Hardware Compatibility Matrix

Real-world edge performance benchmarks using the Quantized Labs Resonance Engine. No cloud GPUs required.

Hardware / ChipsetSmolLM-135MUltra-Tiny • 150MB RAMGemma-2-2BIoT / Mobile • 600MB RAMLlama-3.1-8BReasoning • 2.5GB RAMQwen-2.5-14BHeavy-Duty • 4GB RAM
AWS Graviton3 (c7g.2xlarge)
Cloud Server / ARM Neoverse-V1
2,800 T/s620 T/s145 T/s72 T/s
Apple M3 Max
MacBook Pro / Desktop ARM
2,400 T/s450 T/s110 T/s65 T/s
AMD Ryzen 9 7950X
Desktop x86 / AVX-512
2,100 T/s380 T/s98 T/s54 T/s
Intel Core Ultra 7 155H
Meteor Lake / Modern AI Laptop
1,950 T/s360 T/s92 T/s50 T/s
Intel Core i7-13700K
Desktop x86 / AVX2
1,800 T/s320 T/s85 T/s45 T/s
Snapdragon 8 Gen 3
Galaxy S24 / Flagship Android
950 T/s160 T/s42 T/s24 T/s
MediaTek Dimensity 9300
Flagship Android
880 T/s150 T/s38 T/s20 T/s
Apple A17 Pro
iPhone 15 Pro
800 T/s90 T/s28 T/s16 T/s
NVIDIA Jetson Orin Nano
Edge Robotics
700 T/s120 T/s30 T/s15 T/s
Google Pixel 8 Pro
Tensor G3
650 T/s115 T/s28 T/s14 T/s
Steam Deck
AMD Custom APU
600 T/s110 T/s25 T/s12 T/s
Snapdragon 6 Gen 3
Mid-Range Android
350 T/s45 T/s12 T/sOOM
Rockchip RK3588
SBC / Orange Pi 5
220 T/s35 T/s8 T/sOOM
Apple Watch Ultra 2
S9 SiP
120 T/sOOMOOMOOM
Real-Time Inference (>30 T/s)
Usable Inference (10-30 T/s)
Out of Memory (OOM)

Zero Thermal Throttling

Standard edge runtimes (like CoreML and NNAPI) burn through battery and overheat the device within minutes, causing massive frame-rate drops. The Quantized Labs's Symbiotic Runtime prevents thermal runaway entirely.

45°C40°C35°C30°C
44.2°C
Standard CoreML
Throttles at 3 mins
33.8°C
Quantized Labs
60 mins continuous

Battery Consumption (mAh / 10k Tokens)

Thermals are great, but for mobile and wearables, battery life is the ultimate constraint. The Quantized Labs uses raw integer execution units, consuming up to 80% less power than float-based Neural Engine frameworks.

100 mAh75 mAh50 mAh25 mAh
88 mAh
Standard CoreML
iPhone 15 Pro
17 mAh
Quantized Labs
iPhone 15 Pro

Real-World Performance Simulator

Watch how Quantized Labs's Symbiotic Engine radically outperforms standard float-based pipelines on identical hardware (simulated iPhone 15 Pro).

CoreML (Float16)
Explain Quantum Entanglement
Thermal Throttling Detected12 T/s
Quantized Labs (Int2)
Explain Quantum Entanglement
Nominal Temp (33°C)38 T/s

Memory Bandwidth Bottlenecks

Edge AI is bound by memory bandwidth, not just compute. The following matrix shows the theoretical minimum bandwidth required to achieve 10 Tokens/Sec for each compressed architecture.

ArchitectureMin Bandwidth for 10 T/sRecommended RAM TypeCompatible Hardware
SmolLM-135M1.5 GB/sDDR4 / LPDDR4IoT / Raspberry Pi 4
Gemma-2-2B6.0 GB/sLPDDR4xMid-Range Phones
Llama-3.1-8B25.0 GB/sLPDDR5 (3200MHz+)Flagship Phones / Laptops
Qwen-2.5-14B40.0 GB/sLPDDR5xApple M-Series / High-End Laptops

Sustained Load Profiling

Need to know exactly how Quantized Labs performs on your proprietary hardware? Submit your silicon architecture details and our engineering team will provide a comprehensive Time-to-Throttling analysis.