Normalized claim
Benchmark PoC performance score: 96% increase
Alibaba Cloud's PAI Lingjun AI infrastructure achieved a high-performance score of 96%.
Futureverse: PAI Lingjun Intelligent Computing for foundation model training
Use case typeAI model trainingUpdated 1 day ago
Futureverse is a New Zealand-based AI and metaverse technology unicorn that uses Alibaba Cloud's PAI Lingjun Intelligent Computing Service as the heterogeneous computing foundation for foundation model training. The company is developing JEN-1, a text-to-music generator that aims to deliver higher fidelity audio, longer and more complex musical compositions, and efficient large language model training.
- Organization
- Futureverse
- Industry
- Tech & Comms
- Location
- New Zealand
- Published
- July 2026
Reported outcomes
85.7 score
JEN-1 perceptual qualityQuality & accuracy
+96%benchmark PoC performance score83.8 scoreexisting methods perceptual quality1 ratioLLM training speedup
Strategic outcomes
Cost efficiencyFoundation model training was optimized and acceleratedOther strategic outcomeImproved text-to-music generation quality and complexity
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
JEN-1 perceptual quality: 85.7 score increase
achieved a notably higher perceptual quality (85.7/100) compared to existing methods (83.8/100)
Normalized claim
Existing methods perceptual quality: 83.8 score
compared to existing methods (83.8/100)
Normalized claim
LLM training speedup: 1 ratio increase
achieve linear speedup close to 1 for their LLM training
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Futureverse
- Provider
- Alibaba
- Maturity
- Exploring
- Linked source
- Alibaba Cloud Customers
Alibaba Cloud provided integrated computing, storage, networking, and full lifecycle AI engineering capabilities to support JEN-1 evaluation, demos, and production training optimization
Customer identity supportedSource describes one deploymentMaturity supported
Primary read
Use case focus
Showing 2 of 2
- 1AI model training
- 2Training infrastructure modernization
- Futureverse needed robust, reliable AI training infrastructure for foundation model training.
- The company required high computational efficiency and perceptual quality for JEN-1, its text-to-music generation model.
- Futureverse used Alibaba Cloud Platform for AI (PAI) Lingjun Intelligent Computing Service as on-demand, single-tenant AI infrastructure for foundation model training.
- Alibaba Cloud provided integrated computing, storage, networking, and full lifecycle AI engineering capabilities to support JEN-1 evaluation, demos, and production training optimization.
- Alibaba Cloud's PAI Lingjun AI infrastructure achieved a high-performance score of 96% in benchmark PoCs versus multiple vendors.
- JEN-1 achieved higher perceptual quality at 85.7/100 compared with existing methods at 83.8/100.
- Futureverse reported linear speedup close to 1 for LLM training.
- The foundation model training process was significantly optimized and accelerated on Alibaba Cloud.
Architecture
PAI Lingjun Intelligent Computing Service provided single-tenant AI infrastructure and heterogeneous computing foundation for foundation model training, with integrated computing, storage, networking, and full lifecycle AI engineering support.
Sources & evidence1
Evidence: Medium65/100Evidence strength
- Customer explicitly identified
- Deployment status explicitly supported
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Jul 28, 2026Publisher: Alibaba CloudEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
Browse the catalog
Was this useful?
Community
Comments
No published comments yet.