AWS now renting monster HPE servers, even in clusters of 7,680-vCPUs and 128TB
Heir to Superdome goes cloudy for those who run large in-memory databases and apps that need them
Amazon Web Services usually stays schtum about the exact disposition of the servers it rents in its Elastic Compute Cloud, but made an exception on Tuesday with the announcement it is offering instances based on a single HPE server: the Compute Scale-up Server 3200.
HPE bills the server, launched in 2023, as ideal for applications such as large in-memory databases – a workload the Silicon Valley pioneer learned to handle in the late 1990s and early 2000s. At the time it was devising the Superdome architecture it offered as the highly scalable platform on which to run its HP-UX operating system.
Superdome persists as an architecture that allows modular scaling and memory pooling – if you use the right HPE servers – not the sort of trick AWS's commodity servers can perform. Some customers appear to be frustrated by that, because the cloud champ's announcement of the new offering declares "customers that currently run on-premises with HPE servers have also asked how we can help them migrate to AWS to take advantage of cloud benefits while continuing to use HPE hardware."
Other punters want SAP-certified rigs in the Amazonian cloud.
Installing the HPE servers solves both problems – it's well and truly ready for SAP and is the same hardware HPE offers on-prem.
- AWS says AI could disrupt everything – and hopes it will do just that to Windows
- We'll answer any questions DoJ has, HPE CEO tells us as Feds probe $14B Juniper buy
- HPE lets loose VM Essentials to run on third-party platforms
- BlackBerry squashes plan to spin out its IoT biz
The result is the poetically named U7inh instance type – each of which offers 1,920 vCPUs and 32TB of memory. Just like HPE's model 3200 server, the cloudy machines run fourth generation Intel Xeon Scalable processors. HPE's promo for the machine states it can host up to 16 CPU sockets, each capable of hosting 60 cores. With 1,920 vCPUs on offer, it appears these rent-a-servers are fully loaded.
The instances are U7inh certified to run Business Suite on HANA (SoH), Business Suite S/4HANA, Business Warehouse on HANA (BW), and SAP BW/4HANA in production environments.
The machines can also be clustered into groups of four – a scenario in which they are certified to run what AWS describes as "scale-out SAP HANA OLTP workloads such as S/4HANA."
The Amazonian cloud's "Nitro" platform is in place alongside these servers, providing its usual combination of isolation and security services.
These are not pay-as-you-go boxes, nor the sort of thing you can whistle up for some elastic capacity: AWS will only offer them on three-year Instance Savings Plans. At the time of writing, Amazon's pricing tools don’t list the U7inh instance type, but we can’t imagine it will be cheap, or plentiful. For now they're only available in the US East (N. Virginia) and US West (Oregon) AWS Regions. ®