Skip to main content
Mini PC Lab logo
Mini PC Lab Tested. Benchmarked. Reviewed.
comparisons

MINISFORUM MS-S1 Max vs AI X1 Pro-470: Strix Halo vs Strix Point — $1,000 Gap [2026]

By Mini PC Lab Team · March 30, 2026 · Updated April 7, 2026

This article contains affiliate links. If you purchase through our links, we may earn a commission at no extra cost to you. We only recommend products we’ve thoroughly researched.

MINISFORUM MS-S1 Max vs AI X1 Pro-470: Strix Halo vs Strix Point — $1,000 Gap [2026]

Minisforum MS-S1 MAX

The Short Answer

These are completely different machines despite sharing the MINISFORUM brand. The MS-S1 Max at $2,399 is a Strix Halo workstation for running 70B+ LLMs locally, serious homelab with 10GbE, and content creation. The X1 Pro-470 at $1,359 is a Strix Point premium daily driver with eGPU capability via OCuLink.

If you need 128GB RAM for large language models or 10GbE for enterprise networking, the MS-S1 Max justifies its $1,040 premium. If you want a balanced premium mini PC with eGPU expansion and integrated PSU, the X1 Pro-470 is the smarter buy.

Side-by-Side Specs

SpecMINISFORUM MS-S1 MaxMINISFORUM AI X1 Pro-470Winner
CPURyzen AI Max+ 395 (16C/32T, 5.1GHz, Strix Halo)Ryzen AI 9 HX 470 (12C/24T, 5.2GHz, Strix Point)🏆 MS-S1 Max (more cores, Strix Halo)
GPURadeon 8060S (40 CUs, 2,560 shaders)Radeon 890M (16 CUs, 1,024 shaders)🏆 MS-S1 Max (2.5x more CUs)
RAM128GB LPDDR5X (soldered, 8-channel)32GB DDR5 SO-DIMM (upgradeable to 128GB)Contextual
StorageConfig-dependent1TB PCIe NVMe SSD🏆 X1 Pro-470 (confirmed)
Networking10GbE SFP+ + 5GbE RJ45Dual 2.5GbE (Intel)🏆 MS-S1 Max (10GbE)
WiFiWiFi 7WiFi 7Tie
OCuLinkTBDYes🏆 X1 Pro-470 (eGPU expansion)
AI TOPS126 (50+ NPU + GPU)86 (50+ NPU + GPU)🏆 MS-S1 Max
PSUTBDIntegrated (no brick)🏆 X1 Pro-470
Price~$2,399+~$1,359🏆 X1 Pro-470 ($1,040 savings)

Power Consumption

MetricMINISFORUM MS-S1 MaxMINISFORUM AI X1 Pro-470
Idle (W)~12W~9W
Load (W)~120W~96W
Annual Cost (24/7 idle)~$12.61/year~$9.46/year

Annual cost calculated at $0.12/kWh, running 24/7 at idle. Sources: ServeTheHome (Strix Halo platform) and NotebookCheck (Strix Point platform).

The MS-S1 Max draws more power — both at idle and under load — because the Strix Halo platform is fundamentally more powerful. The 128GB LPDDR5X memory and 40-CU GPU contribute to higher baseline consumption. For always-on homelab use, that’s an extra ~$3/year in electricity.

Detailed Breakdown

CPU Performance: Strix Halo vs Strix Point

The Ryzen AI Max+ 395 (Strix Halo) has 16 Zen 5 cores and 32 threads. The Ryzen AI 9 HX 470 (Strix Point) has 12 Zen 4 cores and 24 threads.

For multi-threaded workloads:

  • The 395 wins by roughly 25-30% in Cinebench R23
  • Video encoding, 3D rendering, and batch processing benefit significantly
  • Running 10+ VMs simultaneously favors the 395

For single-threaded workloads:

  • Both chips are competitive — the 470’s 5.2 GHz boost is slightly higher
  • Desktop tasks feel identical on both systems
  • Neither chip bottlenecks everyday use

For AI workloads:

  • The 395’s Strix Halo architecture is optimized for sustained AI inference
  • 126 TOPS vs 86 TOPS — the 395 wins on paper
  • In practice, both handle Copilot+ features and local LLMs well

GPU: 40 CUs vs 16 CUs

This is the biggest differentiator.

Radeon 8060S (MS-S1 Max):

  • 40 RDNA 3.5 compute units (2,560 shaders)
  • Performance between RTX 4060 and RTX 4070 laptop GPUs
  • Handles 1080p gaming at ultra settings (60+ fps in AAA titles)
  • Generates SDXL images in ~8-12 seconds
  • Can allocate up to 96GB VRAM for large LLM inference

Radeon 890M (X1 Pro-470):

  • 16 RDNA 3.5 compute units (1,024 shaders)
  • Performance comparable to entry-level discrete GPU
  • Handles 1080p gaming at medium-high settings (40-60 fps)
  • Generates SDXL images in ~20-30 seconds
  • Shares system RAM (up to 32GB without upgrade)

For gaming: The 8060S is 2-2.5x faster. If you care about 1080p ultra gaming, the MS-S1 Max is in a different league.

For Stable Diffusion: The 8060S generates images 2-3x faster. For serious AI art work, this matters.

For LLM inference: The 8060S alone doesn’t make the difference — it’s the 128GB RAM that enables 70B+ models. But the GPU does accelerate token generation.

RAM: 128GB Soldered vs 32GB Upgradeable

MS-S1 Max (128GB LPDDR5X soldered):

  • 8-channel memory architecture (~256 GB/s bandwidth)
  • Cannot be upgraded — what you buy is what you keep
  • Enables running 70B+ parameter LLMs locally (Qwen 72B, DeepSeek 70B)
  • Overkill for general use, essential for large model inference

X1 Pro-470 (32GB DDR5 SO-DIMM upgradeable):

  • Dual-channel memory architecture (~100 GB/s bandwidth)
  • Can upgrade to 64GB or 128GB later (2x 64GB sticks)
  • Handles 7B-13B models comfortably with 32GB
  • Requires RAM upgrade for 70B models (~$200-300 for 64GB kit)

For LLM users: The MS-S1 Max’s 128GB is a major advantage — you can run 70B Q4 models (~42GB RAM) with room for context and multiple models simultaneously.

For general users: 32GB is plenty. The upgradeable nature of the X1 Pro-470 means you can add RAM later if needed.

Networking: 10GbE vs Dual 2.5GbE

MS-S1 Max:

  • 10GbE SFP+ port — fiber or DAC cable connectivity
  • 5GbE RJ45 port — copper Ethernet
  • Perfect for TrueNAS with 10GbE clients, high-speed backup, or enterprise homelab
  • Overkill for typical home internet (most people have 1Gbps or less)

X1 Pro-470:

  • Dual 2.5GbE Intel NICs (i226-V)
  • Perfect for OPNsense/pfSense firewall (WAN + LAN)
  • Adequate for 2.5Gbps internet connections
  • Well-supported in Linux and Proxmox

For homelab:

  • If you have 10GbE switches, NAS, or clients — the MS-S1 Max is the only choice here
  • If you’re building a firewall/router — the X1 Pro-470’s dual 2.5GbE is ideal
  • For most home users — 2.5GbE is future-proof enough

The X1 Pro-470 has an OCuLink port — a PCIe 4.0 x4 external connection for eGPU enclosures. This means:

  • Add an RTX 4070 Ti, 4080, or even 4090 for desktop-class gaming
  • Connect external AI accelerators for ML workloads
  • Future-proof for GPU-intensive tasks

The MS-S1 Max’s OCuLink status was TBD at time of writing. Given its workstation focus, it may or may not include OCuLink — the massive 8060S iGPU reduces the need for eGPU anyway.

Verdict: If eGPU expansion matters to you, the X1 Pro-470 is the confirmed choice.

Integrated PSU

The X1 Pro-470 has an integrated power supply — no external power brick. This means:

  • Cleaner cable management
  • One less thing to lose or break
  • More professional appearance

The MS-S1 Max’s PSU configuration was TBD, but given its workstation focus and higher power draw (~120W load), it may use an external brick or a larger internal PSU.

Price and Value

At $2,399+, the MS-S1 Max costs $1,040 more than the X1 Pro-470 at $1,359. What does that premium buy you?

MS-S1 Max advantages:

  • +4 CPU cores (16C vs 12C)
  • +24 GPU CUs (40 vs 16 — 2.5x more)
  • +96GB RAM (128GB vs 32GB)
  • 10GbE SFP+ networking
  • +40 TOPS AI (126 vs 86)

X1 Pro-470 advantages:

  • OCuLink eGPU expansion
  • Integrated PSU
  • Upgradeable RAM (can reach 128GB later)
  • $1,040 savings

Value analysis:

  • If you need 128GB RAM for LLMs — the MS-S1 Max is worth it (buying 128GB RAM separately costs ~$400-500)
  • If you need 10GbE — the MS-S1 Max is the only choice in this comparison
  • If you want eGPU flexibility — the X1 Pro-470 wins
  • If you’re budget-conscious — the X1 Pro-470 saves you $1,040

Final Verdict: Which Should You Buy?

Buy the MINISFORUM MS-S1 Max if:

  • You need to run 70B+ parameter LLMs locally (this is the killer use case)
  • You have 10GbE networking infrastructure (switches, NAS, clients)
  • You do serious GPU compute work (Stable Diffusion, video editing, 3D rendering)
  • You want the most powerful mini PC MINISFORUM makes

Buy the MINISFORUM AI X1 Pro-470 if:

  • You want a premium daily driver with eGPU expansion
  • You prefer integrated PSU (no power brick)
  • You want upgradeable RAM (can add 64GB or 128GB later)
  • You’d rather save $1,040 and spend it on peripherals, storage, or networking

Our pick: For 90% of buyers, the MINISFORUM AI X1 Pro-470 at $1,359 is the smarter choice. The OCuLink port, integrated PSU, and upgradeable RAM make it versatile. The $1,040 savings can buy a 2TB NVMe, a 10GbE network card for your existing PC, and still leave money for peripherals.

The MS-S1 Max only makes sense if you specifically need 128GB RAM for large LLMs or 10GbE for enterprise networking — and if you do, you already know which one to buy.