Intel’s PR team is at it again. In the latest set of slides shared with the press, the company claims a performance lead of up to 48% in “browser benchmarks” such as WebXPRT with the Ryzen parts running on DC (battery power) which prioritizes battery life over performance.
Similarly, the PCMark 10 benchmark for Office 365 shows a 38% deficit upon switching to battery power while the Intel parts are largely unaffected.
After that, there a series of other useless tests such as Outlook Mail Merger, Word to PDF, Excel to Word conversions where Intel’s parts are 2 seconds faster on battery. More importantly, all these tests were part of various Intel benchmark RUGs which include a custom script (that Intel doesn’t publicly share) to test the performance of various applications. I reckon all these tests are favored by lower latency which is where Intel parts have an advantage die to their monolithic design.
When it comes to relevant content creation benchmarks, the Ryzen CPUs don’t really show a deficit upon switching to the battery, and here they are nearly twice as fast compared to the closest Intel Tiger Lake-U rivals. In our testing, however, we found that the Ryzen 9 4900U (Zephyrus G14) does show a notable drop of roughly 1,000 points on battery, but it’s still much faster than anything Intel has to offer.
Let me ask you a question? What is more important, a browser test where the deltas between all the chips is a mere 2-4 seconds between all the CPUs or rendering, where the deltas vary from 50-60% between the two rivals, saving content creators roughly half the time compared on Renoir laptops?
Intel won’t say anything about gaming performance here, as the integrated Vega parts on the Ryzen 4000 parts are quite a bit faster due to better driver optimization and the use of a well-known microarchitecture.