All MacBook Air benchmarks from Microsoft

Microsoft is on a mission to take down the MacBook Air with its new line of Copilot Plus PCs. The company is so confident that it finally has Windows on the right track that it spent an entire day pitting its new Surface Laptop against the MacBook Air at its headquarters in Redmond, Washington, last month. The edge was presented with multiple real-world benchmarks and simulated tests to demonstrate that a new Qualcomm-powered Surface Laptop beats Apple’s best-selling laptop.

While I’ve previously discussed Microsoft’s confidence in beating Apple’s M3 processor, I thought it would be useful to examine all the benchmark claims and battery life estimates in detail. Microsoft discussed some of these at the Surface and Windows AI event last week, but the claims on stage weren’t always as detailed as what Microsoft employees showed me last month.

I wasn’t able to run the benchmarks myself, but the results should serve as an important data point as we get closer to the launch of these Copilot Plus PCs on June 18. It’s also important to note that unlike Apple’s MacBook Air, Microsoft’s new Surface Laptop is not fanless, allowing it to deliver more performance. Microsoft only compared its Surface Laptop to the MacBook Air M3 – not to a fan-equipped MacBook Pro.

Either way, benchmarks aren’t everything, and we’ll get a better idea of ​​real-world performance and battery life when we review the Surface Laptop next month.

Microsoft started the day of benchmarks by first measuring the raw performance between the Surface Laptop and MacBook Air M3. It showed two benchmark claims, with sustained performance when using the Cinebench 2024 multithreaded workload and top performance when using Geekbench 6’s multithreaded test.

The Surface Laptop achieved a score of 980 in Cinebench 2024 multithreaded and a score of 14,000 in Geekbench 6 multithreaded. Microsoft avoided highlighting the single-threaded scores of both benchmarks, presumably because the MacBook Air would score slightly better here.

Regardless, Microsoft claims its new Surface Laptop will beat a MacBook Air M3 in Cinebench multithreaded workloads by 50 percent. In Geekbench 6, the Surface Laptop is only 16 percent better. On stage last week, Microsoft also claimed that its range of Copilot Plus PCs will be “58 percent faster than a MacBook Air M3.”

Next, Microsoft addressed what it describes as “real performance.” The main test here was a HandBrake ToS benchmark, which measures how long it takes to encode a 4K video file. The Surface Laptop with a Snapdragon

More importantly, this was twice as fast as the Surface Laptop 5 with a 12th Gen Intel Alder Lake CPU, which took 10 minutes and 30 seconds to complete the task. A Surface Laptop 4 took even longer: 13 minutes and 32 seconds.

Microsoft’s comparisons to the MacBook Air M3 also extend to battery life. During testing, I noticed Microsoft simulating battery life while browsing the web and playing videos. Microsoft uses a script to simulate web browsing. On the 2022 Intel-based Surface Laptop 5, it took eight hours and 38 minutes to completely drain a battery; the new Surface Laptop lasted twice as long, at 16 hours and 56 minutes. That’s better than the same test on a 15-inch MacBook Air M3, which took 15 hours and 25 minutes.

Microsoft ran a similar test for video playback, with the Surface Laptop lasting more than 20 hours, while the MacBook Air M3 managed 17 hours and 45 minutes. That’s also almost eight hours more than the Surface Laptop 5, which lasted 12 hours and 30 minutes.

Microsoft claimed on stage last week that new Copilot Plus PCs with Qualcomm’s Snapdragon

The last benchmarks Microsoft showed me were for NPU performance. Microsoft claims the NPU in the Snapdragon X Elite is nearly two times faster on AI acceleration tasks than Apple’s M3 Neural Engine on the cross-platform Procyon AI Computer Vision benchmark.

The Surface Laptop scored 1,745 on the Procyon AI score, while the MacBook Air achieved 889. Qualcomm’s Snapdragon X Elite has 45 TOPS of AI acceleration performance, much more than the 18 TOPS on the M3.

Microsoft also showed that the Surface Laptop achieved 4.5x inference efficiency for fast processing of the Phi Silica model through the M3, in addition to a peak inference efficiency of 24 TOPS/watt.