{"slug": "geekbench", "name": "Geekbench", "category": "other", "category_name": "Other software", "description": "", "recommenders": 4, "commenters": 16, "warners": 6, "score": -1.025, "rising": -1.31, "first": "2025-11-12", "latest": "2026-09-19", "why": "the only benchmark I use to quickly judge some CPU is geekbench", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 3, "neutral": 7, "compare": 1, "warn": 6}, "spark": [2, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "the only benchmark I use to quickly judge some CPU is geekbench", "kind": "", "icon": false, "domain": "geekbench.com", "github_repo": null, "official_url": "https://www.geekbench.com", "pricing": "", "aliases": ["Geekbench"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 784444480, "item_id": 49027780, "url": "https://news.ycombinator.com/item?id=49027780", "by": "GeekyBear", "date": "2026-07-23", "stance": "recommend", "quote": "People really shouldn't call Geekbench a synthetic benchmark anymore, since it has been using real world workloads for years now.", "excerpt": "Thanks for finding that. > The Clang workload uses the Clang compiler to compile the Lua interpreter, a popular open-source language interpreter. People really shouldn't call Geekbench a synthetic benchmark anymore, since it has been using real world workloads for years now.", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 756693472, "item_id": 47293342, "url": "https://news.ycombinator.com/item?id=47293342", "by": "osti", "date": "2026-03-07", "stance": "recommend", "quote": "the only benchmark I use to quickly judge some CPU is geekbench", "excerpt": "Not true. Geekbench, especially single threaded benchmark, is probably the best we got, it has a bunch of workloads, unlike many other benchmarks like cinebench for example. And they publish all the results on their website, so you can dig into each individual workload and find the ones that apply to you. And like the other poster mentioned, it correlates well with SPEC, so it's basically a easily accessible SPEC. Th", "context": "", "use_case": "quick CPU comparisons", "switched_from": "", "story": "Does Apple\u2018s M5 Max Really \u201cDestroy\u201d a 96-Core Threadripper?"}, {"id": 735067952, "item_id": 45941747, "url": "https://news.ycombinator.com/item?id=45941747", "by": "Aurornis", "date": "2025-11-15", "stance": "recommend", "quote": "For productivity you can check Phoronix or even the GeekBench details in the compiler section", "excerpt": "> - Visit https://www.cpubenchmark.net/single-thread/ and pick the fastest CPU under $400 FYI www.cpubenchmark.com is a running joke for how bad it is. It\u2019s not a good resource. There are a few variations of these sites like userbenchmark that have been primarily built for SEO spam and capturing Google visitors who don\u2019t know where to go for good buying advice. Buying a CPU isn\u2019t really that complicated. For gaming i", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "AMD continues to chip away at Intel's x86 market share"}, {"id": 734536256, "item_id": 45908516, "url": "https://news.ycombinator.com/item?id=45908516", "by": "bigyabai", "date": "2025-11-12", "stance": "recommend", "quote": "Pull up a unified benchmark like the OpenCL Geekbench", "excerpt": "But why is it overdue? It's easy to put the performance profile of a console on an SOC, it's impossible to integrate many desktop GPUs into the same form factor. Pull up a unified benchmark like the OpenCL Geekbench, it makes this obvious. The most powerful SOCs, like the M3 Ultra, pull over 250w to get worse scores than a 4080 laptop dGPU: https://browser.geekbench.com/opencl-benchmarks How are SOCs going to replace", "context": "", "use_case": "", "switched_from": "", "story": "Steam Machine"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 796257504, "item_id": 49766094, "url": "https://news.ycombinator.com/item?id=49766094", "by": "tom_", "date": "2026-09-19", "stance": "warn", "quote": "Geekbench might flatter the Apple chips for multicore a bit", "excerpt": "Judging by my PCs (M4 Max Mac Studio; 2990WX desktop PC), Geekbench might flatter the Apple chips for multicore a bit I think: Geekbench 7 results: * AMD 2990WX: 1384 (single), 13052 (multi) (https://browser.geekbench.com/v7/cpu/181239) * Apple M4 Max: 3552 (single), 29863 (multi) (https://browser.geekbench.com/v7/cpu/390256) For parallelisable stuff that can occupy all cores for an extended period, the 2990WX typica", "context": "", "use_case": "", "switched_from": "", "story": "Apple M6 Pro achieves the highest single-core CPU score in Geekbench 7"}, {"id": 784617120, "item_id": 49038570, "url": "https://news.ycombinator.com/item?id=49038570", "by": "adrian_b", "date": "2026-07-24", "stance": "warn", "quote": "The Geekbench multi-threaded results do not provide any useful information about how a computer will behave for such applications.", "excerpt": "A very frequent real-world workload is the compilation of a software project, which scales almost perfectly with the number of cores. For the other multi-threaded tasks, like running certain CAD/EDA applications, I use well-optimized programs, either by others or by myself. What I do is typical for a professional use of a computer. The Geekbench multi-threaded results do not provide any useful information about how a", "context": "professional computer user running compilation and CAD/EDA", "use_case": "multi-threaded benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 784427792, "item_id": 49026737, "url": "https://news.ycombinator.com/item?id=49026737", "by": "thrownawaysz", "date": "2026-07-23", "stance": "warn", "quote": "Never really liked Geekbench.", "excerpt": "Never really liked Geekbench. Synthetic benchmarks are easy to use and easy to understand but borderline useless. 1000 is better than 990 but it literally doesn't mean anything at all. Just an example: the 2017 Intel Xeon gets more points than an Apple M1 on multicore but everyone knows that they are incomparable. And no one should get a 2017 Intel Xeon because it's scoring higher yet that's the whole point of Geekbe", "context": "", "use_case": "", "switched_from": "", "story": "Geekbench 7"}, {"id": 779722656, "item_id": 48732666, "url": "https://news.ycombinator.com/item?id=48732666", "by": "pocketsand", "date": "2026-06-30", "stance": "warn", "quote": "Geekbench is bursty", "excerpt": "Obvious massive asterisk that benchmarks are not always realworld use: Apple M5, Single core Geekbench: ~4,200 (https://browser.geekbench.com/macs/macbook-pro-14-inch-2025) Apple M5 Pro (15-core, lower core version), multicore: ~26,000 (https://browser.geekbench.com/v6/cpu/18535781) compare this to the top of the line Intel Panther Lake chips, which have comparable battery life. I cherry picked a 16\" Dell XPS machine", "context": "", "use_case": "", "switched_from": "", "story": "Free the Icons"}]}], "who": [], "tags": [], "use_cases": ["cpu benchmarking", "consumer cpu benchmarking", "multi-threaded benchmarking", "quick cpu comparisons"], "compared": [], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49025812", "title": "Geekbench 7", "mentions": 6}, {"url": "https://news.ycombinator.com/item?id=49763883", "title": "Apple M6 Pro achieves the highest single-core CPU score in Geekbench 7", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=47291906", "title": "Does Apple\u2018s M5 Max Really \u201cDestroy\u201d a 96-Core Threadripper?", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=45940413", "title": "AMD continues to chip away at Intel's x86 market share", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=45903404", "title": "Steam Machine", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=48698908", "title": "Free the Icons", "mentions": 1}], "sentiment_by_month": {"2025-11": {"neutral": 1, "recommend": 2, "warn": 1}, "2025-12": {"neutral": 1}, "2026-01": {}, "2026-02": {}, "2026-03": {"neutral": 2, "recommend": 1}, "2026-04": {}, "2026-05": {"warn": 1, "neutral": 1}, "2026-06": {"warn": 1}, "2026-07": {"warn": 2, "neutral": 1, "compare": 1}, "2026-08": {}, "2026-09": {"neutral": 1, "warn": 1}, "2026-10": {}}, "quotes": [{"id": 796257504, "item_id": 49766094, "url": "https://news.ycombinator.com/item?id=49766094", "by": "tom_", "date": "2026-09-19", "stance": "warn", "quote": "Geekbench might flatter the Apple chips for multicore a bit", "excerpt": "Judging by my PCs (M4 Max Mac Studio; 2990WX desktop PC), Geekbench might flatter the Apple chips for multicore a bit I think: Geekbench 7 results: * AMD 2990WX: 1384 (single), 13052 (multi) (https://browser.geekbench.com/v7/cpu/181239) * Apple M4 Max: 3552 (single), 29863 (multi) (https://browser.geekbench.com/v7/cpu/390256) For parallelisable stuff that can occupy all cores for an extended period, the 2990WX typica", "context": "", "use_case": "", "switched_from": "", "story": "Apple M6 Pro achieves the highest single-core CPU score in Geekbench 7"}, {"id": 796234432, "item_id": 49764652, "url": "https://news.ycombinator.com/item?id=49764652", "by": "tristanj", "date": "2026-09-19", "stance": "neutral", "quote": "Leaked prior to launch - Leaked Benchmarks Confirm M2 Chip", "excerpt": "Not accurate, both leak very frequently. Leaked prior to launch - Leaked Benchmarks Confirm M2 Chip is Up to 20% Faster Than M1 https://www.macrumors.com/2022/06/15/m2-geekbench-benchmark/ Leaked prior to announcement - 'M2 Max' Geekbench Scores Leak Online, Revealing Rumored Specs and Performance https://www.macrumors.com/2022/11/30/m2-max-chip-geekbench-results/ Leaked prior to launch - M3 Max Chip Around as Fast a", "context": "", "use_case": "", "switched_from": "", "story": "Apple M6 Pro achieves the highest single-core CPU score in Geekbench 7"}, {"id": 784643328, "item_id": 49040208, "url": "https://news.ycombinator.com/item?id=49040208", "by": "wtallis", "date": "2026-07-24", "stance": "neutral", "quote": "Geekbench has that as a subtest.", "excerpt": "> A very frequent real-world workload is the compilation of a software project, which scales almost perfectly with the number of cores. Geekbench has that as a subtest. > For the other multi-threaded tasks, like running certain CAD/EDA applications, I use well-optimized programs, either by others or by myself. Those seem pretty clearly out of scope. You want a workstation benchmark, not a consumer benchmark. > This w", "context": "", "use_case": "consumer CPU benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 784617120, "item_id": 49038570, "url": "https://news.ycombinator.com/item?id=49038570", "by": "adrian_b", "date": "2026-07-24", "stance": "warn", "quote": "The Geekbench multi-threaded results do not provide any useful information about how a computer will behave for such applications.", "excerpt": "A very frequent real-world workload is the compilation of a software project, which scales almost perfectly with the number of cores. For the other multi-threaded tasks, like running certain CAD/EDA applications, I use well-optimized programs, either by others or by myself. What I do is typical for a professional use of a computer. The Geekbench multi-threaded results do not provide any useful information about how a", "context": "professional computer user running compilation and CAD/EDA", "use_case": "multi-threaded benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 784609840, "item_id": 49038115, "url": "https://news.ycombinator.com/item?id=49038115", "by": "wtallis", "date": "2026-07-24", "stance": "compare", "quote": "cheating on Geekbench is more difficult", "excerpt": "I don't think you've put forth a coherent argument here. SPEC CPU is distributed in source form, to be compiled by the end user. There are restrictions on what kind of compiler options and other games you can play when making an official submission to SPEC, which Intel (and others) get around by publishing \"estimated\" SPEC scores and using those for marketing claims. In practice, the amount of bullshit in Intel's cla", "context": "", "use_case": "", "switched_from": "", "story": "Geekbench 7"}, {"id": 784444480, "item_id": 49027780, "url": "https://news.ycombinator.com/item?id=49027780", "by": "GeekyBear", "date": "2026-07-23", "stance": "recommend", "quote": "People really shouldn't call Geekbench a synthetic benchmark anymore, since it has been using real world workloads for years now.", "excerpt": "Thanks for finding that. > The Clang workload uses the Clang compiler to compile the Lua interpreter, a popular open-source language interpreter. People really shouldn't call Geekbench a synthetic benchmark anymore, since it has been using real world workloads for years now.", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 784439168, "item_id": 49027448, "url": "https://news.ycombinator.com/item?id=49027448", "by": "GeekyBear", "date": "2026-07-23", "stance": "neutral", "quote": "In the past, they have provided information on which workloads they use.", "excerpt": "In the past, they have provided information on which workloads they use. For instance, the file compression workload on version 6: This workload compresses and decompresses the Ruby 3.1.2 source archive (a 75 MB archive with 9,841 files) using different compression codecs (such as LZ4 and ZSTD). It also verifies the files using the SHA1 (Secure Hash Algorithm 1) function. https://www.geekbench.com/doc/geekbench6-cpu-", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Geekbench 7"}, {"id": 784427792, "item_id": 49026737, "url": "https://news.ycombinator.com/item?id=49026737", "by": "thrownawaysz", "date": "2026-07-23", "stance": "warn", "quote": "Never really liked Geekbench.", "excerpt": "Never really liked Geekbench. Synthetic benchmarks are easy to use and easy to understand but borderline useless. 1000 is better than 990 but it literally doesn't mean anything at all. Just an example: the 2017 Intel Xeon gets more points than an Apple M1 on multicore but everyone knows that they are incomparable. And no one should get a 2017 Intel Xeon because it's scoring higher yet that's the whole point of Geekbe", "context": "", "use_case": "", "switched_from": "", "story": "Geekbench 7"}, {"id": 779722656, "item_id": 48732666, "url": "https://news.ycombinator.com/item?id=48732666", "by": "pocketsand", "date": "2026-06-30", "stance": "warn", "quote": "Geekbench is bursty", "excerpt": "Obvious massive asterisk that benchmarks are not always realworld use: Apple M5, Single core Geekbench: ~4,200 (https://browser.geekbench.com/macs/macbook-pro-14-inch-2025) Apple M5 Pro (15-core, lower core version), multicore: ~26,000 (https://browser.geekbench.com/v6/cpu/18535781) compare this to the top of the line Intel Panther Lake chips, which have comparable battery life. I cherry picked a 16\" Dell XPS machine", "context": "", "use_case": "", "switched_from": "", "story": "Free the Icons"}, {"id": 767791680, "item_id": 47986980, "url": "https://news.ycombinator.com/item?id=47986980", "by": "llm_nerd", "date": "2026-05-02", "stance": "neutral", "quote": "That specific Geekbench test is to measure the ANE performance", "excerpt": "\"We might hope that macOS would process AI tasks using the CPU and GPU rather than the neural engine, when running in a VM.\" That specific Geekbench test is to measure the ANE performance, which they did by setting the CoreML run to cpuAndNeuralEngine. They could have set it to all and it would use any hardware available, but that would be counterproductive to a test that hopes to measure the ANE, no? And note that t", "context": "", "use_case": "", "switched_from": "", "story": "How fast is a macOS VM, and how small could it be?"}, {"id": 767571616, "item_id": 47973226, "url": "https://news.ycombinator.com/item?id=47973226", "by": "AnthonyMouse", "date": "2026-05-01", "stance": "warn", "quote": "Geekbench is basically trash.", "excerpt": "Geekbench is basically trash. People keep using it for comparing Mac performance because many of the things people usually benchmark don't run on Macs. But single-number outputs like that are useless. Is the number ~10% higher because it's consistently ~10% faster at everything, or because it's 100% faster on a minority of things and slower at everything else? The first one is pretty unlikely when comparing processor", "context": "", "use_case": "", "switched_from": "", "story": "Apple Says Mac Studio and Mac Mini Will Be in Short Supply for Months"}, {"id": 757866801, "item_id": 47366675, "url": "https://news.ycombinator.com/item?id=47366675", "by": "kotaKat", "date": "2026-03-13", "stance": "neutral", "quote": "Geekbench 6 was around ~2600 single-core with the VM overhead for me", "excerpt": "Geekbench 6 was around ~2600 single-core with the VM overhead for me. That's still punching above single-core power in its class for Windows machines and it makes me giggle. https://browser.geekbench.com/v6/cpu/17011372 This was the latest UTM in the App Store, so native Hypervisor.Framework access for arm64 Windows acceleration.", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Parallels Confirms MacBook Neo Can Run Windows in a Virtual Machine"}, {"id": 756693472, "item_id": 47293342, "url": "https://news.ycombinator.com/item?id=47293342", "by": "osti", "date": "2026-03-07", "stance": "recommend", "quote": "the only benchmark I use to quickly judge some CPU is geekbench", "excerpt": "Not true. Geekbench, especially single threaded benchmark, is probably the best we got, it has a bunch of workloads, unlike many other benchmarks like cinebench for example. And they publish all the results on their website, so you can dig into each individual workload and find the ones that apply to you. And like the other poster mentioned, it correlates well with SPEC, so it's basically a easily accessible SPEC. Th", "context": "", "use_case": "quick CPU comparisons", "switched_from": "", "story": "Does Apple\u2018s M5 Max Really \u201cDestroy\u201d a 96-Core Threadripper?"}, {"id": 756673456, "item_id": 47292091, "url": "https://news.ycombinator.com/item?id=47292091", "by": "Aurornis", "date": "2026-03-07", "stance": "neutral", "quote": "GeekBench probably made the right choice to optimize for more realistic real-world workloads", "excerpt": "GeekBench probably made the right choice to optimize for more realistic real-world workloads than for the more specific workloads that benefit from really high core counts. GeekBench is supposed to be a proxy for common use case performance. High core count CPUs are only useful for specific workloads and should not be purchased as general purpose fast CPUs. Unless you\u2019re doing specific tasks that scale by core count,", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Does Apple\u2018s M5 Max Really \u201cDestroy\u201d a 96-Core Threadripper?"}, {"id": 742219328, "item_id": 46388708, "url": "https://news.ycombinator.com/item?id=46388708", "by": "Telaneo", "date": "2025-12-25", "stance": "neutral", "quote": "Have a look at Geekbench's results.", "excerpt": "Not in raw performance, no, but they're only beat out by i9s and the like, which are very power hungry. If you care even a little bit about performance per watt, the M series are far superior. Have a look at Geekbench's results.[1] Ignore the top ones, since they're invalid and almost certainly cheated (click to check). The iPads and such lower down are all legit, but the same goes for some of the i9s inbetween. And ", "context": "", "use_case": "", "switched_from": "", "story": "Asahi Linux with Sway on the MacBook Air M2"}, {"id": 735072544, "item_id": 45942034, "url": "https://news.ycombinator.com/item?id=45942034", "by": "embedding-shape", "date": "2025-11-15", "stance": "warn", "quote": "just puts a list of all benchmarks, even when the CPU is the same", "excerpt": "> For productivity you can check Phoronix or even the GeekBench details in the compiler section I guess the reason people prefer something like cpubenchmark, is because it seems way easier to get an overview / see data in aggregate. GeekBench (https://browser.geekbench.com/v6/cpu/multicore) for example just puts a list of all benchmarks, even when the CPU is the same. Not exactly conductive for finding the right CPU.", "context": "", "use_case": "", "switched_from": "", "story": "AMD continues to chip away at Intel's x86 market share"}, {"id": 735067952, "item_id": 45941747, "url": "https://news.ycombinator.com/item?id=45941747", "by": "Aurornis", "date": "2025-11-15", "stance": "recommend", "quote": "For productivity you can check Phoronix or even the GeekBench details in the compiler section", "excerpt": "> - Visit https://www.cpubenchmark.net/single-thread/ and pick the fastest CPU under $400 FYI www.cpubenchmark.com is a running joke for how bad it is. It\u2019s not a good resource. There are a few variations of these sites like userbenchmark that have been primarily built for SEO spam and capturing Google visitors who don\u2019t know where to go for good buying advice. Buying a CPU isn\u2019t really that complicated. For gaming i", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "AMD continues to chip away at Intel's x86 market share"}, {"id": 734536256, "item_id": 45908516, "url": "https://news.ycombinator.com/item?id=45908516", "by": "bigyabai", "date": "2025-11-12", "stance": "recommend", "quote": "Pull up a unified benchmark like the OpenCL Geekbench", "excerpt": "But why is it overdue? It's easy to put the performance profile of a console on an SOC, it's impossible to integrate many desktop GPUs into the same form factor. Pull up a unified benchmark like the OpenCL Geekbench, it makes this obvious. The most powerful SOCs, like the M3 Ultra, pull over 250w to get worse scores than a 4080 laptop dGPU: https://browser.geekbench.com/opencl-benchmarks How are SOCs going to replace", "context": "", "use_case": "", "switched_from": "", "story": "Steam Machine"}, {"id": 734502274, "item_id": 45906392, "url": "https://news.ycombinator.com/item?id=45906392", "by": "mhitza", "date": "2025-11-12", "stance": "neutral", "quote": "Using Geekbench I saw only a 5% cpu performance penalty.", "excerpt": "I've used SteamOS as a daily driver for half a year. Immutable distros have limitations and my distrobox images failed to work after a SteamOS update. If you're ok with running work stuff in a separate VM within SteamOS, that works great. Using Geekbench I saw only a 5% cpu performance penalty. Io takes a bigger hit, but that wasn't a blocker for me as I was intending to run VMs with encrypted storage anyway (which a", "context": "", "use_case": "CPU benchmarking", "switched_from": "", "story": "Steam Machine"}]}