RiVA logo

The Most Powerful Petrotechnical Computing Platform in the Oil & Gas Industry

Read and Write data at 5GBs from your Windows workstation

Volante

Volante is an Enterprise-grade, Ultra-High-Performance platform for interactive subsurface workflows. Choose this platform when performance is paramount. Pre-configured models exist for 100s of workstations and storage requirements of up to 10PBs. Compute Clusters can be embedded in this platform.

CODA

CODA is a smaller platform with good performance for interactive subsurface workflows. Choose this platform when cost is sensitive. Pre-configured models exist for storage requirements of up to 5PB.

Computare

Computare is a compute cluster for processing non-interactive subsurface workloads. Cluster nodes can be optimized for core speed, core density, or GPU-acceleration.

Brahma

The Brahma platform excels at parallel processing, which makes it ideal for AI workloads such as machine learning. It is particularly well-suited for training AI models, where complex mathematical operations are performed on massive datasets.

Atlas

Atlas is a cost-effective platform for hosting OSDU, either on-premises or in a private cloud. OSDU is a cross-company data platform for the oil and gas industry that aims to create an open-source, standards-based, and technology-agnostic data platform.

Volante and CODA platforms include Remote Visualization Systems (RVS) that host the petrotechnical workstations. The RVS are separated into two series to address the wide variety of requirements presented to us by our customers. Each RiVA platform can be customized by installing RVS in any combination.

If your application runs on a x86 system, then we can run it on RiVA. We support Linux and Windows equally. Out platforms work with every G&G application, and that's by design. Furthermore, every platform comes with all your applications installed, tested, and ready to go.

windows enterprise 11 logorhel 9 logo

H-Series for virtual workstations

Optimized for high performance, each RVS can host 2, 4 or 8 virtual workstations. The virtual workstation options are as follows:

Option
vCPUs
(3.6GHz)
VM Memory
(5600MHz)
vGPU Memory
nVidia GPU
1
8
128GB
6GB
L4
2
16
256GB
12GB
L4
3
16
256GB
24GB
L40S
4
32
512GB
24GB
L40S
5
32
512GB
48GB
L40S

X-Series for bare metal workstations

Provides extreme performance to address exceptionally demanding workloads. These bare-metal workstations employ the most powerful Intel processors and nVidia GPUs available. The bare-metal workstation options are as follows:

Option
Processor Cores
(3.9GHz)
Memory
(5600MHz)
GPU Memory
nVidia GPU
1
16
512GB
48GB
RTX 6000 Ada

Compute Clusters

Compute Clusters can now be embedded into a RiVA platform allowing them to leverage services provided by the RiVA Operating Environment, such as monitoring, patching, and security. Compute cluster nodes can be configured with or without GPUs. GPU acceleration has become essential for modern workflows, offering significant performance improvements over CPU-only clusters. 400Gbs Infiniband interconnects can also be configured as an option.

There are seven bare metal compute cluster node configurations available:

Option
Profile
Processor Cores
Clock Speed
Memory
(5600MHz)
GPU Memory
nVidia GPU
1
High Clock Speed
32
3.6GHz
256GB
2
32
3.6GHz
512GB
3
High Core Density
64
2.5GHz
512GB
4
64
2.5GHz
1024GB
5
GPU Accelerated
32
3.6GHz
256GB
24GB
L4
6
32
3.6GHz
512GB
48GB
L40S
7
32
3.6GHz
1024GB
96GB
H100

Production (PROD)

︿

All platforms can be implemented in a Production capacity.

Disaster Recovery (DR)

︿

All platforms can be implemented in a Disaster Recovery capacity. DR platforms are typically a pared down version of its PROD sibling.

Replication Target (RT)

︿

Only Volante and CODA platforms can be configured as Replication Targets. RT platforms are like a DR platforms except there are no servers or workstations.

On-Premises Platform

︿

The RiVA platform is physically situated within the customer's data center and is directly connected to the core network, which is overseen by the customer's IT department. The RiVA Operating Environment may leverage various network services provided by the customer, such as Authentication, and will be safeguarded by the customer’s security policies and services.

Hosted Platform - CloudConnect

︿

This variation of the on-premises deployment differs in that GeoComputing manages the data center cage where the RiVA platform is installed. The cage may be located in the same data center as the customer’s core infrastructure or at a remote datacenter connected to the customer’s core network via a private data link. Regardless of the location, GeoComputing provides the cage, physical security, power, and network connectivity.

Single-Tenant Private Cloud

︿

With a single-tenant private cloud platform deployment, all hardware and software resources are dedicated to the customer’s organization. GeoComputing manages all aspects of the platform and offers connectivity over the Internet or via private data links. Unlike public cloud providers, your private cloud can be located wherever you want, which can be an important consideration with respect to data sovereignty.

Multi-Tenant Private Cloud

︿

This variation of the single-tenant private cloud deployment differs in that some hardware and software resources are shared amongst several customers to reduce cost. Having said that, shared hardware resources are logically partitioned and secured from one another. The server and workstation estates remain dedicated resources.

RiVA servers
RiVA server
RiVA servers
RiVA server

Data Management

︿

Diskover Data offers comprehensive visibility and control over all your unstructured petrotechnical data, both within and beyond the RiVA platform. By indexing and enriching metadata across various file systems, cloud storage solutions, and legacy archives, Diskover unveils the complete landscape of your data estate, regardless of its location. When integrated with RiVA, Diskover amplifies GeoComputing’s robust data management capabilities, allowing users to efficiently search, filter, tag, and manage data across the entire enterprise. From initial discovery to informed decision-making, Diskover bridges the gap between storage and scientific analysis, streamlining workflows and fostering sustainable data management practices.

Data Platform (OSDU)

︿

OSDU is a cross-company data platform for the oil and gas industry that aims to create an open-source, standards-based, and technology-agnostic data platform. The OSDU Forum, which was founded in September 2018, provides a vendor-neutral environment for the development of the platform and ecosystem. The platform's goals include: stimulating innovation, industrializing data management, and reducing time to market for new solutions.

OSDU is based on a NoSQL database. A NoSQL database can significantly help with AI pipelines, particularly when dealing with large volumes of unstructured or semi-structured data, thanks to its flexible data model and high scalability, making it suitable for storing and accessing data needed for training and deploying machine learning models rapidly.

Key benefits of using NoSQL for AI pipelines:

  • Flexible data structures: NoSQL allows storing diverse data types like text, images, and JSON documents without rigid schema constraints, which is often necessary for AI applications that deal with complex data formats.
  • High scalability: NoSQL databases can easily scale horizontally to handle large data volumes generated by AI training processes, especially when dealing with distributed computing environments.
  • Fast query performance: Many NoSQL databases are optimized for fast read/write operations, which is crucial for real-time AI applications requiring quick data access.
  • Cost-effectiveness: Depending on the specific implementation, NoSQL can be more cost-effective than traditional relational databases for large-scale AI projects due to its efficient data storage mechanisms.

Compute Cluster Management

︿

Spectrum LSF is a workload management platform and job scheduler for Compute Clusters. This is the default tool for RiVA platforms.

AI Pipeline Development

︿

nVidia AI Enterprise isa cloud-native software platform that provides AI practitioners and IT professionals with the tools and frameworks needed to develop, deploy, and manage AI applications. It includes:

  • nVidia TAO Toolkit: A tool that allows users to combine pre-trained models with their own data to create custom AI models
  • nVidia NIM and CUDA-X microservices: These microservices provide an optimized runtime and building blocks for generative AI development
  • nVidia NeMo: A framework that allows users to customize pre-trained models and select community models for specific use cases
  • nVidia Riva: A GPU-accelerated SDK for speech and translation AI
  • nVidia Maxine: A developer platform for adding augmented reality effects and enhancing audio and video
  • nVidia AI Workflows: A collection of packaged reference applications that include pretrained models, training pipelines, and Jupyter Notebooks
  • nVidia Tensor RT: A tool for GPU-optimized deep learning inference
  • Triton Inference Server: A tool for deploying trained AI models at scale

Security Information & Event Management

︿

Booli is a Security Information and EventManagement (SIEM) software platform that helps organizations identify and respond to security threats. SIEM systems collect and analyze data from various sources, such as servers, firewalls, and applications, to identify potential threats.

GeoComputing partners with Fulcrum Technology, who operate Booli as a Managed Detection & Response (MDR) service.

Vulnerability Scanner

︿

Nessus is a widely used vulnerability scanner tool, which scans computer systems and networks to identify potential security weaknesses and vulnerabilities that could be exploited by attackers, essentially acting as a proactive tool to identify and address security issues before they are exploited; it does this by checking for outdated software, misconfigurations, and known vulnerabilities across various systems and applications.

Endpoint Security

︿

Endpoint security is a set of technologies and practices that protect devices that connect to a network from malicious threats and cyberattacks. These devices are called endpoints, and can include laptops, desktops, mobile phones, tablets, smart watches, printers, servers, and more Palo Alto Cortex XDR is an endpoint security product that can prevent unauthorized access to networks, applications, and data stores. In addition, it can:

  • Flexible data structures: NoSQL allows storing diverse data types like text, images, and JSON documents without rigid schema constraints, which is often necessary for AI applications that deal with complex data formats.
  • High scalability: NoSQL databases can easily scale horizontally to handle large data volumes generated by AI training processes, especially when dealing with distributed computing environments.
  • Fast query performance: Many NoSQL databases are optimized for fast read/write operations, which is crucial forreal-time AI applications requiring quick data access.
  • Cost-effectiveness: Depending on the specific implementation, NoSQL can be more cost-effective than traditional relational databases for large-scale AI projects due to its efficient data storage mechanisms.

Data Encryption

︿

Encryption in Transit is a security measure that protects your data being transmitted between devices over a network from interception, eavesdropping, and tampering. Enabling Encryption in Transit will increase cost and decrease performance.

Encryption at Rest refers to the practice of encrypting data on a storage device. It's a security measure to safeguard data when it's not actively being used or transferred. All storage in a RiVA platform is Encrypted at Rest.

Platform Security

︿

The Platform Security option offers a variety of physical security features including integrated temperature and humidity sensors, intrusion detection sensors, and surveillance cameras that all integrate into the platform’s Event Management software.

Providing the best Petrotechnical Computing in the E&P Industry

To learn more about Geocomputing Group and RiVA, please contact us.
About Us Support Services Contact Us