Hewlett Packard Enterprise (HPE) between now and this summer plans to make available eight ProLiant Compute Gen12 servers that, in addition to enabling higher levels of performance in a smaller footprint, also include a processor that ensures higher levels of security.
Krista Satterthwaite, senior vice president and general manager for compute at HPE, said the servers, based on Intel Xeon 6 processors for data center and edge environments, will provide as much as 41% better performance per watt compared to legacy systems, with one Gen12 server providing the same compute performance as seven Gen10 servers.
The HPE ProLiant Compute Gen12 servers will also deliver up to 65% in power savings per year. HPE is offering optional direct liquid cooling (DLC) on Intel-based HPE ProLiant Compute Gen12 one-socket and two-socket rack servers to remove more than 3,000 times more heat, based on volume.
Those servers are also more secure thanks to the latest version of HPE Integrated Lights Out (iLO) management software that makes it possible to invoke dedicated security processors to encrypt firmware in a way that is robust enough to resist any effort to break it using quantum computers, noted Satterthwaite.
Embedded into the server hardware, the security processor developers by HPE established a secure enclave that protects against firmware attacks, she added.
At the same time, the company is also updating HPE Compute Ops Management, a cloud-based platform for managing servers, to now include an ability to forecast power usage requirements that are surfaced by machine learning algorithms. There is also now a global map view that makes it simpler to identify specific server health issues across distributed IT environments. HPE has also further automated server set-up and ongoing management. Those capabilities will be available across the HPE Proliant Compute Gen 10 through Gen 12 series of servers.
Finally, HPE for new customers is making available a standalone tool, dubbed HPE Power Advisor, that estimates metrics such as energy costs and greenhouse gas emissions.
It’s not clear at what pace IT teams are moving to upgrade legacy servers, but as more artificial intelligence (AI) capabilities are embedded into applications it’s apparent that much of the infrastructure relied on today will soon need to be replaced. Most of those platforms are simply not designed to meet the data processing requirements of AI applications.
The challenge, as always, will be prioritizing the acquisitions of the servers needed to run those applications, versus simply upgrading existing platforms with servers that will run existing applications more efficiently. There are not many IT organizations that can afford to deploy additional servers for new applications while simultaneously replacing legacy servers unless, of course, a legacy application is being retired.
In the meantime, IT teams might want to start identifying which servers are likely to be retired as the next generation of servers become available. Of course, it doesn’t follow that every server based on Intel processors will be replaced by a server based on a faster processor from Intel.
There are, after all, a lot more viable processor options available than any time in recent memory. The issue, as always, is determining what level of disruption can the organization absorb in pursuit of faster, more energy-efficient infrastructure that happens to provide the added bonus of being considerably more secure.
