|
Quamba: Post-training quantization for selective state space models
Background
State Space Models (SSMs) are integral to numerous advanced technologies, including natural language processing, robotics, autonomous vehicles, and edge computing systems. As these applications increasingly demand real-time processing and intelligent decision-making on devices with limited computational resources, there is a pressing...
|
|
Rapid optical metrology of critical dimensions of large-area nanostructure arrays with complex patterns
Background
Scatterometry-based metrology has the capability to perform high-throughput inspection of geometric characteristics of large-area nanopatterned surfaces. It utilizes physics-based dependencies between reflectance of light scattered from nanopatterned surfaces for pre-defined set of wavelengths and Critical Dimensions (CDs) of such nanopatterns.
Current...
|
|
An efficient codec for improved quality streaming video at low bitrates
Background
The ultimate goal of a successful video compression system is to reduce data volume while retaining the perceptual quality of the decompressed data. Standard video compression techniques, such as H.264, HEVC, VP9 and others, have well-known limits and tradeoffs between rate and distortion, especially at low bitrates. Better video compression...
|
|
Hard matrix multiplier blocks in field-programmable gate arrays
Rapidly changing algorithms and network architectures in the field of machine learning make Field Programmable Gate Array (FPGA) based designs an attractive solution for accelerating neural networks. However, the generic building blocks available in current FPGAs limit the acceleration that can be achieved. This technology, developed by researchers...
|
