From the VAST Data edition of November 10, 2026
TACC reports Frontera retired in October 2026 and Horizon in its first phase of early operations with exabyte-capacity storage in place
Texas Advanced Computing Center (TACC), The University of Texas at Austin · Public research university / NSF-funded national research computing facility · Texas, United States (researchers from 30+ institutions)
- Publisher
- TACC News
- Original publication
- 2026-10-28
- Source retrieved
- 2026-11-11
- Event date
- 2026-10-28
What happened
TACC's Oct. 28, 2026 feature on the ninth TACCSTER symposium reports that Frontera was retired in October 2026 after more than eight million jobs and that Horizon, the NSF LCCF centerpiece, has entered its first phase of early operations. Executive Director Dan Stanzione announced that power infrastructure is complete, an exabyte-capacity storage system is in place, and initial deployment of Horizon's AI-focused hardware is underway. The article does not name VAST or give a production date.
Why it matters
Newly relevant tonight because the Nov 10 Horizon User Guide update ties this early-operations milestone to VAST user file systems, and it was resurfaced in Nov 10 regional coverage. For public research universities, it marks the transition of a large VAST-backed academic system from construction to early user operation.
Evidence and measured results
Opened the original TACC news page (by Jorge Salazar, Oct. 28, 2026); read the 'Frontera Sunsets, Horizon Dawns' section. Statements are attributed to TACC leadership. No storage metrics.
Limitations and uncertainty
Predates the last completed run (used for why-now context); institutional news, not an evaluation; VAST not named; 'exabyte-capacity storage system' likely aggregates tiers and should not be attributed wholly to VAST.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-11-10; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Universities retiring flagship HPC systems need a replacement path that keeps researchers productive during migration.
- Stakeholders
- Research computing director
- Vice president for research
- Faculty HPC users
- Facilities and power planning
- Discovery questions
- When does your current primary HPC system reach end of life?
- What interim capacity will users have during migration?
- Which storage tiers are flash versus archive in your plan?
- Value hypothesis
- Interpretation: Horizon's staged opening offers a reference timeline for planning a system transition; it does not show VAST-specific value.
- Potential engagement
- A bounded system-transition planning session that maps an institution's retirement and early-user timeline against Horizon's public milestones.
- Unsupported claims
- That VAST supplied the full exabyte-capacity storage
- Any Horizon production date
- Any performance or user-productivity result
Pre-sales engineering
Role takeaway
- Fit
- Relevant as a milestone reference for large academic systems moving from build-out to early operations.
- Architecture and integration
- Storage was in place before AI hardware deployment finished, implying storage-first staging; integration details come from the user guide, not this article.
- Prerequisites
- Completed facility power
- Storage installed ahead of compute
- Early-user program
- Constraints
- Partial hardware during early operations
- No production date given
- Storage tiers not broken out
- Security
- Not addressed in the article.
- Proof of value
- Confirm storage-tier breakdown with TACC publications
- Track early-user performance reports before citing Horizon
- Validate production date announcements
Delivery
Role takeaway
- Work
- Manage the overlap between retiring a flagship system and opening a successor in early operations.
- Dependencies
- Data migration from the retired system
- Early-user allocations
- User training sessions
- Ownership
- University research-computing center leadership and user-services staff.
- Skills and adoption
- Users porting from CPU-only systems to Grace Blackwell and Vera nodes need training and staff help, as the symposium tutorial on Horizon illustrates.
- Governance checkpoints
- Approve retirement date and data-migration window
- Approve early-user admission criteria
- Production readiness review
- Acceptance criteria
- Proposed: all active projects offered a migration path before retirement
- Proposed: early-user support tickets resolved within an agreed time
- Proposed: production declared only after storage and compute acceptance tests
- Risks
- Users stranded between systems
- Early-operations instability affecting research deadlines
- Misattributing storage capacity to a single vendor in planning
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Confirms system-level staging (power complete, storage in place, AI hardware deployment underway) rather than a finished system; the storage described as exabyte-capacity is not broken out between flash and tape tiers.
Governance
Who approves, reviews and stays accountable for outcomes?
Retirement of the prior flagship forces user migration on the operator's timeline; institutions should plan data migration windows and allocation bridging.
Security and privacy
What data, permissions and controls need testing?
Not addressed.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
The symposium included a getting-started tutorial on Horizon's CPU-GPU architecture, signaling user training as part of the transition.
Procurement
What should contracts, pricing and exit terms secure?
No procurement details; useful only as timeline context for institutions planning successor systems.
Operating model
Which teams own the service once it runs?
Center runs overlapping retirement and early operations, with storage installed ahead of full compute deployment.
Publication history
- 2026-11-10VAST Data · Issue 062 resources
Stable resource ID: tacc-taccster-horizon-early-operations-frontera-retired-20261028