Intel’s Binary Optimization Tool Skews Geekbench 6 Results
The latest Intel CPUs might appear more impressive than they actually are, thanks to a new tool from Intel that is impacting benchmark scores. Specifically, Intel’s Binary Optimization Tool is altering Geekbench 6 results, artificially boosting performance metrics for the company’s newest processors.
Primate Labs, the developer of the widely-used Geekbench CPU and GPU benchmarking software, recently announced that Intel’s Binary Optimization Tool is creating inaccurate scores for Intel CPUs. The tool works by modifying instruction sequences to improve performance. Whereas performance improvements are generally positive, benchmarks are only reliable when they measure the same exact workload consistently. Altered benchmarks render score comparisons between different devices meaningless.
How Much Are Scores Affected?
When Geekbench 6 is run through the Binary Optimization Tool, some workload scores can increase by up to 40%, with overall scores seeing boosts of up to 8%. These inflated scores do not accurately reflect real-world computing performance. Currently, the Intel Binary Optimization Tool supports a limited number of applications, with Geekbench 6 being one of the few.
According to Primate Labs, “The techniques used are not publicly documented, and it is unclear how widely applicable these techniques are across different applications.” The tool is currently compatible with certain Core Ultra Series 3 processors (“Panther Lake”) and Core Ultra 200 Plus chips (“Arrow Lake Refresh”).
What’s Being Done About It?
Because Geekbench currently has “no way to detect if a Geekbench 6 result was run with or without the Binary Optimization Tool,” all benchmark results from CPUs that support the tool will now display a warning message indicating the result might be invalid.
This isn’t the first time benchmark manipulation has come to light. Similar practices have been observed in the smartphone industry, with companies like Samsung, LG, and OnePlus temporarily overclocking devices during benchmark tests. Intel itself was previously accused of using a compiler hack to inflate benchmark results for its Xeon processors.
What Does This Mean for Consumers?
If you’re comparing CPU performance, especially when considering newer Intel CPUs, it’s crucial to verify whether the Geekbench 6 test was conducted without the influence of Intel’s Binary Optimization Tool. Currently, Geekbench 6.3 is the only version affected, but Intel may expand the tool’s compatibility to other applications in the future.
Source: Geekbench Blog