Raw nanopore signal analysis is a common approach in genomics to provide fast and resource-efficient analysis without translating the signals to bases (i.e., without basecalling). However, existing solutions cannot interpret raw signals directly if a reference genome is unknown due to a lack of accurate mechanisms to handle increased noise in pairwise raw signal comparison. Our goal is to enable the direct analysis of raw signals without a reference genome. To this end, we propose Rawsamble, the first mechanism that can 1) identify regions of similarity between all raw signal pairs, known as all-vs-all overlapping, using a hash-based search mechanism and 2) use these to construct genomes from scratch, called de novo assembly. Our extensive evaluations across multiple genomes of varying sizes show that Rawsamble provides a significant speedup (on average by 16.36x and up to 41.59x) and reduces peak memory usage (on average by 11.73x and up to by 41.99x) compared to a conventional genome assembly pipeline using the state-of-the-art tools for basecalling (Dorado's fastest mode) and overlapping (minimap2) on a CPU. We find that 36.57% of overlapping pairs generated by Rawsamble are identical to those generated by minimap2. Using the overlaps from Rawsamble, we construct the first de novo assemblies directly from raw signals without basecalling. We show that we can construct contiguous assembly segments (unitigs) up to 2.7 million bases in length (half the genome length of E. coli). We identify previously unexplored directions that can be enabled by finding overlaps and constructing de novo assemblies. Rawsamble is available at https://github.com/CMU-SAFARI/RawHash. We also provide the scripts to fully reproduce our results on our GitHub page.
The recent growth of Artificial Intelligence (AI), particularly large language models, requires energy-demanding high-performance computing (HPC) data centers, which poses a significant burden on power system capacity. Scheduling data center computing jobs to manage power demand can alleviate network stress with minimal infrastructure investment and contribute to fast time-scale power system balancing. This study, for the first time, comprehensively analyzes the capability and cost of grid flexibility provision by GPU-heavy AI-focused HPC data centers, along with a comparison with CPU-heavy general-purpose HPC data centers traditionally used for scientific computing. A data center flexibility cost model is proposed that accounts for the value of computing. Using real-world computing traces from 7 AI-focused HPC data centers and 7 general-purpose HPC data centers, along with computing prices from 3 cloud platforms, we find that AI-focused HPC data centers can offer greater flexibility at 50% lower cost compared to general-purpose HPC data centers for a range of power system services. By comparing the cost to flexibility market prices, we illustrate the financial profitability of flexibility provision for AI-focused HPC data centers. Finally, our flexibility and cost estimates can be scaled using parameters of other data centers through algebraic operations, avoiding the need for re-optimization.
Sparse Mixture of Experts (MoE) models, while outperforming dense Large Language Models (LLMs) in terms of performance, face significant deployment challenges during inference due to their high memory demands. Existing offloading techniques, which involve swapping activated and idle experts between the GPU and CPU, often suffer from rigid expert caching mechanisms. These mechanisms fail to adapt to dynamic routing, leading to inefficient cache utilization, or incur prohibitive costs for prediction training. To tackle these inference-specific challenges, we introduce ExpertFlow, a comprehensive system specifically designed to enhance inference efficiency by accommodating flexible routing and enabling efficient expert scheduling between CPU and GPU. This reduces overhead and boosts system performance. Central to our approach is a predictive routing path-based offloading mechanism that utilizes a lightweight predictor to accurately forecast routing paths before computation begins. This proactive strategy allows for real-time error correction in expert caching, significantly increasing cache hit ratios and reducing the frequency of expert transfers, thereby minimizing I/O overhead. Additionally, we implement a dynamic token scheduling strategy that optimizes MoE inference by rearranging input tokens across different batches. This method not only reduces the number of activated experts per batch but also improves computational efficiency. Our extensive experiments demonstrate that ExpertFlow achieves up to 93.72\% GPU memory savings and enhances inference speed by 2 to 10 times compared to baseline methods, highlighting its effectiveness and utility as a robust solution for resource-constrained inference scenarios.
FlashAttention series has been widely applied in the inference of large language models (LLMs). However, FlashAttention series only supports the high-level GPU architectures, e.g., Ampere and Hopper. At present, FlashAttention series is not easily transferrable to NPUs and low-resource GPUs. Moreover, FlashAttention series is inefficient for multi- NPUs or GPUs inference scenarios. In this work, we propose FastAttention which pioneers the adaptation of FlashAttention series for NPUs and low-resource GPUs to boost LLM inference efficiency. Specifically, we take Ascend NPUs and Volta-based GPUs as representatives for designing our FastAttention. We migrate FlashAttention series to Ascend NPUs by proposing a novel two-level tiling strategy for runtime speedup, tiling-mask strategy for memory saving and the tiling-AllReduce strategy for reducing communication overhead, respectively. Besides, we adapt FlashAttention for Volta-based GPUs by redesigning the operands layout in shared memory and introducing a simple yet effective CPU-GPU cooperative strategy for efficient memory utilization. On Ascend NPUs, our FastAttention can achieve a 10.7$\times$ speedup compared to the standard attention implementation. Llama-7B within FastAttention reaches up to 5.16$\times$ higher throughput than within the standard attention. On Volta architecture GPUs, FastAttention yields 1.43$\times$ speedup compared to its equivalents in \texttt{xformers}. Pangu-38B within FastAttention brings 1.46$\times$ end-to-end speedup using FasterTransformer. Coupled with the propose CPU-GPU cooperative strategy, FastAttention supports a maximal input length of 256K on 8 V100 GPUs. All the codes will be made available soon.
For over half a century, computer-aided structural elucidation systems (CASE) for organic compounds have relied on complex expert systems with explicitly programmed algorithms. These systems are often computationally inefficient for complex compounds due to the vast chemical structural space that must be explored and filtered. In this study, we present a proof-of-concept transformer based generative chemical language artificial intelligence (AI) model, an innovative end-to-end architecture designed to replace the logic and workflow of the classic CASE framework for ultra-fast and accurate spectroscopic-based structural elucidation. Our model employs an encoder-decoder architecture and self-attention mechanisms, similar to those in large language models, to directly generate the most probable chemical structures that match the input spectroscopic data. Trained on ~ 102k IR, UV, and 1H NMR spectra, it performs structural elucidation of molecules with up to 29 atoms in just a few seconds on a modern CPU, achieving a top-15 accuracy of 83%. This approach demonstrates the potential of transformer based generative AI to accelerate traditional scientific problem-solving processes. The model's ability to iterate quickly based on new data highlights its potential for rapid advancements in structural elucidation.
Proprietary large language models (LLMs) demonstrate exceptional generalization ability across various tasks. Additionally, deploying LLMs on edge devices is trending for efficiency and privacy reasons. However, edge deployment of proprietary LLMs introduces new security threats: attackers who obtain an edge-deployed LLM can easily use it as a base model for various tasks due to its high generalization ability, which we call foundational capability stealing. Unfortunately, existing model protection mechanisms are often task-specific and fail to protect general-purpose LLMs, as they mainly focus on protecting task-related parameters using trusted execution environments (TEEs). Although some recent TEE-based methods are able to protect the overall model parameters in a computation-efficient way, they still suffer from prohibitive communication costs between TEE and CPU/GPU, making it impractical to deploy for edge LLMs. To protect the foundational capabilities of edge LLMs, we propose CoreGuard, a computation- and communication-efficient model protection approach against model stealing on edge devices. The core component of CoreGuard is a lightweight and propagative authorization module residing in TEE. Extensive experiments show that CoreGuard achieves the same security protection as the black-box security guarantees with negligible overhead.
Nearly 900 million people live in low-lying coastal zones around the world and bear the brunt of impacts from more frequent and severe hurricanes and storm surges. Oceanographers simulate ocean current circulation along the coasts to develop early warning systems that save lives and prevent loss and damage to property from coastal hazards. Traditionally, such simulations are conducted using coastal ocean circulation models such as the Regional Ocean Modeling System (ROMS), which usually runs on an HPC cluster with multiple CPU cores. However, the process is time-consuming and energy expensive. While coarse-grained ROMS simulations offer faster alternatives, they sacrifice detail and accuracy, particularly in complex coastal environments. Recent advances in deep learning and GPU architecture have enabled the development of faster AI (neural network) surrogates. This paper introduces an AI surrogate based on a 4D Swin Transformer to simulate coastal tidal wave propagation in an estuary for both hindcast and forecast (up to 12 days). Our approach not only accelerates simulations but also incorporates a physics-based constraint to detect and correct inaccurate results, ensuring reliability while minimizing manual intervention. We develop a fully GPU-accelerated workflow, optimizing the model training and inference pipeline on NVIDIA DGX-2 A100 GPUs. Our experiments demonstrate that our AI surrogate reduces the time cost of 12-day forecasting of traditional ROMS simulations from 9,908 seconds (on 512 CPU cores) to 22 seconds (on one A100 GPU), achieving over 450$\times$ speedup while maintaining high-quality simulation results. This work contributes to oceanographic modeling by offering a fast, accurate, and physically consistent alternative to traditional simulation models, particularly for real-time forecasting in rapid disaster response.
Large language models (LLMs) with long context windows have gained significant attention. However, the KV cache, stored to avoid re-computation, becomes a bottleneck. Various dynamic sparse or TopK-based attention approximation methods have been proposed to leverage the common insight that attention is sparse. In this paper, we first show that TopK attention itself suffers from quality degradation in certain downstream tasks because attention is not always as sparse as expected. Rather than selecting the keys and values with the highest attention scores, sampling with theoretical guarantees can provide a better estimation for attention output. To make the sampling-based approximation practical in LLM generation, we propose MagicPIG, a heterogeneous system based on Locality Sensitive Hashing (LSH). MagicPIG significantly reduces the workload of attention computation while preserving high accuracy for diverse tasks. MagicPIG stores the LSH hash tables and runs the attention computation on the CPU, which allows it to serve longer contexts and larger batch sizes with high approximation accuracy. MagicPIG can improve decoding throughput by up to $5\times$ across various GPU hardware and achieve 54ms decoding latency on a single RTX 4090 for Llama-3.1-8B-Instruct model with a context of 96k tokens. The code is available at https://github.com/Infini-AI-Lab/MagicPIG.
Large Language Models (LLMs) have emerged as powerful tools in mathematical theorem proving, particularly when utilizing formal languages such as LEAN. The major learning paradigm is expert iteration, which necessitates a pre-defined dataset comprising numerous mathematical problems. In this process, LLMs attempt to prove problems within the dataset and iteratively refine their capabilities through self-training on the proofs they discover. We propose to use large scale LEAN problem datasets Lean-workbook for expert iteration with more than 20,000 CPU days. During expert iteration, we found log-linear trends between solved problem amount with proof length and CPU usage. We train a critic model to select relatively easy problems for policy models to make trials and guide the model to search for deeper proofs. InternLM2.5-StepProver achieves open-source state-of-the-art on MiniF2F, Lean-Workbook-Plus, ProofNet, and Putnam benchmarks. Specifically, it achieves a pass of 65.9% on the MiniF2F-test and proves (or disproves) 17.0% of problems in Lean-Workbook-Plus which shows a significant improvement compared to only 9.5% of problems proved when Lean-Workbook-Plus was released. We open-source our models and searched proofs at https://github.com/InternLM/InternLM-Math and https://huggingface.co/datasets/internlm/Lean-Workbook.
Current and future large scale structure surveys aim to constrain the neutrino mass and the equation of state of dark energy. We aim to construct accurate and interpretable symbolic approximations to the linear and nonlinear matter power spectra as a function of cosmological parameters in extended $\Lambda$CDM models which contain massive neutrinos and non-constant equations of state for dark energy. This constitutes an extension of the syren-halofit emulators to incorporate these two effects, which we call syren-new (SYmbolic-Regression-ENhanced power spectrum emulator with NEutrinos and $W_0-w_a$). We also obtain a simple approximation to the derived parameter $\sigma_8$ as a function of the cosmological parameters for these models. Our results for the linear power spectrum are designed to emulate CLASS, whereas for the nonlinear case we aim to match the results of EuclidEmulator2. We compare our results to existing emulators and $N$-body simulations. Our analytic emulators for $\sigma_8$, the linear and nonlinear power spectra achieve root mean squared errors of 0.1%, 0.3% and 1.3%, respectively, across a wide range of cosmological parameters, redshifts and wavenumbers. We verify that emulator-related discrepancies are subdominant compared to observational errors and other modelling uncertainties when computing shear power spectra for LSST-like surveys. Our expressions have similar accuracy to existing (numerical) emulators, but are at least an order of magnitude faster, both on a CPU and GPU. Our work greatly improves the accuracy, speed and range of applicability of current symbolic approximations to the linear and nonlinear matter power spectra. We provide publicly available code for all symbolic approximations found.
In this paper, we have gone through different AI-Based frameworks used for various educational tasks like digital customized assignment allotment and performance monitoring, identifying slow-learners and fast-learners, etc. application describes a novel invention, digital screen-integrated tables, designed specifically for educational settings. The tables feature integrated digital screens controlled by a central processing unit (CPU), enabling synchronized display of educational content such as textbooks, presentations, exam questions, and interactive learning materials. Additionally, the invention facilitates the collection of student performance data during classroom activities and assessments. The gathered data is utilized for analysis using machine learning models to identify patterns and trends in student learning behaviours. By leveraging machine learning algorithms, educators can ascertain whether a student is a fast learner or a slow learner, based on which, the teacher can allocate more resources to the slow learners. This innovative approach aims to address the evolving needs of modern classrooms by providing a dynamic and data-driven learning environment. The unique integration of digital screens into traditional classroom furniture represents a significant advancement in educational technology. This patent filing encompasses the design, functionality, and method of operation of the digital screen-integrated tables, emphasizing their innovative features and applications in educational institutions.
Quadratic Unconstrained Binary Optimization (QUBO) sits at the heart of many industries and academic fields such as logistics, supply chain, finance, pharmaceutical science, chemistry, IT, and energy sectors, among others. These problems typically involve optimizing a large number of binary variables, which makes finding exact solutions exponentially more difficult. Consequently, most QUBO problems are classified as NP-hard. To address this challenge, we developed a powerful feedforward neural network (FNN) optimizer for arbitrary QUBO problems. In this work, we demonstrate that the FNN optimizer can provide high-quality approximate solutions for large problems, including dense 80-variable weighted MaxCut and random QUBOs, achieving an average accuracy of over 99% in less than 1.1 seconds on an 8-core CPU. Additionally, the FNN optimizer outperformed the Gurobi optimizer by 72% on 200-variable random QUBO problems within a 100-second computation time limit, exhibiting strong potential for real-time optimization tasks. Building on this model, we explored the novel approach of integrating FNNs with a quantum annealer-based activation function to create a quantum-classical encoder-decoder (QCED) optimizer, aiming to further enhance the performance of FNNs in QUBO optimization.
Two-stage object detectors exhibit high accuracy and precise localization, especially for identifying small objects that are favorable for various edge applications. However, the high computation costs associated with two-stage detection methods cause more severe thermal issues on edge devices, incurring dynamic runtime frequency change and thus large inference latency variations. Furthermore, the dynamic number of proposals in different frames leads to various computations over time, resulting in further latency variations. The significant latency variations of detectors on edge devices can harm user experience and waste hardware resources. To avoid thermal throttling and provide stable inference speed, we propose Lotus, a novel framework that is tailored for two-stage detectors to dynamically scale CPU and GPU frequencies jointly in an online manner based on deep reinforcement learning (DRL). To demonstrate the effectiveness of Lotus, we implement it on NVIDIA Jetson Orin Nano and Mi 11 Lite mobile platforms. The results indicate that Lotus can consistently and significantly reduce latency variation, achieve faster inference, and maintain lower CPU and GPU temperatures under various settings.
Activation functions introduce non-linearity into Neural Networks, enabling them to learn complex patterns. Different activation functions vary in speed and accuracy, ranging from faster but less accurate options like ReLU to slower but more accurate functions like SiLU or SELU. Typically, same activation function is used throughout an entire model architecture. In this paper, we conduct a comprehensive study on the effects of using mixed activation functions in YOLO-based models, evaluating their impact on latency, memory usage, and accuracy across CPU, NPU, and GPU edge devices. We also propose a novel approach that leverages Neural Architecture Search (NAS) to design YOLO models with optimized mixed activation functions.The best model generated through this method demonstrates a slight improvement in mean Average Precision (mAP) compared to baseline model (SiLU), while it is 22.28% faster and consumes 64.15% less memory on the reference NPU device.
The rapid growth in machine learning models, especially in natural language processing and computer vision, has led to challenges when running these models on hardware with limited resources. This paper introduces Superpipeline, a new framework designed to optimize the execution of large AI models on constrained hardware during both training and inference. Our approach involves dynamically managing model execution by dividing models into individual layers and efficiently transferring these layers between GPU and CPU memory. Superpipeline reduces GPU memory usage by up to 60% in our experiments while maintaining model accuracy and acceptable processing speeds. This allows models that would otherwise exceed available GPU memory to run effectively. Unlike existing solutions that focus mainly on inference or specific model types, Superpipeline can be applied to large language models (LLMs), vision-language models (VLMs), and vision-based models. We tested Superpipeline's performance across various models and hardware setups. The method includes two key parameters that allow fine-tuning the balance between GPU memory use and processing speed. Importantly, Superpipeline does not require retraining or changing model parameters, ensuring that the original model's output remains unchanged. Superpipeline's simplicity and flexibility make it useful for researchers and professionals working with advanced AI models on limited hardware. It enables the use of larger models or bigger batch sizes on existing hardware, potentially speeding up innovation across many machine learning applications. This work marks an important step toward making advanced AI models more accessible and optimizing their deployment in resource-limited environments. The code for Superpipeline is available at https://github.com/abbasiReza/super-pipeline.
Inference with transformer-based language models begins with a prompt processing step. In this step, the model generates the first output token and stores the KV cache needed for future generation steps. This prompt processing step can be computationally expensive, taking 10s of seconds or more for billion-parameter models on edge devices when prompt lengths or batch sizes rise. This degrades user experience by introducing significant latency into the model's outputs. To reduce the time spent producing the first output (known as the ``time to first token'', or TTFT) of a pretrained model, we introduce a novel method called KV Prediction. In our method, a small auxiliary model is used to process the prompt and produce an approximation of the KV cache used by a base model. This approximated KV cache is then used with the base model for autoregressive generation without the need to query the auxiliary model again. We demonstrate that our method produces a pareto-optimal efficiency-accuracy trade-off when compared to baselines. On TriviaQA, we demonstrate relative accuracy improvements in the range of $15\%-50\%$ across a range of TTFT FLOPs budgets. We also demonstrate accuracy improvements of up to $30\%$ on HumanEval python code completion at fixed TTFT FLOPs budgets. Additionally, we benchmark models on an Apple M2 Pro CPU and demonstrate that our improvement in FLOPs translates to a TTFT speedup on hardware. We release our code at https://github.com/apple/corenet/tree/main/projects/kv-prediction .
Explainable molecular property prediction is essential for various scientific fields, such as drug discovery and material science. Despite delivering intrinsic explainability, linear models struggle with capturing complex, non-linear patterns. Large language models (LLMs), on the other hand, yield accurate predictions through powerful inference capabilities yet fail to provide chemically meaningful explanations for their predictions. This work proposes a novel framework, called MoleX, which leverages LLM knowledge to build a simple yet powerful linear model for accurate molecular property prediction with faithful explanations. The core of MoleX is to model complicated molecular structure-property relationships using a simple linear model, augmented by LLM knowledge and a crafted calibration strategy. Specifically, to extract the maximum amount of task-relevant knowledge from LLM embeddings, we employ information bottleneck-inspired fine-tuning and sparsity-inducing dimensionality reduction. These informative embeddings are then used to fit a linear model for explainable inference. Moreover, we introduce residual calibration to address prediction errors stemming from linear models' insufficient expressiveness of complex LLM embeddings, thus recovering the LLM's predictive power and boosting overall accuracy. Theoretically, we provide a mathematical foundation to justify MoleX's explainability. Extensive experiments demonstrate that MoleX outperforms existing methods in molecular property prediction, establishing a new milestone in predictive performance, explainability, and efficiency. In particular, MoleX enables CPU inference and accelerates large-scale dataset processing, achieving comparable performance 300x faster with 100,000 fewer parameters than LLMs. Additionally, the calibration improves model performance by up to 12.7% without compromising explainability.
This paper introduces a novel bionic intelligent optimisation algorithm, Octopus Inspired Optimization (OIO) algorithm, which is inspired by the neural structure of octopus, especially its hierarchical and decentralised interaction properties. By simulating the sensory, decision-making, and executive abilities of octopuses, the OIO algorithm adopts a multi-level hierarchical strategy, including tentacles, suckers, individuals and groups, to achieve an effective combination of global and local search. This hierarchical design not only enhances the flexibility and efficiency of the algorithm, but also significantly improves its search efficiency and adaptability. In performance evaluations, including comparisons with existing mainstream intelligent optimisation algorithms, OIO shows faster convergence and higher accuracy, especially when dealing with multimodal functions and high-dimensional optimisation problems. This advantage is even more pronounced as the required minimum accuracy is higher, with the OIO algorithm showing an average speedup of 2.27 times that of conventional particle swarm optimisation (PSO) and 9.63 times that of differential evolution (DE) on multimodal functions. In particular, when dealing with high-dimensional optimisation problems, OIO achieves an average speed of 10.39 times that of DE, demonstrating its superior computational efficiency. In addition, the OIO algorithm also shows a reduction of about 5\% in CPU usage efficiency compared to PSO, which is reflected in the efficiency of CPU resource usage also shows its efficiency. These features make the OIO algorithm show great potential in complex optimisation problems, and it is especially suitable for application scenarios that require fast, efficient and robust optimisation methods, such as robot path planning, supply chain management optimisation, and energy system management.
Dense Convolutional Network has been continuously refined to adopt a highly efficient and compact architecture, owing to its lightweight and efficient structure. However, the current Dense-like architectures are mainly designed manually, it becomes increasingly difficult to adjust the channels and reuse level based on past experience. As such, we propose an architecture search method called Dense Optimizer that can search high-performance dense-like network automatically. In Dense Optimizer, we view the dense network as a hierarchical information system, maximize the network's information entropy while constraining the distribution of the entropy across each stage via a power law, thereby constructing an optimization problem. We also propose a branch-and-bound optimization algorithm, tightly integrates power-law principle with search space scaling to solve the optimization problem efficiently. The superiority of Dense Optimizer has been validated on different computer vision benchmark datasets. Specifically, Dense Optimizer completes high-quality search but only costs 4 hours with one CPU. Our searched model DenseNet-OPT achieved a top 1 accuracy of 84.3% on CIFAR-100, which is 5.97% higher than the original one.
Resilience serves to assess the ability of the financial market to resist external shocks and to recover from shocks. The intensity and duration, used to indicate resilience, are calculated for China's financial market in this paper, focusing on the performance of each financial sub-market during and after several crises. Also, given that climate issues have been recognized as an important source of risk by financial market, we investigate the connectedness and mechanism of China's climate policy uncertainty (CPU) to its financial market resilience. We have found that the two resilience indicators of each market have a relatively consistent trend, but connectedness among markets has different sensitivities to both. In addition, China's CPU affects its financial market resilience by increasing the investor sentiment index and the non-performing loan ratio of commercial banks, and by reducing the capital and financial account balance. It is further found that China's financial markets' consensus on the unswerving implementation of climate policy, which provides the reference for other countries on how to balance climate policy's introduction and financial market development.