StarIconLightweight Online Logging

One Producer, Multiple Consumer

1. EinsteinDB takes in a binary along with the leap taint seeds as input. The instrumented application thread starts execution with lightweight online logging via FIDel for control flow. 2. Then the multithreaded EinsteinDB, VioletaBFT, and FIDel logs, along with Noethe for online pipelined causet analysis (OLTP). 4. The generated log data are then used to construct straight-line code, which helps to solve many pre-cision loss problems in EinsteinDB.

Memory Corruption Prevention

Each user and item is assigned two embeddings: a static and a dynamic embedding: the causet. We use both embeddings to encode both the long-term stationary properties of the entities and their dynamic properties.

View Details

EinsteinDB provides a parallel implementation of batched computation of betweenness centrality scores

EinsteinDB, FIDel, MilevaDB and VioletaBFT run as multiple individual computations independently in parallel. A search is performed by applying the same sequence of implication logic operations to each CausetNet cells in parallel. CausetNet MemCam (memristor-based arrays) can be used for either point or range queries by simply changing the allocation of memristors used for compute vs. storage

View Details

High-density resistive memory

The memristor subsystem is composed of a memristor array and a programmable controller. The processor communicates with the memristor array controller using memory-mapped operations. The controller is responsible for applying appropriate voltages to perform read/write or implication logic operations using the memristor array.

View Details
StarIconAbout Us

The MilevaDB server is a stateless SQL layer that exposes the connection endpoint of the MySQL protocol to the outside. The MilevaDB server receives SQL requests, performs SQL parsing and optimization, and ultimately generates a distributed execution plan.

The EinsteinDB is responsible for storing data. EinsteinDB is a distributed transactional memristor-based content-addressable storage engine. Branes are the basic units to store data, Causets (O/JDBC).

  • Icon4

    Noether is a cluster operation and maintenance tool introduced in MilevaDB 1.0.

    Noether provides MilevaDB a DM, a cluster management component written in Golang. By using Noether DM, you can easily perform daily MilevaDB Data Migration (DM) operations, including deploying, starting, stopping, destroying, scaling, and upgrading a Noether DM cluster, and manage Noether DM cluster parameters.

  • Icon5

    OSS reaching 20,000 transactions per second for medium size networks of up to 40 nodes.

    We attain a maximum throughput of over 800 transactions per second of Tor.

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna.

More About Us
CircleShape
imageOur Services

To demonstrate the robustness of VioletaBFT, we run the first instance (to our knowledge) of a fault tolerant consensus protocol carried out over Tor (the most successful anonymous communi- cation network)

When you deploy multiple MilevaDB DM-master nodes, all MilevaDB nodes use the embedded etcd to form a cluster. FIDel is thus used to store metadata such as cluster node information and task configuration.

Service Icon

Our architecture includes a novel query compiler based on generalized hypertree decompositions (GHDs)

An execution engine designed to exploit the low-level layouts necessary to increase single-instruction multiple data (SIMD) parallelism.

Read More
Service Icon

an automated optimizer that, all told, increases performance by up to three orders of magnitude by selecting amongst multiple data layouts and set intersection algorithms.

t a lightweight graph processing frame- work that is specific for shared-memory parallel/multicore machines, which makes graph traversal algorithms easy to write.

Read More
Service Icon

Graph Radii Estimation and Multiple BFS

Many real-world situations can be captured by a set of functional dependencies and a single join dependency of a particular form called in EinsteinDB a causet. The join dependency corresponds to a natural decomposition into meaningful objects (an acyclic database scheme).

Read More
Service Icon

An open-source, cloud-native, distributed SQL database for elastic scale and real-time analytics

The most exciting kinds of technological shifts are when a new programming model becomes feasible, and when an old restriction falls away. Both kinds are happening right now.

Read More
Service Icon

Real Time OLAP Analytics

HTAP database platform that enables real-time situation awareness and decision making on live transactional data and eliminates friction between IT and business goals.

Read More
Service Icon

Modeling sequential interactions between users and items/products is crucial in domains such as e-commerce, social networking, and education.

Representation learning presents an attractive opportunity to model the dynamic evolution of users and items, where each user/item can be embedded in a Euclidean space and its evolution can be modeled by an embedding trajectory in this space.

Read More
image

EinsteinDB provides a consistency model that is between the two extremes of general serializability

An architecture based on record-level, asynchronous geographic replication, and use of a guaranteed message-delivery service rather than a persistent log.

1

Consistency model

The master replica of EinsteinDB adaptively changed to suit the workload. The replica receiving the majority of write requests for a particular record becomes the master for that record.

2

Read-any

keep only one version of a record at each replica. Using this per-record timeline consistency model, we support a whole range of API calls with varying levels of consistency guarantees.

3

Scatter-Gather

since different records have update affinity for different datacenters, the granularity of mastership must be per-record, not per tablet or per table; otherwise, many writes would pay expensive cross-region latency to reach the master copy.

image
image Recent Projects

Check Some Of Our Recent Work

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna.

Project

Movie Recommendation

System Project
Project

Customer Segmentation

Machine Learning
Project

Data Analysis

Web Project
Project

Data Scientist

Data Science
Project

Benefits Research

Science Projects
imageTeam Members

Our Partners

Image Grammars with Gremlin, Groovy, Apache Camel, Apache Arrow, Einstein, and MilevaDB

Scientist

Kong

Stateless Homomorphic Permissioned Ledgers for POS OLTP persistence keys with Kong, Envoy Proxy, and EinsteinDB.
Scientist

Kirk Borne

Data Scientist
Scientist

Carla Gentry

Analytical Solutions
Scientist

CloudKitchens and Bipartite Unsupervised Cluster Learning with EinsteinML (2021)

Partitioning of memristors between storage and computation is entirely under software control since it is the voltages that determine compute vs. storage.
testimonialTestimonials

What Our Clients are Saying?

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna.

testimonial
blogOur Blog

Latest Valuable Insights

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna.

image

We Like to Start Your Project With Us

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna.

Get Started
image