A new independent benchmark study published by Windows Central on October 1, 2026 is forcing a hard question onto Intel’s roadmap: does the Core Ultra brand still have a performance edge over Qualcomm’s second generation of Arm-based laptop silicon, or has that edge quietly disappeared. The study, run by testing firm Signal65 and commissioned by Qualcomm, pitted a standard 18-core Snapdragon X2 Elite against Intel’s Core Ultra X7 358H (the Panther Lake chip Intel positions as its premium thin-and-light part) and AMD’s Ryzen AI 7 445. The chips were tested inside three real retail 14-inch laptops, not reference boards, which is where this story gets interesting for anyone shopping this fall.
The headline numbers are blunt. Qualcomm’s mainstream chip beat Intel’s flagship thin-and-light part on multi-core and single-core CPU throughput, on two separate AI benchmarks, on most battery tests, and on price. Intel held onto one category cleanly: graphics. The result matters well beyond one laptop comparison. It is the second Signal65 report in two months to show Qualcomm’s Oryon-based silicon outperforming x86 rivals, and this time the win extends from a single halo SKU down into the laptops people actually buy.
What Signal65 actually tested
Signal65 is an independent benchmarking firm that Qualcomm has used twice in recent months to generate head-to-head performance data against Intel and AMD. The first report, covered by PCMag and by Windows Central the month before, compared Qualcomm’s flagship Snapdragon X2 Elite Extreme chip against premium 16-inch Intel and AMD laptops and found Qualcomm ahead across most workloads. That earlier win was easy to dismiss as a halo-product story, since the Extreme chip ships in essentially one laptop.
This second report targets a different, more consequential segment: mainstream 14-inch premium laptops, the size and price tier where most Windows buyers actually shop. Signal65 tested three specific retail machines rather than engineering samples:
- Lenovo Slim 7x running the 18-core Snapdragon X2 Elite X2E-88-100, priced at $1,999
- Dell XPS 14 running Intel’s Core Ultra X7 358H (Panther Lake), priced at $2,749
- Lenovo Yoga 7a 14 running AMD’s Ryzen AI 7 445
Daniel Rubino, who wrote up the findings for Windows Central, framed the shift in scope as the real story: Qualcomm’s Extreme chip proving a point is one thing, but a standard 18-core part beating Intel’s best thin-and-light silicon in laptops people can actually afford is a different and more durable claim. One methodological detail stands out across the whole report: the Snapdragon system reached its top-tier scores without needing a higher power profile, while both x86 systems needed to be switched into performance modes to compete.
CPU performance: Geekbench 7 and Cinebench 2026
Raw CPU throughput is where Intel’s Core Ultra lineup has traditionally defended itself against Arm challengers, and it is also where the gap was widest in this report. In Geekbench 7, the Snapdragon-powered Slim 7x posted a multi-core score of 24,016 against 18,360 for the Intel-powered XPS 14, a 31% lead for Qualcomm. Single-core results told the same story: 3,237 for Snapdragon versus 2,602 for Intel, a 24% advantage.
That single-core gap is the number worth sitting with. Core-count advantages explain multi-threaded wins easily enough, but single-core performance is architecture, not thread count. Signal65’s data points to real gains from Qualcomm’s second-generation Oryon CPU core design, not just a wider die. Intel’s Core Ultra X7 358H also carried a meaningful disadvantage in Cinebench 2026 rendering tests, where Qualcomm maintained what Windows Central described as a substantial lead across both multi-thread and single-thread runs.
Productivity workloads, measured through UL Procyon’s Office test suite covering Word, Excel, Outlook, PowerPoint, Edge, and Teams, again favored Qualcomm, with a 10% lead over Intel in Best Performance mode and a smaller advantage in the default Balanced profile. For a buyer whose day is spent in email and documents rather than synthetic benchmarks, that 10% gap in the exact software stack they use daily may be more persuasive than any Geekbench chart.
Table 1: Signal65 benchmark results, Snapdragon X2 Elite vs. Intel Core Ultra X7 358H
| Benchmark | Snapdragon X2E-88-100 (Lenovo Slim 7x) | Intel Core Ultra X7 358H (Dell XPS 14) | Qualcomm lead |
|---|---|---|---|
| Geekbench 7 multi-core | 24,016 | 18,360 | 31% |
| Geekbench 7 single-core | 3,237 | 2,602 | 24% |
| UL Procyon AI Computer Vision | 4,360 | 2,302 | ~89% |
| AI CV 2.0 (WinML) | Leading | Trailing | ~31% |
| UL Procyon Office (Best Performance) | Leading | Trailing | 10% |
| Local video playback battery | 16.1 hours | 14.4 hours | 1.7 hours longer |
| Chrome browsing rundown battery | 11.2 hours | 8.6 hours | 2.6 hours longer |
| Procyon Office sustained battery | 14.3 hours | 17.4 hours | Intel wins by 3.1 hours |
| 3DMark Steel Nomad Light (GPU) | Adreno X2-90, trailing | Arc B390, leading | Intel wins by ~23% |
| Retail price | $1,999 | $2,749 | $750 cheaper |
Source: Signal65 report, commissioned by Qualcomm, as reported by Windows Central, October 1, 2026.
AI acceleration is Qualcomm’s clearest advantage
Qualcomm has spent two years marketing Snapdragon X-series chips as AI-first hardware, and the second-generation numbers back that positioning more than the first generation did. In UL Procyon’s AI Computer Vision benchmark, the Snapdragon system scored 4,360 against Intel’s 2,302, roughly double Intel’s result. More telling is the AI CV 2.0 WinML test, which routes inference through Microsoft’s own Windows ML framework instead of a vendor-specific software path. Qualcomm still held a 31% lead there. A benchmark that favors proprietary optimization tricks is useful marketing; a benchmark that preserves an advantage through a shared Windows API is a stronger signal about how AI features will actually perform for ordinary users running Copilot+ workloads, image generation, or local inference inside everyday apps.
This AI lead is also the backbone of Microsoft’s own hardware strategy. The company’s refreshed Surface Pro and Surface Laptop lineup shipped with Snapdragon X2 chips and an 80 TOPS NPU target starting at $1,149, and Microsoft has now gone roughly five years without shipping a single Surface device built on AMD silicon. Rubino’s piece makes the same point explicitly: Microsoft’s continued avoidance of AMD in its own flagship hardware line looks less like a supply issue and more like a deliberate signal about where the company sees the best combination of AI performance and battery efficiency sitting right now.
Battery life: a win with an asterisk
Battery performance has been Qualcomm’s calling card since the first Snapdragon X launched in 2024, and the pattern held in this report. Across six separate benchmarks run on battery power, the Snapdragon system retained an average of 97% of its plugged-in performance in Balanced mode. Intel retained 95%. AMD’s Ryzen AI 7 445 system, by contrast, retained only 74%, a gap large enough to reframe the entire AMD conversation in this category.
Direct runtime numbers favored Qualcomm in two of three tests: 16.1 hours of local video playback versus 14.4 hours for the Intel system, and 11.2 hours of Chrome browsing versus 8.6 hours for Intel. But Intel actually won the sustained UL Procyon Office battery test, posting 17.4 hours against Qualcomm’s 14.3 hours. Signal65’s own explanation is that the Snapdragon machine completed more productivity work during that test window, muddying a simple hours-won comparison, but it is still one of Intel’s cleanest victories in the entire report and worth taking seriously rather than waving away.
There is also a confound Rubino flags directly: the two laptops in this test do not have matched displays. The Dell XPS 14 runs a 2,880 x 1,800 OLED panel at up to 120Hz with variable refresh rate, while the Lenovo Slim 7x runs a lower-resolution 1,920 x 1,200 OLED panel at a fixed 60Hz. Display output is consistently the single largest power draw in a modern laptop, so some of Qualcomm’s battery advantage reflects a less demanding screen rather than a purely silicon-level efficiency win. That does not erase the result, but it does mean the comparison is laptop-versus-laptop as much as chip-versus-chip.
Where Intel still wins: graphics
Intel’s cleanest victory in the report has nothing to do with CPU cores. The Core Ultra X7 358H‘s integrated Arc B390 graphics beat Qualcomm’s Adreno X2-90 GPU by roughly 23% in 3DMark’s Steel Nomad Light test, across both DirectX 12 and Vulkan test paths. That is a meaningful result, because graphics has been Intel’s biggest internal investment priority over the past two generations. The same Arc B390 silicon also powers the MSI Claw 8 EX AI+ handheld, where it has shown a clear lead over AMD’s competing handheld graphics.
What this means in practice is narrower than Intel’s marketing might suggest. The GPU win only shows up when the Intel system is plugged in and running in the right performance mode, while Qualcomm’s CPU and AI wins hold regardless of power state or profile selection. For gaming, GPU-accelerated creative work, or heavy rendering, Intel’s integrated graphics edge is real and reproducible. For everyday productivity, browsing, video calls, and AI-assisted work, that edge barely matters, and Qualcomm’s chip wins on nearly every other axis that affects day-to-day use.
Price-performance: the number Qualcomm wants you to remember
Strip away the individual benchmark charts and one figure does the most work in this report: Signal65 calculated that the Snapdragon-powered Lenovo Slim 7x delivered roughly 2.15 times Intel’s value when performance was normalized against price across the tested CPU, rendering, and AI benchmarks. At retail, that comparison plays out as a $1,999 laptop beating most benchmarks run against a $2,749 laptop, a $750 gap that is hard to justify on CPU and AI grounds alone.
Dell’s XPS 14 is not a bad laptop being unfairly compared. It remains one of the most highly regarded premium Windows notebooks on the market, with build quality, display, and industrial design that reviewers consistently rate above most competitors, Snapdragon-powered machines included. Some buyers will reasonably pay the $750 premium for that OLED 120Hz panel, the integrated graphics headroom, and Dell’s build quality. Others will look at the same chart and conclude Intel’s remaining advantages are not worth the price difference. Both conclusions are defensible; the point is that Qualcomm has now built a genuine argument on price that it did not have as cleanly a year ago.
Table 2: Qualcomm’s Signal65 report scorecard, by category
| Category | Winner | Margin |
|---|---|---|
| Multi-core CPU (Geekbench 7) | Snapdragon X2 Elite | 31% |
| Single-core CPU (Geekbench 7) | Snapdragon X2 Elite | 24% |
| AI inference (Procyon AI CV) | Snapdragon X2 Elite | ~89% |
| AI inference (WinML CV 2.0) | Snapdragon X2 Elite | ~31% |
| Office productivity (Procyon) | Snapdragon X2 Elite | 10% |
| Video playback battery | Snapdragon X2 Elite | 1.7 hours |
| Web browsing battery | Snapdragon X2 Elite | 2.6 hours |
| Sustained office battery | Core Ultra X7 358H | 3.1 hours |
| Integrated graphics (3DMark) | Core Ultra X7 358H | ~23% |
| Price-performance ratio | Snapdragon X2 Elite | ~2.15x |
Historical context: how we got here
It is worth remembering how recently this comparison would have been laughable. Early Windows-on-Arm laptops, going back to the original Snapdragon 835-based machines in 2017, struggled with app compatibility, sluggish emulation, and CPU performance well behind even mid-range Intel chips. The first Snapdragon X Elite, launched in 2024, was the point where Qualcomm first posted competitive CPU numbers, but it still leaned heavily on battery life and AI marketing to make its case, because raw performance parity with Intel was not fully there yet.
This report represents a second inflection point. A year ago, beating Intel required Qualcomm’s most expensive, hardest-to-find chip in one laptop. Now a standard, mainstream 18-core Snapdragon X2 Elite is doing it in a $1,999 retail machine against Intel’s current flagship thin-and-light part. That shift from flagship-only to mainstream-capable is the difference between a proof of concept and a sustainable competitive position, and it is the central argument Rubino makes in the Windows Central piece: flagship chips make headlines, but mainstream chips make markets.
Competitive comparison: how the three platforms actually stack up
AMD is the platform this report is quietly hardest on. The Ryzen AI 7 445 in the Lenovo Yoga 7a 14 retained only 74% of its plugged-in performance on battery, a full 21 points behind Intel and 23 points behind Qualcomm. Combined with the broader pattern already visible across the Windows laptop market, where Microsoft has not shipped a Surface device with an AMD chip in roughly five years, this report adds more evidence that AMD’s mobile silicon is losing ground on both ends: it is not winning the efficiency argument Qualcomm owns, and it is not winning the ecosystem or OEM-design-win argument Intel still holds through sheer scale.
Intel’s position is more nuanced than “losing.” The Panther Lake generation that includes the Core Ultra X7 358H has already been pushing OLED displays and higher price points across its laptop lineup, and this report’s one clean Intel win, integrated graphics, is in a category Intel has specifically targeted for investment through the Arc brand. Intel is not trying to win every category; it is defending the categories where it believes it has defensible, durable advantages, and graphics is one of them.
Qualcomm, for its part, is clearly playing a longer game than any single benchmark chart. Its mobile chip business continues shipping aggressively updated silicon, including the Snapdragon 8 Elite Gen 6 announced for phones on a 2nm process, and the company’s Oryon CPU architecture, originally developed in-house, now underpins both its phone and laptop silicon lines. That shared architecture lets Qualcomm iterate CPU design once and deploy the gains across two very different product categories, a structural advantage neither Intel nor AMD can currently replicate as cleanly, since both rely on separate design teams and node strategies for mobile versus laptop-class silicon.
Why a commissioned study still matters
The most important caveat in this entire story is one Rubino raises directly: Signal65 was paid by Qualcomm to run this test. That does not make the numbers false, but it does mean the methodology, laptop selection, and power-mode configurations were chosen by a firm with a commercial relationship to the company whose chip won most categories. Signal65 is described as an independent testing organization rather than an in-house Qualcomm lab, and the benchmarks used (Geekbench 7, Cinebench 2026, UL Procyon, 3DMark) are industry-standard tools rather than custom Qualcomm software, which adds some credibility. Readers and IT buyers evaluating this data for procurement decisions should still treat it the way they would treat any vendor-commissioned study: useful directional evidence, not a substitute for independent third-party testing.
It also matters that OEMs retain real control over how a chip performs on battery. Rubino notes that companies like Lenovo can tune AC-versus-DC performance profiles differently depending on what balance of battery life and raw speed they want a given SKU to project, meaning two laptops with the same chip can post different battery results purely due to firmware tuning choices, not silicon differences. That is a reminder that this report compares shipped retail products, not isolated chips, and buyers should weigh the whole package rather than a single spec sheet number.
Market impact: what this means for Windows OEMs
For PC makers, the practical consequence is pricing pressure. If a mainstream Snapdragon X2 Elite laptop can beat a premium Intel Panther Lake laptop on CPU, AI, and most battery metrics while undercutting it by $750, Intel’s remaining margin for charging a premium on thin-and-light flagships narrows considerably. That pressure will likely show up first in Dell, HP, and Lenovo product lines that currently ship both Intel and Snapdragon configurations of the same chassis, where buyers can make an apples-to-apples price comparison within a single product family rather than across brands.
It also strengthens Microsoft’s own hardware bet. Microsoft has staked its current Surface lineup heavily on Snapdragon silicon, and reports like this one give the company independent-feeling validation (even with the commissioning caveat) for a strategy that, until recently, looked like a gamble on an unproven Arm ecosystem. Software compatibility, long the biggest knock against Windows-on-Arm, continues to matter less each cycle as more applications ship native Arm64 builds or run acceptably well under emulation.
Five predictions for the Snapdragon X2 vs. Intel fight
1. Independent, non-commissioned reviews will narrow the gap somewhat. Signal65’s numbers were generated under conditions Qualcomm had some influence over through laptop and test selection. Expect outlets like Notebookcheck and Tom’s Hardware to publish their own head-to-head numbers over the next few months, and expect those results to show real but somewhat smaller Snapdragon advantages once methodology is fully independent.
2. Intel will lean harder into graphics marketing. With CPU and AI parity increasingly contested, expect Intel’s Panther Lake follow-up messaging to foreground Arc B390 and future Arc generations as the clearest, most defensible differentiator against Arm competitors in thin laptops.
3. AMD faces the most pressure to respond. A 74% battery-retention score, trailing both rivals by double digits, is a problem AMD cannot market around. Expect AMD’s next mobile Ryzen AI refresh to prioritize efficiency tuning over raw clock speed gains.
4. OEM pricing on Snapdragon configurations will stay aggressive. With Signal65’s data showing a roughly 2.15x price-performance edge, expect Lenovo, Dell, and HP to keep Snapdragon SKUs priced meaningfully below comparable Intel configurations through at least the next product cycle, to keep driving adoption in a segment Windows-on-Arm is still working to normalize.
5. Expect a third Signal65-style report focused on AMD specifically. Given the pattern of two Qualcomm-commissioned reports targeting different Intel segments (flagship, then mainstream), a dedicated Snapdragon-versus-AMD report, isolating that comparison rather than treating AMD as a secondary data point, seems like a logical next step for Qualcomm’s marketing cadence.
What buyers should actually do with this data
If a buyer’s workload is CPU-bound productivity, AI-assisted writing or image work, and long battery life away from a charger, this report gives a real, benchmark-backed reason to consider a Snapdragon X2 Elite laptop over a comparably priced Intel Panther Lake machine, with the understanding that app compatibility should still be checked for any specialized, non-mainstream software before buying. If the workload leans on integrated graphics for light gaming, GPU-accelerated creative tools, or anything that benefits from DirectX 12 or Vulkan performance, Intel’s Arc B390 advantage is real and the Core Ultra X7 358H remains the safer choice. For AMD-based laptops specifically, the battery data here is a legitimate reason for caution until AMD’s next refresh addresses the gap directly.
Frequently Asked Questions
What is the Snapdragon X2 Elite X2E-88-100?
It is Qualcomm’s standard, mainstream 18-core laptop chip in the second-generation Snapdragon X2 family, built on the company’s second-generation Oryon CPU architecture. It is distinct from the higher-end Snapdragon X2 Elite Extreme chip, which currently ships in only one laptop.
Who ran the Snapdragon X2 vs. Intel benchmark study?
Signal65, an independent testing firm, ran the benchmarks. The study was commissioned by Qualcomm, which Windows Central disclosed directly in its coverage.
Which laptops were compared in the test?
A Lenovo Slim 7x with the Snapdragon X2 Elite X2E-88-100, a Dell XPS 14 with Intel’s Core Ultra X7 358H, and a Lenovo Yoga 7a 14 with AMD’s Ryzen AI 7 445, all in the 14-inch premium category.
Does Intel still win anything in this comparison?
Yes. Intel’s integrated Arc B390 graphics beat Qualcomm’s Adreno X2-90 GPU by roughly 23% in 3DMark Steel Nomad Light, and Intel also won the sustained UL Procyon Office battery test by about three hours.
Is a Snapdragon X2 laptop cheaper than a comparable Intel laptop?
In this test, yes. The Snapdragon-powered Lenovo Slim 7x retailed at $1,999 compared to $2,749 for the Intel-powered Dell XPS 14, a $750 difference that Signal65 says translates to roughly 2.15 times better price-performance for the Snapdragon system.
Can I trust a benchmark study commissioned by one of the chipmakers involved?
Treat it as directional rather than definitive. The benchmarks used are standard industry tools, but Qualcomm selected and paid for the testing, which can influence laptop selection and configuration choices. Independent, non-commissioned reviews from outlets like Notebookcheck are worth checking as a cross-reference.
Does this affect Microsoft’s Surface lineup?
Indirectly, yes. Microsoft’s current Surface Pro and Surface Laptop models already ship with Snapdragon X2 chips, and this report adds independent-feeling (if commissioned) performance validation for that strategy at a moment when Microsoft has gone roughly five years without shipping a Surface device built on AMD silicon.
What should I check before buying either laptop?
Confirm that any specialized software you rely on runs natively on Windows on Arm or performs acceptably under emulation before buying a Snapdragon machine, and confirm your workload actually needs the Arc B390’s graphics headroom before paying Intel’s price premium for it.
