On Tuesday , Primate Labs revealedGeekbench 6 , the late version of its pop bench mark tool . Used to gauge the CPU and GPU performance of Macs , iPhones , iPads , and non - Apple computing devices , Geekbench 6 ’s suite of run has update that better mull the workloads that today ’s gimmick can handle .
“ A lot has changed in the tech world in the past three years . Smartphone cameras take bigger and better mental picture . Artificial intelligence , especially motorcar learning , has become ubiquitous in oecumenical and fluid applications,”wroteJohn Poole in a blog post . “ The number of cores in computers and roving gimmick continues to uprise . And how we interact with our computers and mobile devices has changed dramatically – who would have guessed that video conferencing would suddenly surge in 2020 ? ”
To better muse New workflows , Geekbench 6 implement new tasks , such as background fuzz in videoconferencing , filtering and adjust images for social media , mechanically removing unwanted objects from photo , detecting and tagging objects in photos using simple machine encyclopaedism mannikin , and using scripting terminology to analyze , process , and commute textbook . Data sets have also been updated with larger files and high - resolution images .
Geekbench explain that the new CPU bench mark scores “ are used to appraise and optimize central processor and memory performance using workload that include data point compressing , epitome processing , machine learning , and compiling code . ”
The fresh tests are similar in nature to Geekbench 5 , but do newer tasks with new data set and depository library to better act forward-looking program . you may not directly compare Geekbench 5 and Geekbench 6 score . Multi - core performance , in particular , uses new methods of share work between cores that should considerably represent the performance of CPUs that have “ high performance ” and “ gamy efficiency ” essence . Inan interview with Ars Technica , Poole say , “ We really want this to be a tool that multitude can use to calculate out interrogation they have about performance . ”
Geekbench 6 features new examination criteria to intimately measure real - world performance of your Apple devices .
Primate Labs
Macworld use Geekbench in our iPhone , iPad , and Mac brushup , and we ’ll be using Geekbench 6 . However , previous upshot record with Geekbench 5 ca n’t be compared to Geekbench 6 , since the two role dissimilar test suites . In the short full term , we may include Geekbench 5 results to provide a historical setting for old devices we are unable to essay with Geekbench 6 . it ’s hard to compare the scores without doing extensive examination , but a trial of the 14 - inch MacBook Pro with an 8 - meat CPU and 14 - core GPU returned scores of 2138 ( exclusive - core ) and 8475 ( multi - core ) versus an average of 1750 ( single ) and 9550 ( multi ) using Geekbench 5 . We also test the 14 - inch MacBook Pro with a 10 - core M1 Pro and 16 - core GPU and got Geekbench 6 scores of 2399 ( single ) and 12289 ( multi ) , versus Geekench 5 sexual conquest of 1778 ( single ) and 12544 ( multi ) .
Geekbench 6 is detached for personal use and can be downloaded from theGeekbench site(Mac ) or theApp Store(iOS ) . The Pro variation allows user to start the app from a portable storage gimmick or internet private road , save results offline , automate testing , and provides a license for commercial use . Geekbench 6 Pro is $ 99 , but Primate Labs offers 20 percent discount ( $ 79 ) until February 28 .
We talk about how to benchmark your Mac here : Mac Speed Test : Test your Mac ’s public presentation with these benchmark tools .
What do the Geekbench scores mean?
What do these Geekbench scores have in mind ? What is it measuring and why should you give care ?
Geekbench run through a stage set of specific repeatable trial that stress either your mainframe or your nontextual matter processor ( GPU ) in specific agency and turns that into a numerical score . The tests and the data point they employ are the same on every chopine , but some platform may have hardware that can speed up sure part of it ( like file compression ) . A high is better , but it does n’t represent a concrete economic value like “ sentence elapsed ” or “ pixels per second . ” It ’s an entirelysyntheticbenchmark and score .
CPU tests
The CPU test includes thing like file compression ( compress and decompress a 75 Bachelor of Medicine archive with 9,841 data file in a variety of method acting ) , bring forth directions between a chronological succession of location , rendering 8 democratic web varlet , rendering PDFs , running developer scripts , and machine - learning labor like object detection and ground blur .
The CPU run has two consequence score : Single - core , which measures the public presentation of launch the tasks on a single CPU pith , and multi - core , which uses all CPU substance . Both are of import – many applications have their carrying into action limited by a unmarried primary screw thread , so exclusive - core CPU performance will determine how tight they run . But for applications that are well multi - threaded or for melt down multiple app at once , multi - core is a good mensuration of the full maximum mainframe functioning .
GPU Compute tests
You might think that evaluate nontextual matter CPU ( GPU ) public presentation is kindred to knowing what kind of frame rate you’re able to expect in 3D games . When it come to Geekbench , that is not necessarily the example . The Geekbench tests donotmeasure 3D graphics render functioning . Instead , it measures the performance of “ GPU Compute ” tasks – using the GPU to handle computational tasks like edge detection on images , utilize a gaussian fuzz to a 24 - megapixel photo , a particle physics computer simulation , or discover faces in photos .
It ’s quite potential for one product to track down 3D games faster than the other while scoring lower on the Geekbench GPU Compute tests . While GPU Compute operation and 3D graphics performance are often related in some fashion , it ’s clean to say that if you want to measure 3D graphics performance you should run a 3D graphic benchmark , not a GPU Compute benchmark like Geekbench .
When you launch the Geekbench 6 GPU Compute test , you get to choose which API ( program programing interface – the way developers talk to the hardware ) that you need it to apply . On a Mac , you may choose between OpenCL and Metal . On iPhones and iPads , the only support API is Metal . Your alternative does not change the tests at all , only the way that the coating verbalize to the hardware . On the Mac , Metal is plausibly the most important score and should probably be compared to Vulkan performance on Windows PCs and Android .
If you ’re technically inclined , you may read PDF white papers that describe the tests in point . There ’s a white paper for theCPU testsand theGPU Compute tests .
