At launch, Apple already told us that the new 12.9-inch M1 iPad Pro would be 50% faster. The latest benchmarks show that the 50% is only a meager estimate. A number of Geekbench benchmark tests show that the new M1 iPad Pro is more than half faster than the 12.9-inch iPad Pro of the fourth generation. The scores are so high that the iPad beats a top class, the 16-inch MacBook Pro with Intel i9 processor.
According to MacRumors, the tests published on the Geekbench site show a significant speed increase. The new iPad Pro achieves an average single core score of 1,712 and a multi-core score of 7,207. Comparing that to the fourth-generation 12.9-inch iPad Pro scores of 1,121 and 4,656, the M1 model comes out 53% and 56% faster in the test for single and multi-core, respectively.
M1 Ipad Pro Faster than 16-inch i9 Intel MacBook Pro
As mentioned, Apple already said during the presentation of the new iPad Pro that the M1 model would deliver 50% faster CPU performance. The figures below should therefore hardly come as a surprise. The numbers only get really impressive when you compare the scores with those of the fastest Intel MacBook Pro. This top class laptop achieves average scores of ‘only’ 1,091 in single core mode and 6,845 in multi-core. Only in multi-core is the new iPad 6% faster.
The GPU of the new iPad has also become considerably faster. The Geekbench’s Metal API scores reveal a 71% speed increase over the fourth-generation iPad Pro. That is considerably more than the 40% that Apple promised during the “Spring-Loaded” event.
Right now, the new iPad Pro looks like a huge powerhouse. It may even have too much power for an iPad. The muscles of this iPad, in combination with the currently available apps and the current version iPadOS, are still far from being used optimally. However, this could change with the arrival of iPadOS 15 and the release of more comprehensive video editing apps. You might think that the M1 iPad Pro is really a purchase aimed at the future.
Are there any reviews that run Windows 64-bit apps on Windows on ARM on Parallels and compare that to current performance? And can you also run 32-bit along that route?