AI News AI资讯 3h ago Updated 2h ago 更新于 2小时前 46

Geekbench 7 will push your computer or phone even harder for better benchmarking Geekbench 7 将更用力地测试您的电脑或手机,以实现更好的基准测试

Primate Labs has released Geekbench 7, introducing new video and audio encoding/decoding tests with support for AV1 and Opus codecs. The multi-core test has been redesigned to better reflect real-world software behavior by only running multi-threaded when modeled tasks require it. CPU and GPU benchmarks now utilize larger, more demanding datasets to accurately capture the hardware demands of modern applications. Geekbench 7 scores are not compatible with Geekbench 6, requiring users to retest ex Primate Labs发布Geekbench 7,引入新的音视频编解码测试及重新设计的多核测试。 采用更大、更严苛的数据集以准确捕捉现代硬件在真实应用中的负载需求。 新增AV1、Opus等格式支持,涵盖视频会议、流媒体和内容消费等工作负载。 多核基准测试优化,仅在模拟任务需要时才运行多线程,更贴合实际软件行为。 价格维持99美元(Pro版),Geekbench 6分数与新版不兼容,需重新测试。

65
Hot 热度
70
Quality 质量
60
Impact 影响力

Analysis 深度分析

TL;DR

  • Primate Labs has released Geekbench 7, introducing new video and audio encoding/decoding tests with support for AV1 and Opus codecs.
  • The multi-core test has been redesigned to better reflect real-world software behavior by only running multi-threaded when modeled tasks require it.
  • CPU and GPU benchmarks now utilize larger, more demanding datasets to accurately capture the hardware demands of modern applications.
  • Geekbench 7 scores are not compatible with Geekbench 6, requiring users to retest existing hardware for comparison.
  • The Pro license remains priced at $99 (discounted to $79 until August 6th) and supports offline benchmarking on Windows, macOS, and Linux.

Why It Matters

This update addresses critical gaps in synthetic benchmarking by incorporating modern media workloads like AV1 and Opus, which are increasingly prevalent in streaming and content creation. For AI practitioners and hardware reviewers, the shift toward more realistic multi-threading logic and larger datasets provides a more accurate assessment of system performance under actual application loads rather than just synthetic bursts.

Technical Details

  • New Media Workloads: Includes dedicated tests for audio and video encoding/decoding, specifically supporting AV1 and Opus formats to model video conferencing, streaming, and content consumption.
  • Redesigned Multi-Core Logic: The multi-core test now mimics real software behavior by executing multi-threaded operations only when the specific workload requires them, improving relevance over previous versions.
  • Enhanced Datasets: Both CPU and GPU tests employ larger and more demanding data sets to stress-test hardware capabilities more effectively than the lighter loads used in prior iterations.
  • Platform Compatibility: Supports all compatible desktop platforms including Windows, macOS, and Linux, with the Pro version allowing for offline benchmarking without data upload.

Industry Insight

  • Benchmarking Evolution: The inclusion of AV1 and Opus tests signals that media efficiency is becoming a primary metric for hardware evaluation, urging manufacturers to optimize for these specific codecs.
  • Performance Baseline Shift: Since scores are not backward-compatible with Geekbench 6, industry comparisons will need to reset baselines, making this a pivotal moment for re-evaluating current-generation hardware against legacy standards.
  • Real-World Relevance: The focus on realistic multi-threading and sustained load simulation suggests a move away from pure synthetic speed towards evaluating how well hardware handles complex, mixed-workload scenarios typical in professional environments.

TL;DR

  • Primate Labs发布Geekbench 7,引入新的音视频编解码测试及重新设计的多核测试。
  • 采用更大、更严苛的数据集以准确捕捉现代硬件在真实应用中的负载需求。
  • 新增AV1、Opus等格式支持,涵盖视频会议、流媒体和内容消费等工作负载。
  • 多核基准测试优化,仅在模拟任务需要时才运行多线程,更贴合实际软件行为。
  • 价格维持99美元(Pro版),Geekbench 6分数与新版不兼容,需重新测试。

为什么值得看

对于AI从业者和硬件评测人员而言,Geekbench 7通过引入更复杂的音视频处理测试和更真实的多线程调度逻辑,提供了评估现代计算设备性能的新标准。其强调的“更严苛数据集”有助于识别设备在长时间高负载下的真实性能表现,而非仅关注瞬时爆发力。

技术解析

  • 新增多媒体测试:引入了针对视频和音频编码/解码的专项测试,明确支持AV1和Opus等现代编解码格式,覆盖模型视频会议、流媒体传输及内容消费场景。
  • 多核测试重构:重新设计了多核测试逻辑,使其不再盲目全核满载,而是根据模拟任务的特性决定是否启用多线程,从而更准确地反映真实软件的行为模式。
  • 数据集升级:CPU和GPU测试使用了更大、更具挑战性的数据集,旨在更精确地衡量硬件在处理现代应用程序时的实际需求和压力承受能力。
  • 兼容性说明:Geekbench 7与Geekbench 6分数体系不互通,用户需进行新一轮的基准测试以建立新的性能对比基线。

行业启示

  • 性能评估维度转变:随着AI和多媒体应用的普及,硬件基准测试正从单纯的算力堆砌转向对特定工作负载(如AV1解码、实时音视频处理)的精细化考量。
  • 真实负载模拟的重要性:行业应关注那些能模拟真实软件行为(如动态多线程使用)的测试工具,避免被仅测试瞬时峰值性能的基准所误导。
  • 持续测试的必要性:由于基准版本迭代导致历史数据不可比,企业和评测机构需建立新的性能基线数据库,以便在新硬件发布时进行准确的横向对比。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Benchmark 基准测试 GPU GPU Product Launch 产品发布