HP Inc BM6G3UT#ABA Z2 Mini G1i Compact Workstation Desktop PC
Overview
The HP Z2 Mini G1i is a small-footprint workstation built around Intel's 15th-generation Core Ultra 7 265 — a 20-core processor that reaches 5.30 GHz boost — paired with an NVIDIA Quadro RTX A100 GPU and 64 GB of RAM in a chassis that measures just 6.80 x 11.70 x 19.60 inches and weighs 11.2 lbs. The BM6G3UT#ABA configuration ships with 1 TB of M.2 PCIe NVMe 2280 TLC 4X4 SSD storage and Intel AI Boost delivering up to 13 TOPS of on-device inference acceleration. For security operations centers, video analytics deployments, or engineering workstations where rack or desk space is at a premium, this unit delivers full professional-workstation compute in a form factor closer to a NUC than a tower. Explore the broader HP Inc workstation and compute line for complementary configurations.
Key Features
- Intel Core Ultra 7 265 — 20 Cores, up to 5.30 GHz: With 20 cores (15th-generation architecture) and a single-core boost of 5.30 GHz from a 1.80 GHz base, this processor handles both heavily threaded workloads — simultaneous multi-stream video decode, model inference, data processing — and latency-sensitive single-threaded tasks without switching machines. For a compact workstation in a space-constrained installation, that range matters.
- NVIDIA Quadro RTX A100 GPU: The discrete Quadro RTX A100 GPU is purpose-built for professional visualization, compute, and AI inference workloads. Unlike consumer GeForce cards, Quadro-class hardware carries certified driver support for professional applications — important if you're running GPU-accelerated VMS analytics, CAD, or scientific compute in a regulated environment.
- Intel AI Boost — up to 13 TOPS: The integrated NPU delivers up to 13 TOPS of dedicated AI inference throughput on-chip, offloading lightweight inferencing tasks (object classification, scene tagging, local LLM prompting) from the CPU and GPU. This keeps the main cores and GPU headroom free for heavier workloads running concurrently — a practical gain in multi-application SOC or analytics server deployments.
- 64 GB RAM: 64 GB of system memory is enough headroom to run a VMS server process, a GPU-accelerated analytics engine, and a management console simultaneously without memory pressure forcing paging. For NVR or video management server use cases where multiple software stacks share one host, this is the configuration to reach for.
- 1 TB M.2 PCIe NVMe 2280 TLC 4X4 SSD: The NVMe 4x4 interface (PCIe Gen 4 x4) delivers sequential read/write speeds that keep OS, application, and working-dataset I/O from becoming a bottleneck. TLC NAND at this capacity offers a reasonable endurance profile for workstation OS and application use — though for sustained 24/7 high-write surveillance recording at scale, pair this unit with external or NAS-based storage rather than relying solely on this drive.
- Mini Form Factor — 11.2 lbs, 6.80 x 11.70 x 19.60 in: The Z2 Mini G1i fits under a monitor, on a shelf, or inside a small equipment cabinet where a standard tower workstation won't go. At 11.2 lbs it's light enough to wall-mount or VESA-mount behind a display in a control room. For deployments where every cubic inch of rack or furniture space is allocated, this form factor removes a real constraint.
- SmartBuy Configuration (BM6G3UT#ABA): HP SmartBuy SKUs are channel-standardized configurations — meaning consistent bill-of-materials across units, predictable imaging and deployment scripting, and no per-unit configuration variance to manage. For IT teams managing fleet deployments or refresh cycles, that consistency reduces provisioning overhead.
Integration and Compatibility
The Z2 Mini G1i's Intel Core Ultra platform supports standard enterprise connectivity and management stacks. The Quadro RTX A100 GPU is compatible with NVIDIA's professional driver ecosystem, including CUDA for GPU-accelerated analytics and compute frameworks commonly used in video intelligence and AI inference pipelines. The NPU's 13 TOPS of AI Boost throughput is accessible via Intel's OpenVINO toolkit and compatible Windows AI frameworks, making this unit a reasonable host for edge inference workloads alongside its primary workstation role.
For physical installation, the unit's compact footprint pairs naturally with a KVM switch in multi-workstation control environments, or deploys headless behind a display in a security operations center. The SmartBuy SKU designation supports volume imaging and deployment scripting workflows common in enterprise IT environments.
When evaluating storage for video-intensive deployments, note that the 1 TB NVMe SSD is appropriate for OS and application hosting. High-retention video recording workloads should be directed to dedicated network-attached or direct-attached storage rather than the internal drive to preserve drive endurance and ensure adequate capacity.
Frequently Asked Questions
Q: What processor does the BM6G3UT#ABA include, and how many cores does it have?
A: The BM6G3UT#ABA is configured with an Intel Core Ultra 7 265, a 15th-generation processor with 20 cores running at a base of 1.80 GHz and boosting up to 5.30 GHz.
Q: What GPU is included in the HP Z2 Mini G1i BM6G3UT#ABA?
A: This configuration includes an NVIDIA Quadro RTX A100 GPU, a professional-class discrete graphics card suited for visualization, compute, and GPU-accelerated analytics workloads.
Q: What does Intel AI Boost (up to 13 TOPS) mean for this workstation?
A: Intel AI Boost refers to the integrated NPU (Neural Processing Unit) in the Core Ultra 7 265, which delivers up to 13 TOPS of dedicated on-chip AI inference throughput. This allows lightweight AI inference tasks to run on the NPU independently, freeing CPU and GPU resources for other concurrent workloads.
Q: How much RAM and storage does the BM6G3UT#ABA ship with?
A: It ships with 64 GB of system memory and a 1 TB M.2 PCIe NVMe 2280 TLC 4X4 SSD.
Q: Is the Z2 Mini G1i suitable as a VMS or video analytics server?
A: The combination of a 20-core CPU, NVIDIA Quadro RTX A100 GPU, 64 GB RAM, and Intel AI Boost makes it capable of handling VMS hosting and GPU-accelerated video analytics. For high-retention video recording, supplement the internal 1 TB NVMe SSD with external or network-attached storage.
Q: What are the physical dimensions and weight of the BM6G3UT#ABA?
A: The unit measures 6.80 x 19.60 x 11.70 inches (L x W x H) and weighs 11.20 lbs, making it compact enough for under-desk, shelf, or VESA-mount installation.
The BM6G3UT#ABA is one of those configurations that earns a second look once you work out what's actually inside. A 20-core Intel Core Ultra 7 265 boosting to 5.30 GHz, an NVIDIA Quadro RTX A100 GPU, 64 GB of RAM, and Intel AI Boost at up to 13 TOPS — all in a chassis you can hold with one hand. For security integrators and IT architects evaluating compact hosts for analytics-heavy deployments, the Z2 Mini G1i punches significantly above its size.
Technical Highlights:
- 20-Core CPU at 5.30 GHz Boost: The Core Ultra 7 265's 20-core count handles parallel workloads — multi-stream decode, concurrent application stacks, background management tasks — without the single-core performance penalty you'd accept from a lower-clocked embedded processor. 5.30 GHz boost keeps latency-sensitive operations responsive.
- NVIDIA Quadro RTX A100 GPU: Quadro-class hardware means certified professional drivers and ISV-validated support for compute and visualization software — not just gaming-optimized firmware. For GPU-accelerated VMS analytics or AI inference pipelines, driver stability and certification matter as much as raw CUDA throughput.
- Intel AI Boost — 13 TOPS NPU: 13 TOPS on the NPU isn't replacing a data-center GPU for heavy model training, but it's meaningful for edge inference: running a local classification model or scene-analysis pipeline on the NPU leaves the Quadro and CPU free for rendering and application processing simultaneously.
Deployment Considerations:
- The 1 TB NVMe SSD is appropriate for OS, applications, and working data — not as the primary destination for sustained high-volume video recording. Plan external or network-attached storage for any retention requirement beyond short-term buffering.
- At 11.2 lbs and 6.80 x 19.60 x 11.70 inches, the unit fits on a shelf or VESA-mounts behind a display, but confirm your mounting hardware and cable management before deployment — the depth (19.60 in) is larger than typical NUC-class devices despite the 'mini' branding.
This configuration is a strong fit for a security operations center workstation or a compact analytics edge node where GPU-accelerated inference, a multi-core CPU, and a small physical footprint need to coexist in the same deployment — and where a full tower workstation simply isn't an option.
