astra: extract claims from 2025-12-10-starcloud-h100-gpu-orbit-first-llm-trained
Some checks failed
Mirror PR to Forgejo / mirror (pull_request) Has been cancelled

- Source: inbox/queue/2025-12-10-starcloud-h100-gpu-orbit-first-llm-trained.md
- Domain: space-development
- Claims: 1, Entities: 0
- Enrichments: 2
- Extracted by: pipeline ingest (OpenRouter anthropic/claude-sonnet-4.5)

Pentagon-Agent: Astra <PIPELINE>
This commit is contained in:
Teleo Agents 2026-04-14 17:17:54 +00:00
parent 069e41b899
commit dca9a7d5fe

View file

@ -1,7 +1,7 @@
---
type: claim
domain: space-development
description: The H100 demonstration establishes TRL 7 for commercial GPUs in low-altitude LEO but does not validate the 500-1800km radiation environment proposed for large-scale orbital data center constellations
description: The H100 demonstration establishes TRL 7 for commercial GPUs in low LEO but does not validate the 500-1800km radiation environment proposed for large-scale ODC constellations
confidence: experimental
source: CNBC, Starcloud-1 mission December 2025
created: 2026-04-14
@ -9,11 +9,11 @@ title: Starcloud-1 validates commercial GPU viability at 325km LEO but not highe
agent: astra
scope: structural
sourcer: CNBC
supports: ["orbital-data-centers-activate-bottom-up-from-small-satellite-proof-of-concept-with-tier-specific-launch-cost-gates", "modern AI accelerators are more radiation-tolerant than expected because Google TPU testing showed no hard failures up to 15 krad suggesting consumer chips may survive LEO environments"]
challenges: ["radiation-hardening-imposes-30-50-percent-cost-premium-and-20-30-percent-performance-penalty-on-orbital-compute-hardware"]
related: ["orbital-data-centers-activate-bottom-up-from-small-satellite-proof-of-concept-with-tier-specific-launch-cost-gates", "modern AI accelerators are more radiation-tolerant than expected because Google TPU testing showed no hard failures up to 15 krad suggesting consumer chips may survive LEO environments", "radiation-hardening-imposes-30-50-percent-cost-premium-and-20-30-percent-performance-penalty-on-orbital-compute-hardware"]
supports: ["modern-ai-accelerators-are-more-radiation-tolerant-than-expected-because-google-tpu-testing-showed-no-hard-failures-up-to-15-krad-suggesting-consumer-chips-may-survive-leo-environments"]
challenges: ["orbital-data-centers-require-five-enabling-technologies-to-mature-simultaneously-and-none-currently-exist-at-required-readiness"]
related: ["starship-achieving-routine-operations-at-sub-100-dollars-per-kg-is-the-single-largest-enabling-condition-for-the-entire-space-industrial-economy", "orbital-data-centers-require-five-enabling-technologies-to-mature-simultaneously-and-none-currently-exist-at-required-readiness", "modern-ai-accelerators-are-more-radiation-tolerant-than-expected-because-google-tpu-testing-showed-no-hard-failures-up-to-15-krad-suggesting-consumer-chips-may-survive-leo-environments", "radiation-hardening-imposes-30-50-percent-cost-premium-and-20-30-percent-performance-penalty-on-orbital-compute-hardware", "starcloud-1-validates-commercial-gpu-viability-at-325km-leo-but-not-higher-altitude-odc-environments", "modern AI accelerators are more radiation-tolerant than expected because Google TPU testing showed no hard failures up to 15 krad suggesting consumer chips may survive LEO environments"]
---
# Starcloud-1 validates commercial GPU viability at 325km LEO but not higher-altitude ODC environments
Starcloud-1 successfully operated an NVIDIA H100 GPU in orbit at 325km altitude from November-December 2025, training NanoGPT, running Gemini inference, and fine-tuning models. This establishes TRL 7 (system prototype demonstration in operational environment) for commercial datacenter-grade GPUs in space. However, the 325km altitude is significantly more benign than the 500-1800km range proposed by SpaceX and Blue Origin for large-scale ODC constellations. At 325km, the satellite operates well inside Earth's magnetic shielding and below the Van Allen belts' intense radiation zones. The 11-month expected mission lifetime is naturally limited by atmospheric drag at this altitude, meaning long-term radiation degradation curves remain unknown. Neither Starcloud nor NVIDIA disclosed radiation-induced error rates or performance degradation metrics. The demonstration proves commercial GPUs can survive LEO's vacuum and thermal cycling, but the radiation environment at higher altitudes—where most ODC proposals target—remains unvalidated.
Starcloud-1 successfully operated an NVIDIA H100 GPU in orbit at 325km altitude from November-December 2025, training NanoGPT, running Gemini inference, and fine-tuning models. This establishes Technology Readiness Level 7 (system prototype demonstration in operational environment) for datacenter-grade GPUs in space. However, the demonstration's altitude is critical: 325km is well inside Earth's magnetic shielding and below the Van Allen belts' intense radiation zones. The proposed ODC constellations from SpaceX (1M satellites) and Blue Origin (Project Sunrise) target 500-1800km altitudes where radiation exposure is significantly higher. The 11-month expected mission lifetime also limits what can be learned about long-term radiation degradation. Starcloud and NVIDIA have not disclosed radiation-induced error rates or performance degradation data. This creates a validated proof-of-concept for the benign LEO environment while leaving the commercial ODC altitude range unvalidated.