arXiv:2605. 08692v2 Announce Type: replace Abstract: Post-training weight-only quantization to 4 bits is widely used to reduce the memory and compute costs of large language model inference.
arXiv:2608. 02700v1 Announce Type: cross Abstract: Analog compute-in-memory (CIM) enables energy-efficient neural network inference, but device variation and read noise can severely degrade low-bit quantized models.
The paper introduces TACO, a new optimizer for fine‑tuning large language models that drastically reduces optimizer state memory while preserving first‑order gradients. TACO selects the sign of the largest magnitude entry in each column of weight matrices, achieving a 174× reduction in persistent optimizer memory compared to AdamW8bit and a 2.9× decrease in peak training memory on OPT‑13B. This allows full‑parameter fine‑tuning of 30–32B‑parameter models on a single 80 GB GPU across multiple model families and tasks, with comparable accuracy and runtime to existing methods.
arXiv:2603. 13042v2 Announce Type: replace Abstract: Digital Compute-in-Memory (DCiM) accelerates neural networks by reducing data movement.
arXiv:2607. 28418v1 Announce Type: cross Abstract: Pruning is a promising approach for improving the efficiency of LLMs.
arXiv:2605. 24391v2 Announce Type: replace-cross Abstract: As the demand for deep learning grows, cost reduction through quantization has become essential for both training and inference.
arXiv:2505. 01043v2 Announce Type: replace Abstract: Large language models (LLMs) have achieved impressive performance across various domains.
Stable 4-bit floating-point (FP4) pretraining is difficult because the E2M1 payload represents only a narrow range of magnitudes. NVIDIA's Transformer Engine \nv{} recipe addresses this with current-tensor scaling, a randomized Hadamard transform (RHT), and bfloat16 (BF16) final layers, adding work outside the FP4 matrix multiplications.
HBQ: Hierarchical Scaling Block Quantization with Hardware‑Efficiency‑Aware Design for Accurate LLM Inference proposes a new block‑quantization scheme that uses large blocks and low‑overhead significand scaling to balance hardware efficiency and accuracy. The authors demonstrate that larger blocks improve efficiency by amortizing dequantization and accumulation costs, while their SIG scaling compensates for the resulting accuracy loss. Experiments on a 28 nm ASIC accelerator show that HBQ achieves up to 4.6× higher area/energy efficiency than state‑of‑the‑art weight‑only quantization, with 1.5–3.0× speedup and 1.6–3.3× system energy reduction over existing BQ methods.
arXiv:2607. 03148v1 Announce Type: cross Abstract: Activation functions are considered an essential primitive for neural nonlinearity, i.
arXiv:2609.36654v1 Announce Type: new Abstract: Large language models make weight storage and memory traffic major inference costs, motivating low-precision formats that represent each weight with on...
The paper introduces H-Scale, a lightweight post-processing technique for refining per-group scaling factors in NVFP4 quantized large language models. By using a diagonal second-order proxy from calibration activations, H-Scale selects hardware-valid scales that directly target layer output perturbation rather than just weight reconstruction error. Experiments on mainstream LLMs show that H-Scale improves NVFP4 baselines and brings several variants closer to BF16 performance without adding inference overhead.