Outside the Box
More Science. Less Money.
Infrastructure Without Compromise
Build impossible servers for HPC + AI workflows — from storage to accelerators to memory — at a fraction of cloud TCO, by optimizing the utilization and efficiency of your existing hardware. Run more workloads faster at lower cost through higher utilization of resources and more agile deployment.
Let your software define your hardware. GigaIO provides workload-defined infrastructure through its dynamic memory fabric, which seamlessly composes rack-scale resources and integrates natively into industry-standard tools.
GigaIO at SC22
Monday,
November 14th
- GigaIO Lighthouse User Group (GLUG) meeting: 11–2pm at Hyatt Regency Dallas
- SDSC Reception Celebrating Vendor Technology: 7–9pm at SDSC booth (#3025)
- Beowulf Bash: 9pm at Gilley’s Dallas
Tuesday,
November 15th
- SDSC National Research Platform (NRP) Panel: 11:00–11:30am at SDSC booth (#3025)
- SC22 Panel: Smackdown! Does HPC Need Composable Infrastructures Now?: 1:30–3:00pm at room C147-148-154
Wednesday,
November 16th
- How Composability Solves Problems for Customers: 11am at Silicon Mechanics booth (#1422)
Thursday,
November 17th
- SC22 Exhibitor Forum: How GigaIO Memory Fabric Enables Dynamic Computing: 1:30–2:00pm (D171)
In GigaIO’s SC22 Booth (#539)
Tuesday,
November 15th
- 12pm: Compose Using Your Existing Tools, (Alan Benjamin, GigaIO)
- 1pm:Â Flexible, Composable, Powerful: GigaIO GigaPod Engineered Solutions, (Brady Black, AMD)
- 3pm: Memory Today Through Memory Tomorrow, (Matt Demas, GigaIO)
- 5pm: Optimizing Parallel Workflows Using a Memory Fabric, (Matt Demas, GigaIO)
Wednesday,
November 16th
- 10am: Memory Today Through Memory Tomorrow, (Alan Benjamin, GigaIO)
- 1:30pm: Prototype National Research Platform, (Frank Würthwein, SDSC)
- 3pm: Time to Buy Into Composable Infrastructure, (Brock Taylor, CIQ)
- 5pm: Preparing for Massively Distributed, Massively Heterogeneous Computing, (Eric Sindelar, Source Code)
Thursday,
November 17th
- 10am: Optimizing Parallel Workflows Using a Memory Fabric, (Guy Adams, GigaIO)
- 12pm: Theoretical or Usable FLOPs as Measure of Cluster Performance? (Matt Demas, GigaIO)
- 2pm: Compose Using Your Existing Tools, (Alan Benjamin, GigaIO)
Are You an HPC Guru?
Take the quiz and be entered into a drawing to win one of several $100 prizes!
Snap a Pic of Giggy at SC22 to Win!
Pick up your Giggy at our booth (#539) to join the fun and enter a daily drawing to win a LEGO® Technic™ Airbus!
Build Impossible Infrastructure With Ease
Easy to obtain
Easy to deploy
Easy to integrate with existing tools
Easy to get higher performance
Easy to save money
Easy to scale into the future
FabreX Remote Shared
Dynamic Memory Fabric
No longer is the server the basic unit of compute. With FabreX, multiple racks can be your server. Transform static infrastructure into software-defined hardware where resources communicate over a native PCIe memory fabric, just as they would if they were plugged directly into the server motherboard.
See demo at our booth.
GigaIO’s Edge-to-Core Solution:
GigaPod and GigaPod Xtreme
Our edge solution reduces time to results and eliminates stagnant intelligence. FabreX, GigaIO’s dynamic memory fabric, unifies edge-to-core infrastructures in order to dynamically deploy any mission application, for actionable real-time intelligence – anywhere.Â
See demo at our booth.
NVIDIA Bright Cluster Manager 9.2 Native Integration of FabreX
A complete solution for building and managing composable clustered infrastructure that combines Bright Cluster Manager’s ability to build and manage clusters with GigaIO’s ability to disaggregate and compose servers, accelerators, and devices within a seamless memory fabric.Â
See demo at our booth.