Author

Baidu Tech Salon

Baidu Tech Salon, organized by Baidu's Technology Management Department, is a monthly offline event that shares cutting‑edge tech trends from Baidu and the industry, providing a free platform for mid‑to‑senior engineers to exchange ideas.

283

Articles

Likes

466

Views

Comments

Latest from Baidu Tech Salon

100 recent articles max

Baidu Tech Salon

Sep 27, 2024 · Artificial Intelligence

PaddleScience Presents Geometry-Informed Neural Operator Model for Aerodynamic Drag Prediction at 2024 CAE Aerodynamics Branch Academic Annual Meeting

At the 2024 CAE Aerodynamics Branch Academic Annual Meeting, PaddleScience unveiled a geometry‑informed neural operator model that predicts automotive drag with 2.1% average error, runs three orders of magnitude faster than full CFD, earned an Outstanding Paper Award, and showcases the toolkit’s AI‑driven workflow for rapid vehicle‑shape optimization.

AICFDDrag Prediction

0 likes · 7 min read

PaddleScience Presents Geometry-Informed Neural Operator Model for Aerodynamic Drag Prediction at 2024 CAE Aerodynamics Branch Academic Annual Meeting

Baidu Tech Salon

Sep 26, 2024 · Artificial Intelligence

Baidu Comate Leads AI Code Generation Market According to Industry Reports

Baidu's Wenxin KuaiMa (Comate) leads the AI code generation market, ranking first in Sullivan and SuperCLUE evaluations across technical, product and market dimensions, supporting over 200 languages, boosting developer productivity, adopted by thousands of enterprises, with China's market projected to reach 33 billion yuan by 2028.

AI code generationBaidu ComateWenxin KuaiMa

0 likes · 9 min read

Baidu Comate Leads AI Code Generation Market According to Industry Reports

Baidu Tech Salon

Sep 25, 2024 · Backend Development

Innovative Solutions for Reducing Result Inconsistency in Baidu Search System

The paper introduces a production‑grade framework that uses tiny controlled traffic, feature‑flattening experiments, dynamic debugging, and an automated inspection flywheel to measure each component’s contribution to Baidu’s search result diff‑rate, isolate root causes, and dramatically reduce inconsistency without impacting real users.

data flatteningdebuggingdiff rate

0 likes · 13 min read

Innovative Solutions for Reducing Result Inconsistency in Baidu Search System

Baidu Tech Salon

Aug 30, 2024 · Artificial Intelligence

AI4R&D: AI-Driven R&D New Paradigm

The AI4R&D report released on July 25 2024 outlines how artificial‑intelligence technologies are reshaping scientific research and industrial development by boosting computing power, data utilization and toolchains, delivering breakthroughs in drug discovery, materials and simulation, and mapping an emerging ecosystem that promises faster translation of innovation into practical applications.

AI ecosystemAI4R&DArtificial Intelligence

0 likes · 5 min read

Baidu Tech Salon

Aug 27, 2024 · Artificial Intelligence

How PaddleX Enables Early Detection of Malignant Skin Tumors with AI Segmentation

This article examines the urgent need for early skin cancer detection in China, outlines the challenges of dermatological imaging, and details a low‑code PaddleX solution that leverages PP‑LiteSeg‑T for data preparation, model training, optimization, and deployment to improve diagnostic accuracy and efficiency.

AIMedical ImagingPaddleX

0 likes · 10 min read

How PaddleX Enables Early Detection of Malignant Skin Tumors with AI Segmentation

Baidu Tech Salon

Aug 20, 2024 · Artificial Intelligence

PaddlePaddle Neural Network Compiler (CINN): Architecture, Optimization Techniques, and Performance

The PaddlePaddle Neural Network Compiler (CINN) combines a PIR‑based frontend and a hardware‑specific backend to apply graph‑level optimizations, operator fusion, schedule transformations and automatic tuning, delivering up to 4× faster kernels and 30‑60% overall speed‑ups for deep‑learning and scientific workloads.

CINNGPU OptimizationOperator fusion

0 likes · 19 min read

PaddlePaddle Neural Network Compiler (CINN): Architecture, Optimization Techniques, and Performance

Baidu Tech Salon

Aug 15, 2024 · Artificial Intelligence

Implementing PCA for Face Recognition with PaddlePaddle: A Step‑by‑Step Guide

This article walks through a complete PCA‑based face‑recognition pipeline using the PaddlePaddle framework, covering dataset preparation, library installation, image vectorization, PCA dimensionality reduction, training, testing, and performance evaluation with detailed code examples.

PCAPaddlePaddledimensionality reduction

0 likes · 12 min read

Implementing PCA for Face Recognition with PaddlePaddle: A Step‑by‑Step Guide

Baidu Tech Salon

Aug 8, 2024 · Operations

Can Smart Caching Slash SSD Read Latency for High‑Performance Ad Search?

This article examines the necessity of caching in ad‑search workloads, evaluates classic eviction policies such as LRU, LFU, Redis Approx‑LRU, OS PageCache, DoubleClock, ARC, 2Q, TinyLFU, and presents a layered TLS‑centered cache design (SsdEngine) with detailed benchmark results showing hit‑rate and latency improvements across Zipfian and uniform workloads.

CachingSSDStorage Systems

0 likes · 25 min read

Can Smart Caching Slash SSD Read Latency for High‑Performance Ad Search?

Baidu Tech Salon

Jul 23, 2024 · Artificial Intelligence

Linear Algebra Fundamentals and PaddlePaddle Applications

The article reviews core linear algebra concepts—vectors, matrices, eigenvalues, and transformations—and demonstrates how PaddlePaddle’s paddle.linalg API enables practical tasks such as least‑squares regression, image compression via SVD, PCA‑based dimensionality reduction, and broader machine‑learning, graphics, cryptography, and optimization applications.

PCAPaddlePaddleSVD

0 likes · 10 min read

Linear Algebra Fundamentals and PaddlePaddle Applications

Baidu Tech Salon

Jul 11, 2024 · Industry Insights

How Baidu Feed Evolved Its Data Warehouse with Multi‑Version Wide Tables

This article outlines the step‑by‑step evolution of Baidu's Feed data warehouse—from traditional layered modeling to hour‑level core tables, then real‑time wide tables, and finally a flow‑batch integrated multi‑version wide‑table architecture—highlighting the motivations, design choices, challenges, and resulting benefits.

Data ModelingData WarehouseReal-time analytics

0 likes · 10 min read