Cache telepathy: Leveraging shared resource attacks to learn {DNN} architectures

M Yan, CW Fletcher, J Torrellas - 29th USENIX Security Symposium …, 2020 - usenix.org
Deep Neural Networks (DNNs) are fast becoming ubiquitous for their ability to attain good
accuracy in various machine learning tasks. A DNN's architecture (ie, its hyperparameters) …

Practical near-data processing for in-memory analytics frameworks

M Gao, G Ayers, C Kozyrakis - 2015 International Conference …, 2015 - ieeexplore.ieee.org
The end of Dennard scaling has made all systemsenergy-constrained. For data-intensive
applications with limitedtemporal locality, the major energy bottleneck is data …

A deeper look into rowhammer's sensitivities: Experimental analysis of real dram chips and implications on future attacks and defenses

L Orosa, AG Yaglikci, H Luo, A Olgun, J Park… - MICRO-54: 54th Annual …, 2021 - dl.acm.org
RowHammer is a circuit-level DRAM vulnerability where repeatedly accessing (ie,
hammering) a DRAM row can cause bit flips in physically nearby rows. The RowHammer …

[LLIBRE][B] Memory systems: cache, DRAM, disk

B Jacob, D Wang, S Ng - 2010 - books.google.com
Is your memory hierarchy stop** your microprocessor from performing at the high level it
should be? Memory Systems: Cache, DRAM, Disk shows you how to resolve this problem …

Memory access scheduling

S Rixner, WJ Dally, UJ Kapasi, P Mattson… - ACM SIGARCH …, 2000 - dl.acm.org
The bandwidth and latency of a memory system are strongly dependent on the manner in
which accesses interact with the “3-D” structure of banks, rows, and columns characteristic of …

[LLIBRE][B] Computer systems: a programmer's perspective

RE Bryant, DR O'Hallaron - 2011 - thuvienso.hoasen.edu.vn
Computer Systems: A Programmer's Perspective Page 1 Computer Systems A Programmer’s
Perspective Randal E. Bryant Carnegie Mellon University David R. O’Hallaron Carnegie Mellon …

Scheduling techniques for GPU architectures with processing-in-memory capabilities

A Pattnaik, X Tang, A Jog, O Kayiran… - Proceedings of the …, 2016 - dl.acm.org
Processing data in or near memory (PIM), as opposed to in conventional computational units
in a processor, can greatly alleviate the performance and energy penalties of data transfers …

[LLIBRE][B] Reconfigurable computing: the theory and practice of FPGA-based computation

S Hauck, A DeHon - 2010 - books.google.com
Reconfigurable Computing marks a revolutionary and hot topic that bridges the gap
between the separate worlds of hardware and software design—the key feature of …

Self-optimizing memory controllers: A reinforcement learning approach

E Ipek, O Mutlu, JF Martínez, R Caruana - ACM SIGARCH Computer …, 2008 - dl.acm.org
Efficiently utilizing off-chip DRAM bandwidth is a critical issuein designing cost-effective,
high-performance chip multiprocessors (CMPs). Conventional memory controllers deliver …

OWL: cooperative thread array aware scheduling techniques for improving GPGPU performance

A Jog, O Kayiran, N Chidambaram Nachiappan… - ACM SIGPLAN …, 2013 - dl.acm.org
Emerging GPGPU architectures, along with programming models like CUDA and OpenCL,
offer a cost-effective platform for many applications by providing high thread level …