Maximizing Data and Security Protocols
Our custom architecture and mission-critical solutions are designed to maximize data transmission speeds and security protocols above all current standards. By targeting contemporary software design flaws and inefficiencies within current on-chip hardware architecture, our custom generated proprietary solutions generate, as a whole or in unison, the most comprehensive data compression, computing and transmission technology available today.
Products
Compression and Decompression Engines
Our portfolio includes Compression and Decompression Engines (CE/DE) of two sorts. First, devices such as the SigBits CE/DE exploit knowledge, for example knowing that the data is composed of sensor-measurements, to obtain high compression ratios. Second, our general data compression systems, such as FarFetch, provide competitive compression ratios. In both cases the systems present superb throughput, latency, energy consumption and cost-effective performance often uniquely bypassing the current performance measure of alternative solution by orders of magnitudes. Consequently, in some environments such as environments that require high speed data transmission with real time guarantees, our products have no competition.
Compressed Domain Processor
Our proprietary Compressed Domain Processors allow for the computation of compressed data, while still in the compressed domain. Bypassing decompression and reconstruction functionality in the data processing operations, generates tremendous benefits relating to costs, throughput, latency, energy consumption, bandwidth, storage space, and environmental impact.
Solutions
Indexing and Posting Lists
Our SigBits technology is based on high performance, data compression hardware solutions for indexing in high powered information retrieval systems. For example, Google Inverted Indexes, generated through web crawling, cover more than 1011 webpages, including millions of posting lists, indexing millions of basic terms and occupying over 100 Petabytes. Delays associated with the cumbersome processing, ranking, transmission and storage of these enormously large data streams has motivated us in applying new lossless compression techniques to the information retrieval pipeline for yielding high compression ratios, compressed domain processing, search, access, and retrieval.
IoT and Sensor Data Network
Our IoT and sensor data solutions acquire large data sets from distributed heterogeneous sensors via wireline, wireless, and cellular communication networks to compress, decompress, and process sensor data streams at various stages of the transmission and processing for optimized bandwidth, throughput, storage and end user analytics monitoring, and control.
AI
Our lossless and near lossless compression of AI system parameters (e.g., convolutional neural networks coefficients) provides a viable option for real time update of AI systems through low bandwidth communications that span from the sensor, to the edge, to the data center.
General Data
While the solutions described above exploit knowledge about the data, our general data compression and decompression engines (e.g., the FarFetch CE/DE) operate on generic data and provide competitive compression ratios accompanied by extremely high throughput, low latency, low energy consumption, and low cost.