Sihao Liu

Sihao Liu

Auto-generated profile, based on publicly available information

Sihao Liu

University of California, Los Angeles

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

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

Representative Publications

View All Publications
1 May 2020
Verified
Conference Paper
Closed access
JWJian Weng
Jian Weng
SLSihao Liu
Sihao Liu
VDVidushi Dadu
Vidushi Dadu
ZWZhengrong Wang
Zhengrong Wang
PSPreyas Shah
Preyas Shah
TNTony Nowatzki
Tony Nowatzki
Domain-specific hardware accelerators can provide orders of magnitude speedup and energy efficiency over general purpose processors. However, they require extensive manual effort in hardware design and software stack development. Automated ASIC generation (eg. HLS) can be insufficient, because the hardware becomes inflexible. An ideal accelerator generation framework would be automatable, enable deep specialization to the domain, and maintain a uniform programming interface. Our insight is that many prior accelerator architectures can be approximated by composing a small number of hardware primitives, specifically those from spatial architectures. With careful design, a compiler can understand how to use available primitives, with modular and composable transformations, to take advantage of the features of a given program. This suggests a paradigm where accelerators can be generated by searching within such a rich accelerator design space, guided by the affinity of input programs for hardware primitives and their interactions. We use this approach to develop the DSAGEN framework, which automates the hardware/software co-design process for reconfigurable accelerators. For several existing accelerators, our evaluation demonstrates that the compiler can achieve 89% of the performance of manually tuned versions. For automated design space exploration, we target multiple sets of workloads which prior accelerators are design for; the generated hardware has mean 1.3× perf2/mm2 over prior programmable accelerators.

12 October 2019
Verified
Conference Paper
Open access
PDF file
VDVidushi Dadu
Vidushi Dadu
JWJian Weng
Jian Weng
SLSihao Liu
Sihao Liu
TNTony Nowatzki
Tony Nowatzki

1 February 2020
Verified
Conference Paper
Closed access
JWJian Weng
Jian Weng
SLSihao Liu
Sihao Liu
ZWZhengrong Wang
Zhengrong Wang
VDVidushi Dadu
Vidushi Dadu
TNTony Nowatzki
Tony Nowatzki
Dense linear algebra kernels are critical for wireless, and the oncoming proliferation of 5G only amplifies their importance. Due to the inductive nature of many such algorithms, parallelism is difficult to exploit: parallel regions have fine-grain producer/consumer interaction with iteratively changing depen-dence distance, reuse rate, and memory access patterns. This makes multi-threading impractical due to fine-grain synchronization, and vectorization ineffective due to the non-rectangular iteration domain. CPUs, DSPs, and GPUs perform order-of-magnitude below peak. Our insight is that if the nature of inductive dependences and memory accesses were explicit in the hardware/software interface, then a spatial architecture could efficiently execute parallel code regions. To this end, we first develop a novel execution model, inductive dataflow, where inductive dependence patterns and memory access patterns (streams) are first-order primitives. Second, we develop a hybrid spatial architecture combining systolic and tagged dataflow execution to attain high utilization at low energy and area cost. Finally, we create a scalable design through a novel vector-stream control model which amortizes control overhead both in time and spatially across architecture lanes. We evaluate our design, REVEL, with a full stack (compiler, ISA, simulator, RTL). Across a suite of linear algebra kernels, REVEL outperforms equally-provisioned DSPs by 4.6×-37×. Compared to state-of-the-art spatial architectures, REVEL is mean 3× faster. Compared to a set of ASICs, REVEL is only 2× the power and half the area.

25 April 2018 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
SZSiyuan Zhang
Siyuan Zhang
SSShaoqiang Shang
Shaoqiang Shang
YHYuqiang Han
Yuqiang Han
CGChunming Gu
Chunming Gu
SWShan Wu
Shan Wu
MWMingxi Wan
Mingxi Wan
The feasibility of ultrasonic Nakagami imaging to evaluate thermal lesions by high-intensity focused ultrasound (HIFU) and microwave ablation (MWA) was explored in ex vivo andin vivo liver models. Dynamic changes of the ultrasonic Nakagami parameter in thermal lesions were calculated, and ultrasonic B-mode and Nakagami images were reconstructed simultaneously. The contrast-to-noise ratio (CNR) between thermal lesions and normal tissue was used to estimate the contrast resolution of the monitoring images. After thermal ablation, a bright hyper-echoic region appeared in the ultrasonic B-mode and Nakagami images, identifying the thermal lesion. During thermal ablation, mean values of Nakagami parameter showed an increasing trend from 0.72 to 1.01 for theex vivo model and 0.54 to 0.72 for the in vivo model. After thermal ablation, mean CNR values of the ultrasonic Nakagami images were 1.29 dB (ex vivo) and 0.80 dB (in vivo), significantly higher (p<0.05) than those for B-mode images. Thermal lesion size, assessed using ultrasonic Nakagami images, shows a good correlation to those obtained from the gross-pathology images (for ex vivo model: length, r =0.96; width, r =0.90; for in vivo model: length, r =0.95; width, r =0.85). This preliminary study suggests that ultrasonic Nakagami parameter may have a potential use in evaluating the formation of thermal lesions with better image contrast. Moreover, ultrasonic Nakagami imaging combined with B-mode imaging may be utilized as an alternative modality in developing monitoring systems for image-guided thermal ablation treatments.

12 October 2019
Verified
Conference Paper
Open access
ASAmirali Sharifian
Amirali Sharifian
RHReza Hojabr
Reza Hojabr
NRNavid Rahimi
Navid Rahimi
SLSihao Liu
Sihao Liu
AGApala Guha
Apala Guha
ASArrvindh Shriraman
Arrvindh Shriraman

12 July 2022 in IEEE Micro
Verified
Journal Article
Closed access
JWJian Weng
Jian Weng
SLSihao Liu
Sihao Liu
DKDylan Kupsh
Dylan Kupsh
TNTony Nowatzki
Tony Nowatzki
Spatial accelerators provide high performance, energy efficiency, and flexibility. Recent design frameworks enable these architectures to be quickly designed and customized to a domain. However, constructing a compiler for this immense design space is challenging, first because accelerators express programs with high-level idioms that are difficult to recognize. Second, it is unpredictable whether certain transformations are beneficial or will lead to infeasible hardware mappings. Our work develops a general spatial-accelerator compiler with two key ideas. First, we propose an approach to recognize and represent useful dataflow idioms, along with a novel idiomatic memory representation. Second, we propose the principle of modular compilation, which combines hardware-aware transformation selection and an iterative approach to handle uncertainty. Our compiler achieves 2.3× speedup, and 98.7× area-normalized speedup over high-end server CPU.