AMD, a company famous for its subtlety and restraint when dealing with competitors, is once again accusing Intel of unfair business practices, this time around dodgy benchmarks. Rather than provide useful metrics to consumers, AMD says Intel's use of the SYSmark benchmarking tool unfairly favours Intel products by putting too much emphasis on raw CPU performance. It even went as far to compare said practice to the recent VW emissions scandal, in which diesel engines could detect when they were having their emissions tested, changing their performance accordingly to improve results.
"Information provided by even the most established organisations can be misleading," explains AMD's John Hampton in a video titled "Truth or Myth?: Is SYSmark a Reliable Benchmark?" on YouTube. The video, in which a po-faced Hampton and technical sidekick Tony Salinas explain the Intel problem in true 60 Minutes style, shows SYSmark on a pair of PCs, one powered by an Intel Core i5 processor and the other by an AMD FX chip. The Core i5 system scored 987, while a "comparable AMD platform" running an FX processor scored 659. "That's a delta of 50 percent. Quite astonishing and not realistic in what real life performance is like," says Salinas.
To try and prove that AMD's chips are comparable to Intel's in real-world usage scenarios, the pair then boot up PCMark8, which has "activity going on on the CPU, GPU, and video sub-components of the system." The performance delta drops down to around seven percent. To hammer the point home they then load up a custom script, which looks at the time taken to complete a set of tasks within Microsoft Office. The Intel system completed it in 62 seconds, while the AMD system completed it in 64, thus providing the "better representation of real performance" the company is so keen to share.
That AMD has dug up these issues with SYSmark is odd given the company already won a judgement from the US Federal Trade Commission (FTC) regarding the benchmark back in 2010 (PDF link). Intel is now forced to include the following statement when citing benchmark results:
"Software and workloads used in performance tests may have been optimised for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations, and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products."
While AMD does have a point that there's little difference between AMD and Intel chips for mainstream users, to completely snub the CPU-focused benchmarks of SYSmark isn't entirely fair either. For enthusiasts that want the fastest system, regardless of cost or other factors, the raw CPU performance that SYSmark tests is just as important as overall system performance. That's especially true for games, which currently favour higher single-threaded performance over more cores, at least until DirectX 12 arrives.
This isn't the first time that AMD has spoken out over Intel's business practices. Back in the the mid-2000s, AMD believed Intel was issuing payments to manufacturers like Toshiba and NEC, on the promise that they wouldn’t use AMD processors in their products. More recently, the company said Nvidia "completely sabotaged" Witcher 3 performance with its GameWorks APIs, and also implied sites like the The Tech Report wrote unfair reviews.
reader comments
169