandwidth limitations. A standard 7B parameter model consumes 14 GB Video Random Access Memory (VRAM) in half-precision (FP16), precluding co-location of vector indices and Key-Value (KV) caches. This paper presents an engineering integration and empirical evaluation of a multimodal RAG framework optimized for consumer-grade Turing hardware (e.g., NVIDIA Tesla T4). Our system-level analysis identifies a critical trade-off: while generative pre-trained transformer quantization (GPTQ) offers superior throughput (51.6 tok/s), it exhibits significant context instability at 2048+ tokens on Turing architectures. Conversely, NormalFloat 4-bit (NF4) maintains robust long-context performance with viable latency (81 ms/tok). By coupling NF4 with a decoupled Florence-2 vision integration, we achieve visual reasoning with only 430 MB VRAM overhead. Furthermore, empirical observations of prompt-induced search space pruning show that structured prompting reduces latency by 47% without altering computational intensity (1.88 GFLOPs/token). The integrated architecture maintains an 8.37 GB footprint, establishing a viable RAG pipeline for 12-16 GB GPUs. Finally, structured few-shot prompting effectively compensates for quantization-induced reasoning loss, restoring mathematical accuracy from 35.0% to 92.0% across 200 trials.

0.30

FAMPWQ: Fisher Information-based Adaptive Mixed Precision Weight Quantization for Effective LLM Inference

cs.LG cs.AIcs.DC ·2026-08-24 ·Gongwei Lee, Ji Liu, Juncheng Jia, Ji Wu
top venue similarity 0.66
Abstract · 21 pages, to appear in EMNLP 2026

Recent years have witnessed remarkable achievements of Large Language Models (LLMs) in multiple domains, while the excessive resource requirements of LLMs hinder the deployment on resource-constrained devices. Although model quantization stands out as an effective approach, conventional quantization approaches typically incur severe performance degradation due to uniform bit-width or simple heuristic sensitivity evaluation. In this paper, we propose a novel Fisher information-based Adaptive Mixed Precision Weight Quantization approach, i.e., FAMPWQ, which performs layer-adaptive weight quantization for effective LLM inference on commodity GPUs. First, we propose a system model with a novel Fisher information metric to measure the layer-wise sensitivity to quantization. Second, we propose a reinforcement learning-based bit-width allocator in FAMPWQ, which generates an adaptive bit-width allocation strategy based on the Fisher information sensitivity metric. Extensive experiments on 7 models and 5 benchmarks demonstrate that FAMPWQ significantly outperforms 7 baseline approaches in terms of PPL (up to 3.39 smaller), accuracy (up to 6.87% higher), and LLM-as-a-judge comparison (up to 76% win rate).

0.41