New York, NY, United States

Reservoir Labs, Inc.

www.reservior.com
New York, NY, United States
SEARCH FILTERS
Time filter
Source Type

A signal pre-compensation system analyzes one or more properties of a communication medium and, taking advantage of the locality of propagation, generates using sparse fast Fourier transform (sFFT) a sparse kernel based on the medium properties. The system models propagation of data signals through the medium as a fixed-point iteration based on the sparse kernel, and determines initial amplitudes for the data symbol(s) to be transmitted using different communication medium modes. Fixed-point iterations are performed using the sparse kernel to iteratively update the initial amplitudes. If the iterations converge, a subset of the finally updated amplitudes is used as launch amplitudes for the data symbol(s). The data symbol(s) can be modulated using these launch amplitudes such that upon propagation of the pre-compensated data symbol(s) through the communication medium, they would resemble the original data symbols at a receiver, despite any distortion and/or cross-mode interference in the communication medium.


Patent
Reservoir Labs, Inc. | Date: 2016-10-05

A system for compiling programs for execution thereof using a hierarchical processing system having two or more levels of memory hierarchy can perform memory-level-specific optimizations, without exceeding a specified maximum compilation time. To this end, the compiler system employs a polyhedral model and limits the dimensions of a polyhedral program representation that is processed by the compiler at each level using a focalization operator that temporarily reduces one or more dimensions of the polyhedral representation. Semantic correctness is provided via a defocalization operator that can restore all polyhedral dimensions that had been temporarily removed.


We present the architecture of a high-performance constraint solver R-Solve that extends the gains made in SAT performance over the past fifteen years on static decision problems to problems that require on-the-fly adaptation, solution space exploration and optimization. R-Solve facilitates collaborative parallel solving and provides an efficient system for unrestricted incremental solving via Smart Repair. R-Solve can address problems in dynamic planning and constrained optimization involving complex logical and arithmetic constraints.


Patent
Reservoir Labs, Inc. | Date: 2015-10-29

In a sequence of major computational steps or in an iterative computation, a stencil amplifier can increase the number of data elements accessed from one or more data structures in a single major step or iteration, thereby decreasing the total number of computations and/or communication operations in the overall sequence or the iterative computation. Stencil amplification, which can be optimized according to a specified parameter such as compile time, rune time, code size, etc., can improve the performance of a computing system executing the sequence or the iterative computation in terms of run time, memory load, energy consumption, etc. The stencil amplifier typically determines boundaries, to avoid erroneously accessing data elements not present in the one or more data structures.


Grant
Agency: Department of Energy | Branch: | Program: SBIR | Phase: Phase II | Award Amount: 1000.00K | Year: 2015

According to a 2013 report published by FireEye, on average it took 229 days between the breach of a computer system and the detection of said breach. In order to effectively mitigate these breaches, traditional corporate border security needs to transition to a defense in depth cyber security posture. As networks get faster, such as the DOE ESNet and corporate datacenters such as Target and The Home Depot, it becomes more difficult to gain network visibility across the whole enterprise. The technology developed in this proposal will enable such network visibility, even on the largest and fastest networks available today.We have developed a scalable cybersecurity technology called R-Scope that at its core can inspect every packet transiting a network to determine where that packet came from, where it is going, and the type of communication it is carrying. This data can then be used to craft policy which can alert when anomalous events take place, such as the exfiltration of data (credit cards, personal identification, critical intellectual property, or even the latest movies) off premises to foreign computer systems. Through the Phase I and Phase II efforts, we engaged in basic R&D to address the notion of "performance collapse" in a network cyber security appliance. This R&D was then rolled into a commercial appliance offering using the Bra open source engine to provide cyber security aware network visibility. Our work addressed the core feature set necessary for a product and began the marketing and sales process to bring the appliance to market culminating with a few early sales of the product. The follow-on funding will be used to complete commercialization, enabling the product leapfrogging across the "market chasm" where many highly technical products die because of their inability to address the larger market beyond the early adopter. We will address product areas that include customer usability, enterprise corporate fit and competitiveness that overall lead us to enabling the product to fit well in a corporate environment and allows us to scale our support of the product as we grow. Commercial Applications and Other Benefits This work will enhance collaborative science, making more efficient and secure use of existing and future networking technologies that will help usher in scientific breakthroughs. At the same time, it will help protect corporate and government datacenters from malicious activity that costs the economy millions of dollars.


Patent
Reservoir Labs, Inc. | Date: 2016-01-04

A system for allocation of one or more data structures used in a program across a number of processing units takes into account a memory access pattern of the data structure, and the amount of total memory available for duplication across the several processing units. Using these parameters duplication factors are determined for the one or more data structures such that the cost of remote communication is minimized when the data structures are duplicated according to the respective duplication factors while allowing parallel execution of the program.


Patent
Reservoir Labs, Inc. | Date: 2016-01-04

A compilation system using an energy model based on a set of generic and practical hardware and software parameters is presented. The model can represent the major trends in energy consumption spanning potential hardware configurations using only parameters available at compilation time. Experimental verification indicates that the model is nimble yet sufficiently precise, allowing efficient selection of one or more parameters of a target computing system so as to minimize power/energy consumption of a program while achieving other performance related goals. A voltage and/or frequency optimization and selection is presented which can determine an efficient dynamic hardware configuration schedule at compilation time. In various embodiments, the configuration schedule is chosen based on its predicted effect on energy consumption. A concurrency throttling technique based on the energy model can exploit the power-gating features exposed by the target computing system to increase the energy efficiency of programs.


A compilation system can compile a program to be executed using an event driven tasks (EDT) system that requires knowledge of dependencies between program statement instances, and generate the required dependencies efficiently when a tiling transformation is applied. To this end, the system may use pre-tiling dependencies and can derive post-tiling dependencies via an analysis of the tiling to be applied.


Grant
Agency: Department of Energy | Branch: | Program: STTR | Phase: Phase II | Award Amount: 999.79K | Year: 2014

The Polyhedral User Mapping Assistant and Visualizer (PUMA-V) project is addressing the challenges of developing optimized computer software for heterogeneous supercomputing systems in support of physics simulations needed for bringing theoretical calculations to the level of the newest precision experiments. Current Lattice Quantum Chromodynamics (LQCD) and Lattice Beyond Standard Model (LBSM) simulations require four orders of magnitude greater computational rates to reach this level. Heterogeneous hardware can provide orders of magnitude improvements in computational rates but presents significantly more complex code development and porting challenges. The project will develop the capability for optimizing LQCD software bases in particular the heavily templated QDP++ data parallel libraries through an interactive control and visualization interface for advanced software optimization technologies. The approach will leverage the polyhedral model for precision of dependence analysis and increased range of optimized mappings. Phase I research and development (1) demonstrated capabilities for recovering high-level array semantics from the templated software base by overloading templates in the QDP++ library (2) developed a visualization and control interface within the Eclipse Interactive Development Environment (IDE) for polyhedral parallelization and optimization. (1) Implementation of translation tools interfacing heavily templated LQCD libraries to polyhedral mapping (2) application of the tools to current theoretical physics simulations related to testing Standard Model (SM) theories (3) development of new optimized solver algorithms for LQCD (4) development and quality improvements. Commercial Applications and Other Benefits: Acceleration of theoretical science in support of Standard Model and Beyond Standard Model theories. Energy and cost savings in high performance computing systems generally. Reduced cost and defects for high performance computing software development through advanced development tools and productivity methodologies. Increased rate of scientific research and innovation.


Grant
Agency: Department of Energy | Branch: | Program: SBIR | Phase: Phase II | Award Amount: 999.99K | Year: 2014

As supercomputing speeds increase and storage costs drop, scientists running applications in various disci- plines, such as high-energy physics, genomics, climate studies, etc., generate datasets of ever-increasing sizes at distributed locations around the world. Backbone Research-and-Education Network (REN) providers such as DOE’s Energy Sciences Network (ESnet) and Internet2, provide high-speed, reliable network services to support these scientists located at various US Department of Energy (DOE)’s national laboratories and universities. Our work will identify, in real-time, very large data flows across these networks, moving them to alternative network paths or queues so that interactive applications such as voice, video, and applications can retain a high quality of service thereby extending the interval to the next inevitable costly equipment upgrade cycle. We are developing a system called real-time Hybrid Network Traffic Engineering System (rHNTES) which will tap the traffic between networking installations to determine which network flows are large, ongoing, and of high bandwidth; i.e. elephant flows. We consider the timely completion of these elephant flows less important than interactive network traffic such as voice, video or applications and as such, change the network priority such that elephant flows are lower priority than other more import network traffic. Our elephant flow identification is accomplished in real-time, using off-the-shelf network processors and innovative software algorithms. Once identified, we will reprogram the switching infrastructure to ensure these flows do not impact the higher priority flows. This project will develop a software packet processing engine using readily available commodity network processors which will be able to analyze network packets at up to 100 billion bits per second. We will use this software engine to develop our real-time HNTES (rHNTES) system based on prior HNTES algorithm work. The network routers will be reprogrammed on-the-fly to mitigate the impact of each of the elephant flows we identify. This system will then be integrated into existing network performance monitoring tools to help identify where network bottles occur. Commercial Applications and Other Benefits: This work will enhance collaborative science making more efficient use of existing and future networking technologies that will help usher in scientific breakthroughs.

Loading Reservoir Labs, Inc. collaborators
Loading Reservoir Labs, Inc. collaborators