Snapdragon X2 Elite Extreme Shines, Rivals Top Laptop Chips
PCMag reported on April 17, 2026, that Qualcomm’s new Snapdragon X2 Elite Extreme delivered strong results in real‑world benchmarks. In the outlet’s lab testing of the laptop SoC, which uses Qualcomm’s Oryon CPU cores, the X2 Elite Extreme showed competitive performance against systems running Apple’s M5, Intel’s Panther Lake Core Ultra and AMD’s Ryzen AI.
The coverage frames Qualcomm as targeting the high‑end laptop performance tier, not just efficiency. PCMag emphasizes that these are real‑world test results and that outcome will depend on system configuration and specific workloads, but the findings suggest Qualcomm is narrowing the performance gap with established x86 and Apple silicon rivals.
Key takeaways, concise summary
- Benchmarks: early tests place Snapdragon X2 Elite Extreme competitive with Intel/AMD on peak benchmarks, Apple silicon remains strong on efficiency.
- Test unit: Asus Zenbook A16, Snapdragon X2E-96-100, 48GB LPDDR5X, 3K 120Hz OLED, 2.65 lb, $1,699.99.
- Design: 18 cores, boost near 5.0 GHz.
- NPU: vendor figures show ~45 TOPS to ~80 TOPS.
What PCMag tested: Snapdragon X2 Elite Extreme on the Asus Zenbook A16
PCMag reviewed a high-end Asus Zenbook A16 configured around Qualcomm’s Snapdragon X2 Elite Extreme, listed as X2E-96-100, paired with 48GB of LPDDR5X memory. The unit they tested is a 16-inch model with a 3K, 120Hz OLED panel, and it tipped the scales at about 2.65 pounds. Important ports for creators and travelers include USB4 support and a full-size SD card reader.
The Zenbook A16 starts at $1,699.99, while the specific configuration PCMag measured reflects that review unit and its test conditions. Those reported specs, weight, and I/O details describe the tested SKU rather than every available configuration.
What’s new with the Snapdragon X2 Elite Extreme
Qualcomm’s Snapdragon X2 Elite Extreme upgrades the platform with a 3rd-gen Oryon CPU architecture in an 18-core, hybrid core design, larger caches and boost clocks reaching about 5.0 GHz. The chip also sees a substantial NPU uplift, moving from roughly 45 TOPS to about 80 TOPS, and the “Extreme” label is appended to the Snapdragon X name for the first time.
Those changes matter because more cores, higher clocks and expanded cache reduce bottlenecks in multicore workloads and improve sustained CPU performance. The hybrid core design aims to balance peak speed with efficiency, while the nearly doubled NPU throughput can materially increase on-device AI task capacity and responsiveness without relying on cloud processing.
For official details, see Qualcomm’s announcement at Qualcomm.
Benchmark context: how the X2E compares with Apple, Intel and AMD
PCMag’s lab charts show the Snapdragon X2 Elite Extreme delivering notably strong multi-core performance compared with other current premium laptop silicon. In their comparisons the X2E is set against Apple’s M5 as seen in the 14-inch MacBook Pro, Intel’s Panther Lake Core Ultra X-series and other Core Ultra SKUs, and AMD’s Ryzen AI HX and Ryzen AI Max+ parts.
Those charts suggest the X2E narrows the gap with prior Snapdragon X-family chips and competes more closely with some Core Ultra and Ryzen AI models in multi-threaded tests. PCMag also updated the story to include additional Panther Lake comparison results, an editor’s note worth checking for readers who follow Intel closely.
Keep in mind these are lab benchmarks reported by PCMag. They are useful for relative performance signals, but they do not capture broader ecosystem factors such as software optimization, battery life, thermals or platform features. For a full picture, pair the charts with real-world testing and vendor specifications.
On-Device AI in Practice: NPU Benchmarks and Common Workflows
TOPS, or tera-operations per second, describes an NPU’s peak arithmetic throughput. Announced uplifts from roughly 45 TOPS to about 80 TOPS give a sense of raw capability, but that figure is a theoretical ceiling and often overstates real application performance because memory, I/O, and model format matter.
- Local LLM inference, including chat and summarization
- Real-time transcription
- Image upscaling and on-device generation
- Interactive photo edits
Practical evaluation translates TOPS into user-facing metrics: latency, throughput, sustained power draw, and behavior under realistic load. Compatibility with frameworks such as ONNX and PyTorch Mobile, support for quantized models, and model-size limits are equally important. From a buyer perspective, privacy and offline capability matter most when benchmark numbers diverge from measured latency and energy on the actual models you plan to run.
Why it matters for buyers: Qualcomm’s strongest case yet for portable power-user laptops
According to PCMag’s testing, Qualcomm’s latest design narrows the performance gap with high-wattage x86 competitors and Apple M-series chips in demanding workloads, while fitting into a thin-and-light 16-inch chassis. That balance of portable performance and mobility is significant for power users who need higher sustained speed without carrying a bulky device.
The tested laptop weighs 2.65 pounds and starts at $1,699.99 for the reviewed configuration. Keep in mind the results reflect that specific setup, and broader software support and compatibility were not quantified in the testing, so buyers should factor those constraints into their decisions.
Where to find full charts and updated comparisons
Detailed numbers and per-system benchmark charts appear in PCMag’s full review; the comparison set has been refreshed to include additional Panther Lake comparisons and updated results.
The review was updated on April 17, 2026. Readers should consult the primary PCMag review to verify specific benchmark figures and view the full system-by-system breakdowns.
