微电子工程师 · 系统开发者 Microelectronics Engineer & Systems Developer 마이크로전자 엔지니어 · 시스템 개발자 微電子工程師 · 系統開發者
连结硅片与系统，
构建稳健的工程底座 Bridging Silicon and Systems,
Building robust foundations. 실리콘과 시스템을 연결하고,
견고한 기반을 구축하다. 連結晶片與系統，
構建穩健的工程底座
🔥 ICCAD 2026 EDA 求解器 🔥 ICCAD 2026 EDA Solver 🔥 ICCAD 2026 EDA 솔버 🔥 ICCAD 2026 EDA 求解器 ⚡ SM4 密码加速器 ⚡ SM4 Crypto Accelerator ⚡ SM4 암호 가속기 ⚡ SM4 密碼加速器 🐧 RISC-V Linux 移植 🐧 RISC-V Linux Porting 🐧 RISC-V Linux 포팅 🐧 RISC-V Linux 移植 🔬 本地 ASR 引擎 🔬 Local ASR Engine 🔬 로컬 ASR 엔진 🔬 本地 ASR 引擎 动态填充 200QPS Dynamic Padding 200QPS 동적 패딩 200QPS 動態填充 200QPS 动态分块 883MB Dynamic Chunking 883MB 동적 청킹 883MB 動態分塊 883MB RV32M 乘除法器 RV32M Multiply/Divide RV32M 곱셈/나눗셈 RV32M 乘除法器 Python + Rust 异构 Python + Rust Backend Python + Rust 이종 Python + Rust 異構 BTB 分支预测 BTB Branch Prediction BTB 분기 예측 BTB 分支預測 本地 ASR 引擎 Local ASR Engine 로컬 ASR 엔진 本地 ASR 引擎 冒险检测与数据转发 Hazard Detection & Forwarding 해저드 감지 & 포워딩 冒險檢測與數據轉發 LLM Agent EDA LLM Agent for EDA LLM 에이전트 EDA LLM Agent EDA 5级流水线 RISC-V CPU 5-Stage RISC-V CPU 5단계 RISC-V CPU 5級流水線 RISC-V CPU RISC-V Linux 移植 RISC-V Linux Porting RISC-V Linux 포팅 RISC-V Linux 移植 SM4 侧信道防御 SM4 Side-Channel Defence SM4 사이드 채널 방어 SM4 側通道防禦 SM4 流水线 94MHz SM4 Pipeline 94MHz SM4 파이프라인 94MHz SM4 流水線 94MHz 位并行模拟 15× Bit-Parallel 15× Speedup 비트병렬 15× 가속 位元並行模擬 15× AIG 重写引擎 AIG Local Rewriting AIG 로컬 리라이팅 AIG 重寫引擎 EDA 综合引擎 EDA Synthesis Engine EDA 합성 엔진 EDA 綜合引擎 搭建技术博客 Building a Tech Blog 기술 블로그 구축 搭建技術部落格 Linux 服务器加固 Linux Server Hardening Linux 서버 강화 Linux 伺服器加固 Docker 容器安全 Docker Container Security Docker 컨테이너 보안 Docker 容器安全
最新文章 Latest Posts 최신 포스트 最新文章
查看全部 View All 전체 보기 查看全部 asr performance concurrency
Dynamic Padding: 200 QPS with Zero Compute Waste for Short Audio
Why Whisper wastes 90% of compute on padding for 1-second commands, and how dynamic padding eliminates it.
asr memory systems
Dynamic Tensor Chunking: Constant 883MB Memory for Arbitrary-Length Audio
How slicing audio into 30-second tensor chunks prevents OOM on 24-hour recordings while keeping peak memory constant at 883MB.
risc-v mul-div arithmetic
RV32M Multiply/Divide Unit: 3-Cycle Mul, 32-Cycle Div
A multi-cycle RV32M unit implementing pipelined 33×33-bit multiplication (3 cycles) and restoring division with sign correction (32 cycles) — handling divide-by-zero and INT_MIN overflow edge cases.