CloudDatacentre

HPE Discover: ‘Project New Stack’ Unifies Multiple Enterprise Clouds

Matt Broersma is a long standing tech freelance, who has worked for Ziff-Davis, ZDnet and other leading publications

Follow on: Google +

At Discover, HPE takes the wraps off advances in software-defined IT, terabyte-scalable persistent memory and firmware-level security

HPE has announced a project aimed at unifying organisations’ IT deployments across various kinds of infrastructure, including standard IT and clouds that may be hosted on-premises or remotely.

“Project New Stack”, which makes it easier to manage an increasingly complex IT environment, is amongst the announcements at HPE’s semi-annual Discover conference in Las Vegas this week, where the company is also showing advances in its software-defined IT offerings, more scalable persistent memory and a new firmware-level security technology.

‘Project New Stack’

The upcoming Project New Stack approach is aimed at allowing organisations to allocate resources on-demand across a hybrid IT estate, making it easier to analyse cost and data and identify underutilised resources that can be tapped into.

“In short, it will help customers optimize their right mix of hybrid IT,” stated Ric Lewis, senior vice president for HPE’s software-defined and cloud group.

Networking globe
The project’s goals are in line with those of the nearer-term technologies announced at Discover, which include software-defined tools aimed at improving infrastructure flexibility.

HPE calls its approach “composable infrastructure”, which uses OneView management software to treat server hardware and storage as fluid resources that can be deployed or “composed” as needed by software commands.

The company faces stiff competition in the area, in particular from Dell-EMC.

At the conference HPE is announcing OneView 3.1, which introduces features including intelligent system tuning, which allows systems to boost processor performance for particular workloads while ensuring data isn’t lost with a technique called jitter smoothing.

The new software includes more than a dozen present workload settings for applications such as low latency, graphics processing, web e-commerce and virtualised power-efficient workloads.

Terabyte-scalable persistent memory

HPE said its new Scalable Persistent Memory expands on previous offerings by scaling to terabyte capacity, making it usable for large-scale, high-performance systems such as the Hekaton in-memory database and the systems used by high-frequency traders.

Hekaton, also called SQL Server In-Memory OLTP, is an in-memory database for online transaction processing (OLTP) workloads built into Microsoft SQL Server.

HPE also said it will expand the capacity of the persistent memory it offers on NVDIMM modules from 8GB to 16GB.

NetworksPersistent memory offers the performance of DRAM with the persistence of flash, meaning in-memory data is retained in the case of a server crash, for instance.

Use of the memory can reduce overall system costs by improving processor performance and thereby reducing the number of cores required, HPE said.

Firmware security

At the conference HPE announced its silicon-based security, which creates a “fingerprint” for firmware and prevents it from executing if it’s been tampered with by hackers.

The offering, which relies on building cryptographic algorithms and custom code into Integrated Lights Out (iLO) firmware chips at HPE’s own factories, is intended to combat the growing problem of attacks targeting firmware.

The offering builds upon the fact that HPE has complete control over its own custom-made iLO silicon chips and server-essential firmware.

Thee ProLiant Gen10 servers introduced at the event include other security features as well, such as the behaviour-analytics security technology acquired with the purchase of Niara earlier this year.

The Niara technology uses machine learning and big-data analytics to detect security threats including anomalous user behaviour.

The Gen10 platform offers the Commercial National Security Algorithm Suite (CSNA) promoted by the National Security Agency (NSA) and the Federal Information Processing Standard (FIPS) 140-2, used by the US government to approve cryptographic modules.

While the platform was announced at Discover it won’t be officially launched until this summer, meaning details such as processor information aren’t yet available.

The servers are expected to use Intel’s Xeon chips with the upcoming Skylake architecture.

How well do you know the cloud? Try our quiz!