Intel’s next-generation Xeon Skylake server processors have reached general availability on Google Compute Engine, in one of several updates announced in a blog post this week intended to boost the power and flexibility of instances.
Skylake processors are currently generally available in Google Cloud Platform’s Western US, Western Europe, and Eastern Asia Pacific regions, and will be phased in for others. Skylake VMs will be offered at no additional charge for 60 days to show off their performance, after which they will be available for 6 to 10 percent more than other processors, depending on the machine configuration.
Google first launched Skylake for Google Compute in February, and says that with up to 64 vCPUs and 455GB of RAM, the processor supports a wide range of compute-intensive workloads with performance increases confirmed through millions of hours of use by Google Cloud Platform (GCP) customers. The company says its public cloud is the only one offering Skylake processors.
Compute Engine has also launched a Minimum CPU Platform feature, which also brings with it Broadwell CPU support to every region, and the ability to create 64-vCPU VMs. Other Xeon CPUs available include Sandy Bridge, Ivy Bridge, and Haswell.
Google has removed the 6.5GB memory ratio restrictions to allow more memory per instance, which the company says benefits applications that can achieve better price-performance with flexible memory configurations, like in-memory databases, high-performance relational databases and NoSQL databases.
The collaborative partnership between Google and Intel has long given Google access to leading server processor technology, but it remains to be seen how compelling Skylake will be to enterprise customers.