Increased Uptime Reduces TCO for Power Systems Clients
The 2020 ITIC Global Reliability Survey shows the value of POWER9.
By Laura DiDio09/02/2020
Every minute that an organization’s servers are down represents countless customers who take their business elsewhere. In the digital age of interconnected systems, organizations do business nonstop, irrespective of time zones or geographic location. Employees and customers alike expect continuous and secure data access. High availability ensures uninterrupted productivity, supports the business’ bottom line, strengthens security and compliance, and mitigates risk.
According to the 2020 Hourly Cost of Downtime1 survey, conducted by Information Technology Intelligence Consulting (ITIC), the cost of one hour of downtime now exceeds $100,000 for 98% of corporate enterprises. Moreover, 40% of large enterprises say the price tag for a one-hour outage can top $1 million or more. Specifically, one minute of hourly downtime calculated at $100,000 for a single server running a mission-critical application is $1,667. This makes high reliability and uninterrupted data access imperative.
1. Information Technology Intelligence Consulting Corp. (ITIC) 2020 Hourly Cost of Downtime Survey
Top Reliability Rankings
The IBM Z® and the IBM Power Systems™ platforms scored the highest reliability and availability rankings among all mainstream server platforms in ITIC’s 2020 Global Server Hardware and Server OS Reliability survey. This is the 12th straight year the IBM servers topped the rankings.
ITIC polled 1,000 businesses from January through March 2020. All categories of businesses were represented in the survey respondent pool: 27% were small/midsized (SMB) firms with up to 200 users; 28% came from the small/midsized enterprise (SME) sector with 201 to 1,000 users; and 45% were large enterprises with over 1,000 users. No sponsorships were accepted.
Depending on specific configuration and deployment scenarios, IBM Power Systems servers recorded 1.54 minutes of per server/per annum downtime due to any inherent flaws or bugs in the underlying hardware. This is the platform’s best reliability showing to date.
ITIC’s latest survey data polled 1,200 global enterprises from January through March 2020. The study compares and analyzes the reliability and availability of 18 mainstream server platforms and one dozen OS distributions. To obtain the most accurate and objective results, ITIC accepts no vendor sponsorship.
The survey results reveal that IBM Power Systems running Linux* deliver up to 28x more uptime and availability than the least dependable unbranded “white box” servers. And the superior uptime of Power Systems hardware makes the platform up to 34x more economical and cost effective than the least stable competing server.
IBM Power Systems servers once again ranked first or tied for first place in every reliability, performance, manageability, security and client satisfaction category in ITIC’s 2020 Global Server Hardware Server OS Reliability survey. Among the most notable achievements:
Least amount of unplanned downtime over four hours: Only 1% of Power Systems servers experienced unplanned server downtime of more than four hours due to inherent system or component parts failures
Highest server availability ratings: When server outages did occur, it typically took IT managers 10 minutes or less to get the systems back online with no data loss
Robust hardware and components: Less than 1% of clients experienced problems or failures within the first year of usage
Rock-solid security capabilities: Only 4% of Power Systems clients experienced downtime as a result of a security hack or data breach within the last 12 months
As Figure 1, (above) illustrates, IBM Z and IBM Power Systems recorded the lowest per server/per minute monetary costs among all mainstream platforms attributed to inherent flaws in the hardware. When businesses estimate hourly downtime costs at $100,000, the price tag for the 1.54 minutes of per server/per annum downtime due to server problems for the Power Systems platform is merely $2,567. In contrast, the least reliable white box server experienced 53 minutes of downtime per server/per annum, adding up to $88,351 in annual downtime costs due to hardware flaws.
A price of $100,000 for one hour of downtime for a single server appears relatively inexpensive. However, that equates to $1,670 per minute/per server. And that $100,000 hourly cost of downtime statistic is relatively conservative for enterprises. ITIC’s latest hourly cost of downtime statistics indicate 88% of respondents find downtime costs exceed $300,000; and 40% of those polled indicate that hourly downtime costs top $1 million or higher.
For large enterprises in specific sectors (mainly banking, energy, government, healthcare, retail, transportation and utilities) the cost of a more severe outage can exceed $1 million or more per hour, which breaks down to $16,700 per server/per minute.
Downtime costs increase exponentially based on the number of affected servers and the enterprise’s estimate on the hourly cost of downtime. The potential monetary and data losses underscore the need for very high reliability of 99.999% or higher (Figure 2, below).
“What enterprises get for their money is a powerhouse scale-out server that’s purpose-built to support compute-intensive workloads and applications that are fine-tuned for today’s in-memory databases and multicloud environments.”
Reliability Lowers TCO
Hourly downtime costs present a compelling example of how supposedly inexpensive servers can actually cost businesses more in service interruptions and lost productivity.
“IBM POWER9™ servers are very economical in terms of offering a low cost of acquisition and a very low TCO throughout the entire product lifecycle,” says Rob Enderle, principal analyst of the Enderle Group in Bend, Oregon. “What enterprises get for their money is a powerhouse scale-out server that’s purpose-built to support compute-intensive workloads and applications that are fine-tuned for today’s in-memory databases and multicloud environments.”
“POWER9 servers provide organizations with up to 1.5x speed and performance of POWER8® servers. Equally as crucial is that IBM builds security into every layer of the POWER9 stack, starting at the chip level,” Enderle says.
Delivering Greater Value for Enterprises
Server reliability and security capabilities are paramount for businesses. This is especially true in today’s business environment with largely remote workforces.
In this year’s survey, POWER9 servers recorded 1.54 minutes of unplanned downtime per server. That’s a decrease of 21 seconds from the 1.75 minutes of per server unplanned downtime registered in ITIC’s 2019 survey. That additional 21 seconds of uptime equates to a cost savings of $350 per server for an organization that estimates one hour of downtime to cost their firm $100,000.
The economies of scale are even greater for large enterprises that estimate higher hourly downtime costs. Using the $300,000 hourly downtime metric, the cost of one minute of per server/per minute downtime for an IBM Power Systems server declined from $8,747 (based on 1.75 minutes of per server downtime) in the ITIC 2019 Global Server Hardware, Server OS Reliability Survey to $7,701 (based on 1.54 minutes of per server downtime) in the 2020 poll. That’s a cost savings of $1,046 for a single server. The savings quickly add up in enterprises with hundreds of servers or cloud deployments with virtual servers.
Enterprise clients likewise praise IBM for its technical service and support. “I rarely need to phone IBM technical support for anything beyond a simple support call, but when I do need to escalate a call, I get prompt attention from very knowledgeable engineers that are able to get to the root of the issue and resolve it quickly,” says an IT manager at a healthcare organization based in Louisiana.
More Than Speeds and Feeds
In today’s business environment, the performance of an IT infrastructure platform is quantified by more than speeds and feeds; it’s measured by reliability and cost savings as well. Learn more about the advantages of the Power Systems platform.
Laura DiDio is principal analyst at ITIC, a Boston-based research and consulting firm.
Post a Comment
Note: Comments are moderated and will not appear until approvedcomments powered by Disqus