gStor
AI-ready, Media-ready, GPU-enabled Storage Appliance
Apace gStor is an AI-ready, GPU-accelerated storage appliance engineered specifically to supercharge media workflows directly at the storage layer. By leveraging dedicated GPU acceleration, gStor offloads massive parallel computational tasks like real-time proxy generation, AI-driven media indexing, facial recognition, and object detection. This GPU-integrated architecture delivers an astonishing 10x speedup in proxy creation compared to legacy CPU-bound systems, completely bypassing host CPU bottlenecks for instant search, smarter discovery, and friction-free team collaboration.
Compute Without Limits
Every element of gStor is engineered to accelerate compute resources and maximize parallel throughput rates.
AI-optimized Architecture
Establish direct DMA memory paths that connect PCIe Gen5 physical flash namespaces directly with GPU virtual address spaces.
Massive Storage Capacity
Drive unthrottled performance up to 128 Gbps per Gen5 slot, supporting multi-GPU nodes under massive scale.
Enterprise-class Reliability
Seamlessly deploy into high-speed physical network frameworks supporting InfiniBand NDR or active 200G/400G RoCE fabrics.
Intelligent Data Processing
Dynamically expand capacity and system bandwidth linearly by scaling node shelves without causing operations downtime.
Seamless Integration
Deploy a proprietary low-latency metadata engine engineered to process millions of tiny randomized directories instantly.
Scalable & Future-Proof
Ensure continuous uptime with active-active redundant power supplies, triple-mirror parity, and automatic disk scrubbers.
gStor Specification Options
2U Model
inches: 16.93 x 26.8 x 3.5
3U Model
inches: 17 x 26.8 x 5.2
4U Model
inches: 17 x 27.6 x 7