https://sled.lighthouseadvisory.consulting/feeds/vast-data.xml2026-11-11T05:16:08.918ZLighthouse Advisoryhttps://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-11-102026-11-11T05:16:08.918Z2026-11-10T00:00:00.000Z

TACC's Horizon User Guide, updated November 10, documents that early users of the NSF Leadership-Class Computing Facility system at UT Austin work on a shared VAST file system for $HOME and $SCRATCH, without Lustre striping, with a 10-day unbacked scratch purge and a $WORK tier deferred to late 2026. TACC's October 28 report says Horizon has entered only its first phase of early operations as Frontera retires. For public research universities this is an operator design reference, not yet measured performance or reliability evidence.

TACC's Horizon User Guide, updated November 10, documents that early users of the NSF Leadership-Class Computing Facility system at UT Austin work on a shared VAST file system for $HOME and $SCRATCH, without Lustre striping, with a 10-day unbacked scratch purge and a $WORK tier deferred to late 2026. TACC's October 28 report says Horizon has entered only its first phase of early operations as Frontera retires. For public research universities this is an operator design reference, not yet measured performance or reliability evidence. 2 sources · 1 cross-source patterns Horizon's VAST user file systems are in early operations, not yet production evidence Before citing Horizon in a university or state research-computing storage decision, which measured early-user throughput, availability and migration results will TACC publish, and when will $WORK and production status arrive?https://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-11-052026-11-06T11:55:44.271Z2026-11-05T00:00:00.000Z

VAST Data's November 5 release claims more than $10 billion in FY2026 year-to-date commitments from 14 large customers and more than 200 exabytes of flash aligned through 2029, mentioning government organizations among planners. Same-day Blocks & Files reporting qualifies the headline as aggregate commitments (software plus supporting hardware) rather than necessarily orders placed, notes there is no official top-10 frontier-lab ranking, and records VAST's statement that aligned capacity is allocated by advance commitments because demand exceeds supply. For SLED, the signal is supply-chain and procurement timing—not a measured agency or campus outcome.

VAST Data's November 5 release claims more than $10 billion in FY2026 year-to-date commitments from 14 large customers and more than 200 exabytes of flash aligned through 2029, mentioning government organizations among planners. Same-day Blocks & Files reporting qualifies the headline as aggregate commitments (software plus supporting hardware) rather than necessarily orders placed, notes there is no official top-10 frontier-lab ranking, and records VAST's statement that aligned capacity is allocated by advance commitments because demand exceeds supply. For SLED, the signal is supply-chain and procurement timing—not a measured agency or campus outcome. 2 sources · 1 cross-source patterns VAST's $10B demand signal is commitment-scale news that SLED buyers must qualify as delivery and allocation risk For a state or university VAST expansion competing with hyperscaler demand, which written delivery dates, advance-commitment terms and shortage-allocation rules will gate award and acceptance?https://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-11-012026-11-03T03:21:36.884Z2026-11-01T00:00:00.000Z

Data infrastructure (limited SLED relevance): a VAST blog dated October 30 models its Kafka-compatible Event Broker on VAST on Cloud at $238,387 a month against $1.1 million for Confluent Cloud Enterprise at 4 GB/s with five-day retention. It is a vendor model with no customers named, and VAST concedes long retention can cost more. An earlier independent StorageMath analysis says VAST's Event Broker and uptime claims lack the methods needed to check them. No new U.S. state, local or education VAST deployment surfaced, and the CSCS evaluation remains inaccessible.

Data infrastructure (limited SLED relevance): a VAST blog dated October 30 models its Kafka-compatible Event Broker on VAST on Cloud at $238,387 a month against $1.1 million for Confluent Cloud Enterprise at 4 GB/s with five-day retention. It is a vendor model with no customers named, and VAST concedes long retention can cost more. An earlier independent StorageMath analysis says VAST's Event Broker and uptime claims lack the methods needed to check them. No new U.S. state, local or education VAST deployment surfaced, and the CSCS evaluation remains inaccessible. 2 sources · 1 cross-source patterns VAST Event Broker cost and performance claims rest on vendor-only models; public buyers need their own workload baselines Before an agency or university considers VAST Event Broker for telemetry or log streams, what measured workload profile, matched durability settings and contract pricing will it use to test VAST's cost and performance claims?https://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-10-312026-11-02T01:20:58.168Z2026-10-31T00:00:00.000Z

State and higher education: New York's Empire AI issued a vendor-neutral Gamma RFI on October 29 (up to 5 MW, about $160 million across two tranches) that records Beta's 20 PB of VAST storage and asks how new storage would interoperate with it, including across two sites, while its user guide says no data is backed up. Research computing: an LLNL study accepted at PDSW'26 found VAST spared GPU communication from contention but still stalled 12% of training iterations. The CSCS evaluation remains inaccessible.

State and higher education: New York's Empire AI issued a vendor-neutral Gamma RFI on October 29 (up to 5 MW, about $160 million across two tranches) that records Beta's 20 PB of VAST storage and asks how new storage would interoperate with it, including across two sites, while its user guide says no data is backed up. Research computing: an LLNL study accepted at PDSW'26 found VAST spared GPU communication from contention but still stalled 12% of training iterations. The CSCS evaluation remains inaccessible. 2 sources · 1 cross-source patterns Public AI compute buyers need measured storage-path evidence, not rated VAST capacity or throughput For a public AI cluster extending or adding VAST storage, which representative-workload tests of tail latency, fabric placement, cross-site access and restore will gate acceptance?https://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-10-302026-10-31T04:40:20.360Z2026-10-30T00:00:00.000Z

State and higher education: the Massachusetts AI Hub's state-funded AICR cluster runs on about 14.3 PiB of VAST all-flash storage with encryption at rest, but its operator documents snapshot-only protection, no off-site backups and Beta status, with no published performance outcomes. Health and human services: a VAST blog maps encryption, retention locks, WORM and immutable snapshots to HIPAA; it is a vendor feature claim, and HHS's Security Rule overhaul is now projected for July 2027. The CSCS independent evaluation remains inaccessible.

State and higher education: the Massachusetts AI Hub's state-funded AICR cluster runs on about 14.3 PiB of VAST all-flash storage with encryption at rest, but its operator documents snapshot-only protection, no off-site backups and Beta status, with no published performance outcomes. Health and human services: a VAST blog maps encryption, retention locks, WORM and immutable snapshots to HIPAA; it is a vendor feature claim, and HHS's Security Rule overhaul is now projected for July 2027. The CSCS independent evaluation remains inaccessible. 2 sources · 1 cross-source patterns VAST data-protection outcomes depend on operator configuration and policy, not the platform alone For a VAST-based public AI or ePHI data store, which protections are actually enabled, who owns keys and backups, and has recovery been tested?https://sled.lighthouseadvisory.consulting/streams/vast-data/editions/2026-10-292026-10-30T03:21:45.629Z2026-10-29T00:00:00.000Z

First VAST Data edition. Higher-ed research computing: Georgia Tech's NSF-funded Nexus opened on an eight-appliance VAST all-flash layer, though it was pre-production and public specs conflict. Agentic AI: VAST and NVIDIA announced an Agent Sandbox Foundation Stack claiming durable agent records, while independent research on the underlying OpenShell runtime shows exfiltration through permitted egress. Sovereign inference: VAST and Hyperfusion plan containerized, air-gapped units, with validation incomplete. No U.S. state/local government VAST evidence surfaced; the CSCS independent evaluation could not be accessed.

First VAST Data edition. Higher-ed research computing: Georgia Tech's NSF-funded Nexus opened on an eight-appliance VAST all-flash layer, though it was pre-production and public specs conflict. Agentic AI: VAST and NVIDIA announced an Agent Sandbox Foundation Stack claiming durable agent records, while independent research on the underlying OpenShell runtime shows exfiltration through permitted egress. Sovereign inference: VAST and Hyperfusion plan containerized, air-gapped units, with validation incomplete. No U.S. state/local government VAST evidence surfaced; the CSCS independent evaluation could not be accessed. 4 sources · 1 cross-source patterns Agent-sandbox persistence and audit claims need independent security verification Before an agency or university pilots sandboxed agents on VAST, how will it prove that every agent action is recorded and that allowed tools cannot exfiltrate data?