Geekbench shootout: Apple’s new 15-inch MacBook Pro (mid 2018) vs. other Macs

“How does the 2018 MacBook Pro 15 inch compare to other Macs?” Rob Art Morgan writes for Bare Feats. “Will the 6-core CPU give it enough punch to compete? And what about the GPU?”

The contenders:
• 2017 iMac Pro 8-core = 3.2GHz 8-core Xeon, 32G RAM, Pro Vega 64 GPU
• 2018 MacBook Pro i9 6-core = 2.9GHz 6-Core i9, 32G RAM, Pro 560X GPU
• 2010 Mac Pro Xeon 12-core = 3.33GHz 12-core Xeon, 96G RAM, Vega Frontier GPU
• 2013 Mac Pro Xeon 8-core = 3.0GHz 8-core Xeon, 64G RAM, FirePro D700 GPUs
• 2018 MacBook Pro i7 6-core = 2.6GHz 6-core i7, 16G RAM, Pro 560X GPU
• 2013 Mac Pro Xeon 6-core = 3.5GHz 6-core Xeon, 32G RAM, FirePro D500 GPUs
• 2010 Mac Pro Xeon 6-core = 3.33GHz 6-core Xeon, 48G RAM, Radeon HD 7950 GPU
• 2017 iMac i7 4-core = 4.2GHz 4-core i7, 64G RAM, 64G RAM, Pro 580 GPU
• 2018 MacBook Pro i7 4-core = 2.7GHz 4-core i7, 16G RAM, Iris Plus 655 GPU
• 2017 MacBook Pro i7 4-core = 3.1GHz 4-core i7, 16G RAM, Pro 560 GPU

“The 2018 MacBook Pro 15 inch is a CPU crunch contender — at least when running brief benchmarks like Geekbench. Extended activity with CPU intensive pro apps have been reported to cause Thermal Throttling, nullifying Turbo Boost and even dropping below the standard clock rating. If that is a concern, you can use the Intel Power Gadget app to monitor the CPU clock frequency during your typical workflow,” Morgan writes. “The 2018 MacBook Pro’s 560X GPU, however, is not impressive. Consider budgeting for an eGPU box with a high-end GPU (like the AMD Radeon Vega series) if your GPU intensive applications flounder.”

Read more, and see the full benchmark results, in the full article here.

MacDailyNews Take:


  1. Always take synthetic benchmarks with a huge truckload of salt. Some of you may remember the old Dhrystone benchmark that was popular in the 80s and early 90s. Intel’s chips of the day skyrocketed in performance on that benchmark during that period. Then people found out that Intel specifically optimized their micro architecture for that specific benchmark. It was part reason for the push for other benchmarks I had a hand in such as the original SPECInt and SPECfp and SLOLOM benchmarks: all still synthetic, but somewhat more reliable than the Dhrystone and Whetstone benchmarks of that day.

    If performance really is a do or die kind of thing for you, load up a simulation of what you need to do onto a large thumb drive and go to an Apple store. Then spend the time to run your simulation on your platform of choice to see how it really performs for you. (Hopefully it runs for 30 minutes or more to really test the system.)

    1. So 28 years ago Intel broke major consumer laws that consumer agencies wouldn’t stand for today and somehow that’s relevant to all chips?

      Composition/Division fallacy if I ever saw one.

      1. Human nature hasn’t changed in the last 28 years. Any time there is a widely used benchmark for any product category, manufacturers will fine tune their product to do well on that benchmark, even if it does not improve other aspects of product performance. That may be misleading, but not illegal.

        More importantly, “your mileage may differ” applies to chips and computers as much as to automobiles. We have all heard about the YouTube video in which somebody sets every possible option in Premier to put a maximum strain on the CPU and GPU of a new MacBook Pro and gets it to overheat. That mightbe relevant to some other use case, or it might not. A wise shopper will determine whether there is a problem for the software they actually use in the circumstances they actually use it.

      2. Did you take a basic class in denial, Gareth? Your overt skepticism is unwarranted. As TxUser accurately states, human nature has not changed. If anything, it has worsened.

        Surely you realize that there have been other examples of deceptive practices used to artificially inflate benchmarks since that time? Samsung comes to mind, detecting the benchmark and shifting into overdrive to make it appear faster for just a short period of time. There are other examples.

        Don’t be so quick to accuse others of logical fallacies when your own assertion is basically worthless.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.