laitimes

AMD Ryzen™ 6000 Mobile Processor Runs Point Exposure, Inferior to 12th Generation Mobile Core?

At CES 2022 earlier this month, AMD officially unveiled its latest generation of Ryzen™ 6000 series mobile processors.

Compared with the Ryzen 5000 series mobile processors, the Ryzen 6000 series CPU architecture is replaced by the new Zen 3+, the GPU architecture is replaced by the new RDNA 2, and the process process is upgraded to 6nm. And the CPU acceleration frequency has reached 5GHz for the first time, and the core size continues to maintain a maximum of 8 cores and 16 threads, in addition to L2, L3 cache and TDP compared with the previous generation has not changed.

AMD Ryzen™ 6000 Mobile Processor Runs Point Exposure, Inferior to 12th Generation Mobile Core?

And just recently, a notebook equipped with a new Ryzen 9 6900HX processor appeared in the Geekbench database, this computer from Lenovo, the outside world speculated that it may be the savior series.

AMD Ryzen™ 6000 Mobile Processor Runs Point Exposure, Inferior to 12th Generation Mobile Core?

In terms of running points, this Ryzen™ 9 6900HX scored 1593 points for a single core and 10151 points for a multi-core. Compared with the previous generation of Ryzen™ 9 5900HX, it is about 12% and 33% higher.

But if you want to compete with your opponents, then this result may not be so good, compared to the 12th generation Core i9-12900H can lead the Ryzen™ 9 6900HX by 20% and 40% respectively.

AMD Ryzen™ 6000 Mobile Processor Runs Point Exposure, Inferior to 12th Generation Mobile Core?

In addition, some time ago, foreign media Chips and Cheese also tested the 3D cache version of the EPYC (Milan-X) processor.

According to tests of amd EPYC 7V73X, in OpenSSL, which does not require a large cache, the actual frequency of the 3D cache version is actually 2% slower.

However, in scenarios such as Gem5 compilation and 7-Zip compression, it is improved by 5 to 7%.

AMD Ryzen™ 6000 Mobile Processor Runs Point Exposure, Inferior to 12th Generation Mobile Core?

Summarizing and considering the frequency loss in virtualization tests, Chips and Cheese finally came to the conclusion that the overall performance of the 3D cache version was improved by 12.5%, which is frankly far lower than the amD official score of more than 50%. Of course, AMD's scenario is the measured EDA RTL validation workload on Synopsys VCS.

Read on