QH-01 · Online Research IntelligenceResearch Intelligence
for Tomorrow’s Technology
Discover breakthrough papers, follow leading researchers, explore top schools, and stay ahead with the latest in computing hardware and AI.
Research
Latest Papers
Award-winning research from ISCA, MICRO, HPCA, and top computer architecture venues.
Cerberus
Junhwan Kim, Jungrae Kim et al. · SKKU
Adaptive Insertion Policies for High Performance Caching
Moinuddin K. Qureshi · Georgia Tech
Precise Exceptions in Relaxed Architectures
Ben Simner et al. · Cambridge
H2-LLM
Cong Li, Yuan Xie et al. · HKUST / Peking Univ.
The XOR Cache
Zhewen Pan, Joshua San Miguel · UW-Madison
LLM.265
Ceyu (Entropy) Xu, Lisa Wu Wills et al. · Duke + CMU
Athena
Rahul Bera, Onur Mutlu et al. · SAFARI / ETH Zürich
Exploiting Intel AMX for LLM Inference
Hyungyo Kim, Nam Sung Kim et al. · —
Torrent
Yunhao Deng, Marian Verhelst et al. · KU Leuven (MICAS)
Logic Synthesis Dissertation
Alessandro Tempia Calvino · EPFL → Synopsys
Wedge-Parallel Triangle Counting for GPUs
Jeffrey Spaan, K.-H. Chen, A.L. Varbanescu · Univ. of Twente
Career Award — Secure Architectures
Srinivas Devadas · MIT
TACO (Transpiler-Architecture Co-Design)
Meng Wang, Prashant Nair, Ang Li et al. · UBC / PNNL / UCSD / UT Austin
Interrupt Side-Channel Attacks on macOS
Xin Zhang, Trevor E. Carlson et al. · Peking Univ. / NUS
Patterns Behind Chaos
Zhongkai Yu, Po-An Tsai et al. · UCSB / NVIDIA
Career — Large-Scale AI Systems
Tushar Krishna · Georgia Tech
Career — Sustainable Cloud Datacenters
Akshitha Sriraman · Carnegie Mellon
Dissertation — Sparse Accelerators
Olivia Hsu (+ Jovan Stojkovic, UIUC) · Stanford
Career — Community Leadership
Sarita Adve · UIUC
People
Leading Researchers
Pioneers shaping the future of computer architecture, hardware design, and AI accelerators.
Bill Dally
Chief Scientist
NVIDIA
Christos Kozyrakis
Bosack-Lerner Professor of Engineering
Stanford
Onur Mutlu
Professor, SAFARI Lab
ETH Zürich
Vivienne Sze
Professor
MIT
Song Han
Associate Professor
MIT / NVIDIA Research
Marian Verhelst
Professor
KU Leuven (MICAS)
Yakun Sophia Shao
Associate Professor
UC Berkeley
Priyanka Raina
Associate Professor
Stanford
Tri Dao
Assistant Professor & Chief Scientist
Princeton / Together AI
Ceyu (Entropy) Xu
Research Assistant Professor
HKUST
Zhiyao Xie
Assistant Professor
HKUST
Academia
Top Schools
Leading universities for electrical engineering, computer engineering, and computer science research.
#1MIT
Only school ranked #1 in all three simultaneously. Song Han, Vivienne Sze, Devadas, CSAIL.
#2Stanford
Raina, Kozyrakis, Olukotun. Dally-adjacent via NVIDIA. Tri Dao PhD alma mater.
#3UC Berkeley
Yakun Sophia Shao, RISE Lab, ADEPT Lab. Birthplace of RISC and BSD Unix.
Carnegie Mellon
#1 overall CS. Akshitha Sriraman (Young Architect Award 2026).
Georgia Tech
Overall engineering #4. Tushar Krishna (Wilkes Award 2026). Qureshi.
UIUC
Sarita Adve. i-acoma group (Torrellas). Deep architecture bench.
Cornell
Broad strength across EE and CS. Cornell Tech NYC adds applied ML-systems.
University of Michigan
Strong VLSI and computer architecture programs. Top public university.
UT Austin
Yale Patt legacy. Jovan Stojkovic now Asst. Professor. LOIS group.
Purdue
Fastest-rising program. New wide-bandgap semiconductor MS launching Fall 2026.
Products
Trending Tech
The latest GPUs, accelerators, EDA tools, and computing hardware driving next-generation research and development.
Hardware
NVIDIA DGX Spark
Desktop AI supercomputer. Grace Blackwell GB10, 128GB unified memory, runs 200B-param models locally. Already in Stanford NLP and MIT CSAIL labs.
Hardware
NVIDIA B200 GPU
Flagship Blackwell datacenter GPU. 208B transistors, 192GB HBM3e, 9,000 TFLOPS FP4. FlashAttention-4 benchmarked here at 71% utilization.
Hardware
AMD Xilinx Versal Adaptive SoCs
Adaptive compute with AI Engines. Standard for FPGA prototyping in architecture labs. Verhelst's DATE 2026 best paper used this platform.
EDA
Cadence Virtuoso
Industry-standard analog/mixed-signal IC design. Used in virtually every university IC lab worldwide. Academic access through Europractice/MUSE programs.
EDA
Synopsys Design Compiler + VCS
RTL synthesis and simulation. Zhiyao Xie's AI-for-EDA research builds on Synopsys flows. Alessandro Tempia Calvino (DATE 2026 PhD Award) now works at Synopsys.
EDA
Siemens Calibre
DRC/LVS/parasitic extraction. The final sign-off tool before any chip goes to fab. No alternative has comparable coverage for physical verification.
Simulation
ChampSim
Trace-driven microarchitecture simulator (C++17). Used by Athena (HPCA 2026), DPC4, and most cache/prefetching research at ISCA/MICRO/HPCA.
Simulation
gem5
Full-system architecture simulator. Standard for CPU/GPU/memory hierarchy research across MIT, Berkeley, Michigan, Georgia Tech.
AI Framework
NVIDIA TensorRT-LLM
Production LLM inference library integrating Song Han's AWQ and SmoothQuant. FlashAttention-4 benchmarks run against this as baseline.
AI Framework
vLLM
High-throughput LLM serving with PagedAttention from UC Berkeley. De facto standard for academic LLM serving research.
Conferences
Upcoming Events
Major conferences, symposiums, and tech summits shaping the industry.
NVIDIA GTC 2026
NVIDIA's flagship developer conference. Keynotes on Blackwell architecture, CUDA advances, and AI infrastructure. 600+ sessions across AI, HPC, and autonomous systems.
Graphics & AI
SIGGRAPH 2026
The world's largest conference on computer graphics and interactive techniques. Neural rendering, real-time ray tracing, and AI-generated content take center stage.
Academic
ISCA 2026
The premier academic conference for computer architecture. Cerberus, Patterns Behind Chaos, and 5 career awards presented this year.
Hardware Summit
Hot Chips 2026
Industry leaders reveal next-generation chip architectures. AMD, Intel, NVIDIA, Google, and startups present silicon details not shared elsewhere.
Academic
MICRO 2026
Top-tier microarchitecture research. Following LLM.265's Best Paper in 2025, expect strong AI-hardware co-design and LLM acceleration papers.
Hardware Summit
ISSCC 2027
The chip industry's Olympics. Marian Verhelst chairs the 2027 program. State-of-the-art in processors, memory, wireless, and power management.