HP Inc BM6M6UT#ABA HP SmartBuy Z2 Mini G1I Workstation Desktop PC Intel Core Ultra 5 245K

HP Inc BM6M6UT#ABA Z2 Mini G1i Compact Workstation Desktop

Overview

The HP Z2 Mini G1i is a sub-6-pound workstation desktop built around Intel's 15th-generation Core Ultra 5 245K processor — the right call when you need certified workstation performance in a form factor that fits behind a monitor arm or inside a rack shelf. The BM6M6UT#ABA ships configured with 32GB of RAM, a 1TB M.2 PCIe NVMe SSD, and NVIDIA Quadro RTX A40 graphics, paired with Intel AI Boost delivering up to 13 TOPS of on-device AI inferencing. For compact workstations handling VMS client decode, forensic review, or AI-assisted video analytics at the edge, this configuration threads the needle between horsepower and physical footprint.

Key Features

• Intel Core Ultra 5 245K (15th Gen, 14 cores, up to 5.20 GHz): The 14-core architecture separates performance and efficiency cores, so background tasks — OS updates, indexing, antivirus scans — don't compete with your active VMS decode or forensic export workloads. Boost to 5.20 GHz matters for single-threaded codec operations that don't parallelize cleanly.
• Intel AI Boost NPU (up to 13 TOPS): The dedicated neural processing unit offloads AI inference tasks — object detection pre-screening, noise reduction, metadata tagging — from the CPU and GPU, keeping both available for rendering and decode. At 13 TOPS, this is meaningful for lightweight edge AI pipelines; heavy model training still belongs on a GPU cluster.
• NVIDIA Quadro RTX A40 discrete graphics: The Quadro RTX A40 is a professional-class GPU with certified driver stacks for video management software and CAD/visualization workloads. Where consumer GeForce cards introduce driver instability under 24/7 VMS loads, Quadro-class hardware is engineered and validated for that operating profile. This is the spec that separates the Z2 Mini G1i from a business desktop.
• 32GB RAM: Running a multi-pane VMS client with simultaneous live views, playback, and analytics overlays is memory-hungry. 32GB gives you enough headroom to run the VMS client, a browser-based management console, and standard OS services concurrently without paging to disk. If your deployment grows to 64+ camera streams on a single client, check whether this SKU supports memory expansion before committing.
• 1TB M.2 PCIe NVMe 2280 TLC 4X4 SSD: NVMe 4x4 lanes deliver sequential read speeds that make opening large video archives, loading forensic clips, or exporting evidence packages feel immediate rather than sluggish. The TLC NAND is the right trade-off here — higher endurance than QLC for a workstation OS drive that gets hammered with random writes from running applications.
• Z2 Mini form factor (5.29 lb, 20.00 x 7.00 x 12.00 in): At under 5.5 pounds, the Z2 Mini mounts behind monitors with VESA adapters or slides into tight equipment cabinet shelves. For NVR and VMS client deployments where rack space is shared with network gear, the footprint matters as much as the specs. Verify your cabinet depth against the 12-inch height dimension before ordering.

Integration and Compatibility

The BM6M6UT#ABA runs on Intel's 15th-generation platform with the Core Ultra 5 245K, which carries the updated Intel Graphics architecture alongside the discrete Quadro RTX A40. For integrators deploying this alongside IP camera systems, confirm your VMS vendor's GPU decode acceleration support for Quadro RTX-series cards — most major platforms (Milestone, Genetec, Axis Camera Station) have certified decode paths for Quadro hardware, but check the specific version matrix for your release. The Intel AI Boost NPU is accessible via OpenVINO and Windows AI APIs; not all VMS platforms expose NPU acceleration today, so treat the 13 TOPS as forward capacity for upcoming platform releases rather than a day-one feature for most deployments. For GPU-accelerated analytics or multi-monitor forensic review stations, pair this unit with a commercial-grade display that matches the Quadro RTX A40's output capabilities.

Frequently Asked Questions

Q: What CPU does the BM6M6UT#ABA ship with, and how many cores does it have?

A: The BM6M6UT#ABA ships with an Intel Core Ultra 5 245K, a 15th-generation processor with 14 cores running at a base of 3.60 GHz and boosting to 5.20 GHz.

Q: Does the HP Z2 Mini G1i BM6M6UT#ABA include dedicated graphics?

A: Yes. This configuration includes an NVIDIA Quadro RTX A40, a professional-class GPU with certified driver support for workstation and VMS workloads, alongside the integrated Intel Graphics included with the Core Ultra 5 245K.

Q: What is the Intel AI Boost feature on the BM6M6UT#ABA?

A: Intel AI Boost refers to the dedicated NPU (Neural Processing Unit) built into the Core Ultra 5 245K processor. It delivers up to 13 TOPS of AI inference performance on-device, offloading lightweight AI tasks from the CPU and GPU.

Q: How much RAM and storage does the BM6M6UT#ABA ship with?

A: The unit ships with 32GB of RAM and a 1TB M.2 PCIe NVMe 2280 TLC 4X4 SSD as the primary storage drive.

Q: What are the physical dimensions and weight of the HP Z2 Mini G1i?

A: The unit measures 20.00 x 7.00 x 12.00 inches (L x W x H) and weighs 5.29 lb, making it suitable for VESA rear-mount or compact cabinet shelf installations.

Q: Is the BM6M6UT#ABA a SmartBuy configuration?

A: Yes. The BM6M6UT#ABA carries the HP SmartBuy designation, which indicates a pre-configured, channel-stocked specification — typically faster to procure than custom-configured build-to-order units.

Ted Perry

The BM6M6UT#ABA is one of the more interesting compact workstation configurations I've seen come through the HP Z-series line — the combination of a 14-core Core Ultra 5 245K with a discrete Quadro RTX A40 in a 5.29-pound chassis is genuinely unusual. Most integrators I work with are choosing between a full tower workstation with real GPU headroom or a mini PC with integrated graphics and constant thermal throttling. The Z2 Mini G1i carves out a third path.

Technical Highlights:

• 14-core Core Ultra 5 245K at 5.20 GHz boost: The performance/efficiency core split means you can pin VMS decode to P-cores and let background services run on E-cores — most modern VMS schedulers handle this automatically. The 5.20 GHz boost ceiling helps on codec operations that don't scale across threads.
• NVIDIA Quadro RTX A40 GPU: Professional GPU driver certification is the real value here — not raw benchmark scores. On 24/7 VMS client workloads, Quadro drivers are validated against the operating profile in ways consumer GeForce stacks simply aren't. This matters when you're running GPU-accelerated decode across 32+ simultaneous streams.
• Intel AI Boost NPU (13 TOPS): This is forward-looking capacity on most current VMS platforms. OpenVINO-integrated analytics pipelines can use it today, but if your VMS doesn't yet expose NPU acceleration, budget the 13 TOPS as runway for the next platform release cycle rather than an immediate productivity gain.

Deployment Considerations:

• The 20.00 x 7.00 x 12.00 in footprint fits tight spaces, but that 12-inch height dimension catches integrators off guard when they're planning cabinet shelves — measure your shelf depth before the unit ships.
• The 1TB NVMe drive is the OS and application drive, not a video archive. Don't spec this unit expecting local long-term storage; pair it with network-attached storage or a dedicated NVR for retention beyond local cache.

This configuration is the right pick for a VMS client workstation in a physical security operations center or a forensic review station where an analyst needs GPU-accelerated multi-stream playback and AI-assisted search in a space-constrained console room.