UNIT · AMD · LAPTOP
128 GB UNIFIEDworkstationReviewed June 2026

HP ZBook Ultra G1a (Ryzen AI Max+ PRO 395)

No editorial image yet — generic vendor mark shown. Credentials in spec table below.

The flagship Strix Halo mobile workstation: a 14" laptop with 128GB LPDDR5X-8000 unified memory (up to ~96GB allocatable to the Radeon 8060S iGPU). The portable form of the 128GB unified-memory class — runs large local LLMs on the go.

Released 2025·256 GB/s memory bandwidth
▼ CHECK CURRENT PRICE· 1 retailer
HP ZBook Ultra G1a (Ryzen AI Max+ PRO 395)

Affiliate disclosure: as an Amazon Associate and partner of other retailers, we earn from qualifying purchases. The verdict on this page is our editorial opinion; affiliate links never influence what we recommend.

RUNLOCALAI SCORE
See full leaderboard →
155/ 1000
DD-tier
Estimated
Throughput
74/ 500
VRAM-fit
0/ 200
Ecosystem
130/ 200
Efficiency
17/ 100

Sub-scores sum to 221 / 1000. Headline = 221 × 0.70 (Estimated-confidence discount) = 155. This is an algorithmic performance-tier score — distinct from, and often lower than, the editorial “Our verdict” below, which weighs value and real-world fit (especially for hardware we haven’t measured yet). How scoring works →

Extrapolated from 256 GB/s bandwidth — 25.6 tok/s estimated. No measured benchmarks yet.

Plain-English: Doesn't fit modern chat models usefully — vision models won't fit.

7B chat
Doesn't fit
14B chat
Doesn't fit
32B chat
Doesn't fit
70B chat
Doesn't fit
Coding agent
Doesn't fit
Vision (≤8B VLM)
Doesn't fit
Long context (32K)
Doesn't fit
Comfortable — fits with headroom
~Tight — works, no slack
Marginal — needs aggressive quant
Doesn't fit usefully

Verdicts extrapolated from catalog VRAM + bandwidth + ecosystem flags. Hover any chip for the rationale. Want measured numbers? Submit your own run with runlocalai-bench --submit.

BLK · VERDICT

Our verdict

OP · Fredoline Eruo|VERIFIED JUN 18, 2026
7.8/10

What it does well

The HP ZBook Ultra G1a is the closest thing to a portable Mac-Studio-class local-AI machine on x86. Its Ryzen AI Max+ PRO 395 exposes up to ~96GB of a 128GB unified pool to the Radeon 8060S iGPU, so this 14" laptop runs 70B-class models locally — capacity no discrete-GPU laptop short of the 24GB 5090 Mobile can approach, and far beyond it for raw model size. For someone who needs to run big models on the move without a cloud, it's nearly unique.

Where it struggles

It's expensive ($4,000+) and shares Strix Halo's bandwidth ceiling (256 GB/s), so token speed on large models is modest — a 'fits big models portably' machine, not a fast one. As a 14" laptop it's thermally constrained versus the desktop Strix Halo boxes, so sustained inference throttles sooner. ROCm-on-Linux is where it shines; Windows GPU-offload support is more limited. No CUDA.

Bottom line

The premium pick for running 70B-class models on a genuine laptop. Worth it only if portability of big-model inference is the specific need; the Strix Halo desktops (Framework, GMKtec) give the same capability for half the price if you don't need it mobile.

BLK · OVERVIEW

Overview

The flagship Strix Halo mobile workstation: a 14" laptop with 128GB LPDDR5X-8000 unified memory (up to ~96GB allocatable to the Radeon 8060S iGPU). The portable form of the 128GB unified-memory class — runs large local LLMs on the go.

Retailers we'd check:Amazon

Some links above are affiliate links. We may earn a commission at no extra cost to you. How we make money.

BLK · SPECS

Specs

System RAM (typical)128 GB
Power draw (peak)120 W
Released2025
MSRP$3999
Backends
ROCm
Vulkan

Models that fit

Open-weight models small enough to run on HP ZBook Ultra G1a (Ryzen AI Max+ PRO 395) with usable context.

Compare alternatives

Hardware worth comparing

The closest alternatives by price, memory bandwidth, and form factor, plus a step up and down — so you can frame the buying decision against real options.

Frequently asked

Does HP ZBook Ultra G1a (Ryzen AI Max+ PRO 395) support CUDA?

No — HP ZBook Ultra G1a (Ryzen AI Max+ PRO 395) is an AMD card. Use ROCm (Linux) or the Vulkan backend in llama.cpp instead. CUDA-only tools won't work.

Where next?

Reviewed by RunLocalAI Editorial. See our editorial policy for how we research and verify hardware specifications.