Figure 2 - uploaded by Mohsan Ali
Content may be subject to copyright.
Download
View publication
Baseline performance (in cycles, per loop iteration) of the code sequence

Baseline performance (in cycles, per loop iteration) of the code sequence

Source publication
Cite
Download full-text

Contexts in source publication

Context 1
... the sample program results in Figure 2 ...
Context 2
... multiprocessor memory systems, lower levels of the memory hierarchy may not be able to be saturated by a single processor but should be able to be saturated by multiple processors working together. Modify the code in Figure 2.29, and run multiple copies at the same time. ...
Context 3
... the memory bandwidth is sometimes 2X this, it would be 5120MB/sec. From Figure 2.14, this is just barely within the bandwidth provided by DDR2-667 DIMMs, so just one memory channel would suffice. ...

Similar publications

FIGURE 1. Comparison of (a) convolution of unrelated feature maps using...
FIGURE 2. The architecture of the two Pooling Block versions: (a)...
FIGURE 3. Sample video frames from the UCF101 and HMDB51 datasets. (a)...
Optimizing Spatiotemporal Feature Learning in 3D Convolutional Neural Networks With Pooling Blocks
Article
Full-text available
  • May 2021
Image data contain spatial information only, thus making two-dimensional (2D) Convolutional Neural Networks (CNN) ideal for solving image classification problems. On the other hand, video data contain both spatial and temporal information that must be simultaneously analyzed to solve action recognition problems. 3D CNNs are successfully used for th...
View
Neuromorphic computing market. Inspired by [14].
A comparison between specialized deep-learning digital electronic...
Traditional NN (left) versus DNN (right).
Conceptual framework of the newly proposed photonic neural network. The...
+8 The developed architectures of complex-valued optical neural...
Neuromorphic Photonics Circuits: Contemporary Review
Article
Full-text available
  • Dec 2023
Neuromorphic photonics is a cutting-edge fusion of neuroscience-inspired computing and photonics technology to overcome the constraints of conventional computing architectures. Its significance lies in the potential to transform information processing by mimicking the parallelism and efficiency of the human brain. Using optics and photonics princip...
View
Fig. 4: Different ways to leak a secret through a Send operation. The...
SoK: Hardware Defenses Against Speculative Execution Attacks
Preprint
Full-text available
  • Jan 2023
Speculative execution attacks leverage the speculative and out-of-order execution features in modern computer processors to access secret data or execute code that should not be executed. Secret information can then be leaked through a covert channel. While software patches can be installed for mitigation on existing hardware, these solutions can i...
View
Fig. 4: Performance of the synthetic models (blue curve associated with...
Fig. 5: Top: Single TPU execution of a model with layers stored on the...
Fig. 6: Speedup of synthetic models using Segm Comp, segmented into 2,...
Fig. 8: InceptionV3 architecture block with four open paths and...
+5 Fig. 9: Example diagram of a segmentation refinement. Top: Description...
Balanced segmentation of CNNs for multi-TPU inference
Preprint
Full-text available
  • Jun 2023
In this paper, we propose different alternatives for CNN (Convolutional Neural Networks) segmentation, addressing inference processes on computing architectures composed by multiple Edge TPUs. Specifically, we compare the inference performance for a number of state-of-the-art CNN models taking as a reference inference times on one TPU and a compile...
View
Error rate for computing the Givens rotation with x, y ∼ N (0, 1)
Fast Compensated Algorithms for the Reciprocal Square Root, the Reciprocal Hypotenuse, and Givens Rotations
Preprint
Full-text available
  • Feb 2021
The reciprocal square root is an important computation for which many very sophisticated algorithms exist (see for example \cite{863046,863031} and the references therein). In this paper we develop a simple differential compensation (much like those developed in \cite{borges}) that can be used to improve the accuracy of a naive calculation. The app...
View