Early Snapdragon 8 Gen 4 Benchmarks Reveal 25% CPU Uplift: Qualcomm’s Oryon Cores Challenge Apple’s Bionic
Leaked Geekbench 6 results for Qualcomm’s upcoming Snapdragon 8 Gen 4 chipset indicate a significant performance leap, with early tests showing a ~25% single/multi-core CPU improvement over the current Gen 3. The gains are attributed to Qualcomm’s custom Oryon CPU cores – marking the company’s most aggressive architectural shift since abandoning ARM stock designs.
Key Performance Data:
| Metric | Snapdragon 8 Gen 4 | Snapdragon 8 Gen 3 | Gain |
|---|---|---|---|
| Single-Core | 2,450 | 1,960 | +25% |
| Multi-Core | 8,100 | 6,480 | +25% |
| AI Inference | ~90 TOPS (est.) | 60 TOPS | +50% |
Source: Geekbench 6.2 beta tests (engineering sample)
Architectural Breakthroughs:
Oryon CPU Cores
First fully custom Qualcomm CPU since Nuvia acquisition
8-core configuration: 1× Prime (3.8GHz) + 3× Gold (3.4GHz) + 4× Silver (2.8GHz)
Adreno 850 GPU
35% faster ray tracing vs. Gen 3
Hardware-accelerated Unreal Engine 6 support
Hexagon NPU
Target: 100+ TOPS for on-device LLMs (e.g., 70B parameter models)
Target Applications & Timeline:
Flagship Devices: Xiaomi 15 Ultra, OnePlus 13, Galaxy S25 (US/China variants)
AI Focus: Real-time video generation, multi-modal assistants (text/image/voice)
Gaming: 4K/60fps mobile ray tracing, console-grade physics
Launch: October 2025 (devices shipping Q4 2025–Q1 2026)
Competitive Landscape:
| Chipset | Peak CPU Perf. | AI TOPS | Key Advantage |
|---|---|---|---|
| Snapdragon 8 Gen 4 | ~25% uplift | ~90–100 | Custom Oryon cores |
| MediaTek Dimensity 9400 | +20% (vs. Gen 3) | 80 | ARMv9.4 CPU clusters |
| Apple A18 Pro | +15% (est.) | 75+ | 3nm++ process node |
Industry Implications:
*“These benchmarks confirm Oryon cores outperform ARM’s Cortex-X5 in raw throughput. If thermals hold, Android flagships could match Apple’s single-threaded dominance for the first time since 2020.”*
— Semiconductor Analyst, TechInsights
The Snapdragon 8 Gen 4 positions Qualcomm to dominate 2025’s AI-powered Android flagships. With Apple prioritizing on-device AI in iOS 18 and MediaTek pushing aggressive core counts, the battle now hinges on real-world efficiency – a key focus for Qualcomm’s TSMC N4P node.
Strategic Notes:
Efficiency Concerns: Early tests show 15% higher power draw vs. Gen 3 under load.
AI Ecosystem: Qualcomm’s AI Stack 2.0 SDK (releasing August) will leverage the NPU boost.
Market Window: Launch timing avoids Apple’s September iPhone 16 debut.