Xu Zhao

Xu Zhao

Auto-generated profile, based on publicly available information

Xu Zhao

Meta, Menlo Park, USA

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 April 2024
Verified
Conference Paper
Open access
JAJason Ansel
Jason Ansel
EYEdward Yang
Edward Yang
HHHorace He
Horace He
NGNatalia Gimelshein
Natalia Gimelshein
AJAnimesh Jain
Animesh Jain
SCSoumith Chintala
Soumith Chintala
This paper introduces two extensions to the popular PyTorch machine learning framework, TorchDynamo and TorchInductor, which implement the torch.compile feature released in PyTorch 2. TorchDynamo is a Python-level just-in-time (JIT) compiler that enables graph compilation in PyTorch programs without sacrificing the flexibility of Python. It achieves this by dynamically modifying Python bytecode before execution and extracting sequences of PyTorch operations into an FX graph, which is then JIT compiled using one of many extensible backends. TorchInductor is the default compiler backend for TorchDynamo, which translates PyTorch programs into OpenAI's Triton for GPUs and C++ for CPUs. Results show that TorchDynamo is able to capture graphs more robustly than prior approaches while adding minimal overhead, and TorchInductor is able to provide a 2.27× inference and 1.41× training geometric mean speedup on an NVIDIA A100 GPU across 180+ real-world models, which outperforms six other compilers. These extensions provide a new way to apply optimizations through compilers in eager mode frameworks like PyTorch.

30 November 2022 in ACM Transactions on Storage
Verified
Journal Article
Open access
RLRuibin Li
Ruibin Li
XRXiang Ren
Xiang Ren
XZXu Zhao
Xu Zhao
SHSiwei He
Siwei He
MSMichael Stumm
Michael Stumm
DYDing Yuan
Ding Yuan
Persistent byte-addressable memory (PM) is poised to become prevalent in future computer systems. PMs are significantly faster than disk storage, and accesses to PMs are governed by the Memory Management Unit (MMU) just as accesses with volatile RAM. These unique characteristics shift the bottleneck from I/O to operations such as block address lookup—for example, in write workloads, up to 45% of the overhead in ext4-DAX is due to building and searching extent trees to translate file offsets to addresses on persistent memory. We propose a novel contiguous file system, ctFS, that eliminates most of the overhead associated with indexing structures such as extent trees in the file system. ctFS represents each file as a contiguous region of virtual memory, hence a lookup from the file offset to the address is simply an offset operation, which can be efficiently performed by the hardware MMU at a fraction of the cost of software-maintained indexes. Evaluating ctFS on real-world workloads such as LevelDB shows it outperforms ext4-DAX and SplitFS by 3.6× and 1.8×, respectively.

8 August 2026
Verified
Conference Paper
Open access
MYMenglu Yu
Menglu Yu
JXJiaqi Xu
Jiaqi Xu
YHYuzhen Huang
Yuzhen Huang
YLYanbo Liang
Yanbo Liang
JLJia Liu
Jia Liu
CLChi-Keung Luk
Chi-Keung Luk

7 November 2026
Verified
Conference Paper
Open access
YLYing Li
Ying Li
YHYueming Hao
Yueming Hao
DBDavid Berard
David Berard
XZXu Zhao
Xu Zhao
BHBarney Huang
Barney Huang
YSYifan Sun
Yifan Sun