Results 261 to 270 of about 1,699,422 (313)
Some of the next articles are maybe not open access.
Implementation of Direct Memory Access for Parallel Processing
2020 4th International Conference on Electronics, Communication and Aerospace Technology (ICECA), 2020There has been an ever growing complexity in design and development of Embedded Systems for various applications. The applications like disaster management in a remote location requires embedded system to perform multiple tasks like data acquisition, processing and wireless transmission.
Chandni Arun +3 more
openaire +1 more source
IEEE Journal of Solid-State Circuits, 2023
Compute-in-memory (nvCIM) macros based on non-volatile memory make it possible for artificial intelligence (AI) edge devices to perform energy-efficient multiply-and-accumulate (MAC) operations by minimizing the movement of data between the processors ...
Je-Min Hung +13 more
semanticscholar +1 more source
Compute-in-memory (nvCIM) macros based on non-volatile memory make it possible for artificial intelligence (AI) edge devices to perform energy-efficient multiply-and-accumulate (MAC) operations by minimizing the movement of data between the processors ...
Je-Min Hung +13 more
semanticscholar +1 more source
Reporting Recollective Experiences: Direct Access to Memory Systems?
Psychological Science, 1995It has been suggested that the use of different memory systems can be identified by asking subjects whether they remember or know a recognized stimulus (e g, Gardiner & Java, 1990, Tulving, 1985) In this view, remember responses indicate the use of episodic memory, and know responses the use of implicit memory (e g, Gardiner & Java, in press ...
Fritz Strack, Jens Forster
openaire +1 more source
Fast and Efficient Model Serving Using Multi-GPUs with Direct-Host-Access
European Conference on Computer Systems, 2023As deep learning (DL) inference has been widely adopted for building user-facing applications in many domains, it is increasingly important for DL inference servers to achieve high throughput while preserving bounded latency. DL inference requests can be
Jinwoo Jeong +2 more
semanticscholar +1 more source
Remote Direct Memory Access (RDMA) over IP Problem Statement
Request for Comments, 2005A. Romanow +3 more
openaire +2 more sources
Optimizing Overlapped Memory Accesses in User-directed Vectorization
Proceedings of the 29th ACM on International Conference on Supercomputing, 2015Current processors incorporate wide and powerful vector units whose optimal exploitation is crucial to reach peak performance. However, present autovectorizing compilers fall short of that goal. Exploiting some vector instructions requires aggressive approaches that are not affordable in production compilers.
Diego Caballero +4 more
openaire +1 more source
Fast and Consistent Remote Direct Access to Non-volatile Memory
50th International Conference on Parallel Processing, 2021For performance benefits, recent trends of modern data centers tend to use NVM (Non-volatile Memory) as storage and utilize one-sided primitives of RDMA (Remote Direct Memory Access) to directly access NVM for I/O requests. However, one-sided RDMA doesn’t have durability semantics currently, and data may partially exist in the NVM when crashes happen ...
Jingwen Du +4 more
openaire +1 more source
2004
In the early 1990s there was no World Wide Web; in fact, many clients didn’t even have email access, and Oracle patches were sent almost exclusively by postal mail. For those clients who did have email, binary attachments were not yet an option. Patches had to be compressed and uuencoded, and even this was pushing the technological envelope for most of
openaire +1 more source
In the early 1990s there was no World Wide Web; in fact, many clients didn’t even have email access, and Oracle patches were sent almost exclusively by postal mail. For those clients who did have email, binary attachments were not yet an option. Patches had to be compressed and uuencoded, and even this was pushing the technological envelope for most of
openaire +1 more source
Direct memory access methods applied to chaotic behavior experiments
IEEE Transactions on Instrumentation and Measurement, 1997This paper deals with the implementation of a PC-controlled benchmark specially suited to real-time experimental investigation of oscillators driven by impulsive parametric pulses. The benchmark performs both the acquisition of the output signal and the generation of the parametric pulses.
H.A. Lorrondo +3 more
openaire +1 more source