Developing an AR game without evaluating the technical limitations of target devices means designing blindly. What works perfectly on HoloLens 2 may deliver 15 FPS on a Meta Quest 3 in passthrough mode, and may not run at all on a budget Android smartphone with ARCore. Analysis of AR device limitations is the first document that should appear in the project, before any architecture. Contact us to get an assessment of your project.
Why is analysis of technical limitations of AR devices necessary?
At first glance, all AR devices display virtual content over the real world. In practice, they do it in fundamentally different ways, and these differences dictate capabilities at the physics level.
Optical see-through (OST) — HoloLens 2, Magic Leap 2. Lenses are transparent, augmented reality is overlaid optically. Consequence: it is impossible to display opaque content, colors are perceived differently (dark shades are almost invisible), field of view is limited (HoloLens 2 — about 52°). For games this means: all content must be bright and contrasty, mechanics cannot rely on occlusion of real objects.
Video see-through (VST) — Meta Quest 3, PICO 4, mobile AR. A camera captures the real world and renders it together with virtual content. Advantage: full control over compositing, occlusion possible. Limitation: latency between the real world and its video image (on Quest 3 about 12-15 ms), color artifacts on object edges, passthrough quality depends on the camera.
Mobile AR (iOS ARKit, Android ARCore) — VST through the main camera. Limitations: lack of depth (no stereoscopic vision), tracking only via SLAM on a single camera, object placement accuracy ~1-2 cm, no controller-level haptic feedback. ARKit with LiDAR provides occlusion 5-10 times more accurate than ARCore without depth sensor, where the error reaches 5-10 cm. Meta Quest 3 offers 2x the field of view of HoloLens 2, enabling more immersive AR experiences.
Which key parameters of AR devices need to be analyzed?
What to know about tracking? — analysis of technical limitations
Determine what the device supports out of the box: plane detection, image tracking, object tracking, face tracking, hand tracking, spatial anchors, scene understanding (mesh reconstruction). ARKit on iPhone 12+ supports LiDAR scanning, giving a mesh of the real world and allowing virtual objects to "hide" behind real surfaces via occlusion. ARCore lacks a depth sensor on most devices and uses monocular depth estimation — significantly less accurate.
Compute budget
AR requires parallel work: camera + SLAM + rendering + game logic. On Snapdragon 888 this works with proper optimization. On Snapdragon 680 (budget segment), ARCore's SLAM algorithm takes 20-30% CPU, leaving less for game logic. Analysis should include specific CPU/GPU budgets for each platform.
Consider an example: on a project with dense geometry and PBR shaders on HoloLens 2, we measured 200k draw calls at 60 FPS — the limit. Increasing triangles beyond 1 million caused drops to 45 FPS. On Quest 3, a similar scene maintained 90 FPS at 300k draw calls thanks to a more powerful GPU. If you plan complex physics or simulations, allocate a separate core for them — otherwise stuttering will occur.
Memory constraints
Textures for AR games are often not compressed as aggressively as in regular games because they overlay the real world and compression artifacts are more noticeable. On devices with 2-3 GB RAM this creates pressure.
Display characteristics
FOV, refresh rate, pixel density — critical for OST devices. HoloLens 2 has ~60 Hz refresh rate and limited FOV, which affects what effects and animations are acceptable. Quest 3 with 110° FOV and 90 Hz offer broader possibilities.
How does the analysis of limitations save project budget?
Timely detection of incompatibilities at the specification stage prevents costly rework. As our technical director notes: Identifying incompatibilities at the specification stage prevents rework that is 10 times more expensive. If your project targets VST devices, you won't waste resources on complex visual effects for OST lenses. Savings at the prototyping stage — up to 40% of time. For a typical project, this translates to savings of $5,000-$15,000 in avoided rework costs. The cost of analysis ranges from $700 to $2,100 per device, a small fraction of the total budget. Get a consultation for your project — we'll assess platform risks and optimization opportunities.
What's included in the work?
The analysis delivers:
- Detailed documentation with benchmark results and capability matrices
- Access to raw data and test scenes for internal verification
- Training session for your team on device-specific optimizations
- Priority support for 2 weeks after delivery for clarification
How we conduct analysis of AR device limitations?
- Collect specifications — study official documentation and performance profiles of target devices.
- Benchmark tests — run test scenes on real hardware, measure FPS, batching, draw calls, memory usage.
- Capability matrix — compile a table per device (tracking, rendering, memory) with noted limitations.
- Form recommendations — determine minimum viable platform and optimization priorities.
Comparison table of popular AR devices
| Device | FOV | RAM | Tracking type | Features |
|---|---|---|---|---|
| HoloLens 2 | 52° | 4 GB | SLAM + IMU | OST, hand tracking, eye tracking |
| Meta Quest 3 | 110° | 8 GB | SLAM + depth | VST, hand tracking, controller |
| iPhone 12+ (LiDAR) | – | 4-6 GB | SLAM + LiDAR | VST, mesh reconstruction |
What is included in the final document?
- Capability matrix per device (hand tracking, plane detection, occlusion, spatial anchors) with statuses: supported/limited/not supported.
- List of constraints for architecture tied to SDK dependencies.
- Definition of minimum viable platform and premium experience.
- Documentation with benchmark results.
- Checklist of technical parameters: CPU/GPU performance, memory bandwidth, tracking latency (motion-to-photon), support for spatial anchors and their stability, compatibility with hand tracking SDKs (ARKit, MRTK, Oculus Interaction), rendering mode limitations (URP/HDRP).
- Each deliverable includes raw data, graphs, and actionable recommendations.
Our analysis covers key AR device constraints such as HoloLens 2 limitations, Meta Quest 3 AR capabilities, iPhone ARKit analysis, and ARCore constraints.
Estimated timelines
| Analysis scope | Time |
|---|---|
| Single device / platform | 3-5 days |
| Comparative analysis of 3-5 devices | 1-2 weeks |
| Full analysis + architecture recommendations | 2-3 weeks |
With over 5 years of AR development experience and a proven track record of 20+ projects, our analysis is trusted by industry leaders. We guarantee that our recommendations will reduce your technical risk. Order an analysis of your project — we'll evaluate platform risks and optimization opportunities.





