GIGABYTE AI TOP ATOM NVIDIA DGX Spark Grace Blackwell Computer - 1TB / 4TB

The Gigabyte AI TOP ATOM ships as a complete personal AI supercomputer: the GB10-based system, the 240W external power adapter, USB-C power cable, and Gigabyte's documentation. For teams that need deeper deployment help, solutions are available from the EMARQUE AI team on request (pre-validation testing, setup assistance, EMARQUE AI stack, workflow tuning).

Gigabyte's AI TOP ATOM is one of NVIDIA's partner-built reference designs in the DGX Spark family (Dell, ASUS, HP, and others all make their own variants). Because all GB10-based DGX Spark units share identical silicon, the practical differences between partner variants come down to thermal design, storage hardware, and chassis layout. Gigabyte's variant ships with the Samsung PM9E1 NVMe SSD (independent testing measured among the strongest storage I/O results in the Spark family via GPU Direct Storage), a chassis tuned to permit aggressive burst thermal envelopes rather than conservative power-capping, and rear-mounted I/O — ConnectX-7 SmartNIC, HDMI 2.1a, USB 3.2 Gen 2×2 — for clean rack or shelf integration.

Published Gigabyte AI TOP ATOM specifications: NVIDIA GB10 Grace Blackwell Superchip (20-core Arm CPU + Blackwell GPU); 128GB LPDDR5X unified memory; up to 4TB PCIe Gen5 NVMe storage; 10GbE networking plus NVIDIA ConnectX-7 SmartNIC for Spark Stacking; 3× USB 3.2 Gen 2×2 Type-C ports (data) and a separate Type-C power input; 240W external power adapter; 150mm × 150mm × 50.5mm chassis at 1.2kg.

Performance ceiling is approximately 1 petaFLOP of AI compute (FP4 sparse). For exact stock-level config questions, email business@emarque.co.

The Gigabyte AI TOP ATOM ships with Gigabyte's 3-year manufacturer warranty covering hardware faults on the unit and its components. EMARQUE's built-in role is local handling of that warranty:

• Local point of contact — pre-sale, post-sale, and warranty all routed through the same EMARQUE team. Real person, no ticket queue.
• RMA routing — if a hardware fault appears, we route the Gigabyte warranty claim on your behalf so you don't have to navigate the manufacturer's process yourself.

For deeper post-sale needs — pre-validation testing, OS reinstall, driver issues, network setup, Spark Stacking configuration, EMARQUE AI stack deployment — solutions are available from the EMARQUE AI team on request. Scope and charges quoted on enquiry.

Independent hands-on testing by StorageReview measured concrete throughput on the Gigabyte AI TOP ATOM:

• Llama 3.1 8B inference — up to 2,757 tokens/sec at batch size 64
• GPT-OSS-120B — sustained ~650 tokens/sec on equal-workload runs, peaking near 2,770 tokens/sec during prefill phases
• Samsung PM9E1 NVMe SSD — 11.6 GB/s read and 12.2 GB/s write via GPU Direct Storage, among the strongest storage results in the DGX Spark group
• Unified memory bandwidth — 273 GB/s LPDDR5x

Reviewers also observed the unit running 80°C+ on the GPU and 90°C on the CPU under aggressive prefill workloads without thermal throttling. Gigabyte's chassis is tuned to permit short burst-acceleration periods rather than conservatively power-capping — a meaningful real-world advantage for time-sensitive inference. Idle temperatures sit at ~38°C, with GPU peak power well within the 240W envelope.