/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Primate Labs releases Geekbench 7, featuring larger, more demanding datasets for CPUs and GPUs, along with new video and audio encoding/decoding tests, and more

The Verge Antonio G. Di Benedetto

Context & Ripple Effects

Geekbench 7 extends Primate Labs’ recurring effort to revise its suite as workloads change. Geekbench 6 shifted toward datasets intended to better reflect real-world work, while the newer release broadens that approach across CPU, GPU, and media tasks.

The update also sits alongside Geekbench AI’s separate coverage of CPU, GPU, and NPU AI performance. That division makes Geekbench 7 chiefly a refreshed general-purpose performance baseline rather than a replacement for AI-specific measurement.

First-order effects

  • Reviewers, device makers, and buyers gain a new Geekbench baseline with heavier CPU/GPU datasets and dedicated video and audio encode/decode tests.
  • Scores produced under Geekbench 7 will need to be identified separately from Geekbench 6 results, since changed workloads can alter relative performance rather than merely raise or lower a score.

Second-order effects

  • Hardware comparisons can place more weight on media engines and GPU behavior, not only conventional CPU throughput, giving platforms with stronger specialized acceleration more opportunities to differentiate.
  • Publishers and benchmark databases will need to update test methodology and maintain version context; cross-device rankings lose meaning when results from different Geekbench generations are mixed.

Third-order effects

  • If benchmark suites continue to track increasingly mixed workloads, headline single-number comparisons will become less sufficient than workload-specific CPU, GPU, media, and AI results.
  • The pattern points to performance evaluation following heterogeneous compute: general benchmarks may stay important, but specialized suites such as Geekbench AI can become necessary complements as more tasks run on different processing blocks.

The trend: Consumer and PC benchmarking is moving from broad CPU-centric scores toward workload-specific measurement across heterogeneous CPU, GPU, media, and AI hardware.

Discussion

  • @mweinbach Max Weinbach on x
    Here's Geekbench 7 for my iPhone 17 Pro Max vs. S26 Ultra My iPhone should probably be a bit higher, but iOS 27 beta and betas usually hit it a bit [image]
  • @mweinbach Max Weinbach on x
    Another quick thing, Geekbench 7 is about as good of an agentic CPU benchmark as I can think of It's testing the workloads agents use your CPU for, and scales it well across multiple threads or single core. For those thinking about AI and agents on the CPU, it's good to use