IBM, Lenovo, HPE and Huawei Servers Maintain Top Reliability Rankings; Cisco Makes Big Gains IBM, Lenovo hardware up to 24x more reliable; 28x more economical vs. least reliable White box servers
ITIC’s latest 2019 Global Server Hardware, Server OS Reliability Mid-Year Update survey results indicate that mission critical servers from IBM, Lenovo, Hewlett-Packard Enterprise (HPE) and Huawei all maintained their top positions, achieving “four to six nines” of uptime.
These findings come at a time when businesses’ demand for high reliability and continuous, uninterrupted data access is at an all-time high.
ITIC’s latest survey data finds that the most reliable mainstream server platforms – the IBM Power Systems, Lenovo ThinkSystem, Hewlett-Packard Enterprise (HPE) and Huawei KunLun deliver up to 24x more uptime and availability than the least dependable unbranded “White box” servers. Additionally, the superior uptime of the above top ranked mission critical hardware makes them up to 28x more economical and cost effective than the least stable White box servers.
High end mission critical server distributions from IBM, Lenovo, HPE and Huawei each recorded just under or approximately two (2) minutes of per server, per annum unplanned downtime due to inherent flaws in the underlying hardware or component parts (See Exhibit 1). By contrast, the least consistent hardware – unbranded White box servers – averaged 49 minutes of unplanned per server, per annum downtime due to problems or failures with the server or its components (e.g. hard drive, memory, cooling systems etc.).
Server hardware reliability directly impacts ongoing daily business transactions and productivity. There are immediate monetary costs associated with server outages of even a few minutes. The disparity in the annual downtime cost comparisons among the top performing and the least reliable server hardware, is eye-opening.
A single hour of downtime calculated at $100,000 equates to $1,667 per server/per minute.
Corporations that deploy the most highly reliable servers: the IBM Power Systems; Lenovo ThinkSystem; HPE Superdome and Huawei KunLun (in that order) that averaged just under or about two (2) minutes of unplanned per server downtime, potentially could expect to lose approximately $3,000 per server/per minute for an hour of downtime calculated at a very conservative $100,000. By contrast, businesses that deploy the least reliable unbranded White box servers which recorded 49 minutes of unplanned per server annual downtime due to the inherent hardware instability could potentially lose $81,683 based on hourly downtime costs of $100,000. The superior economics of the most reliable versus least reliable servers is even more apparent for businesses that estimate or calculate hourly downtime losses of $300,000; $500,000 or $1,000,000 or higher.
Servers are the bedrock upon which the entire network infrastructure and extended network ecosystem rests. When servers fail, data access is denied. Business stops. Productivity ceases. Revenue suffers.
Some 86% of organizations now require a minimum 99.99% reliability for their firms’ server hardware, operating systems and main line-of-business applications to ensure productivity and deliver uninterrupted data access. High reliability and availability also safeguards the corporation’s daily operations, business processes and revenue stream.
IBM Z, IBM POWER, Lenovo ThinkSystem, HPE Integrity and Huawei KunLun Servers Maintain Highest Uptime Rankings
The latest ITIC 2019 Reliability Mid-Year Update survey polled over 800 corporations from July through early September. The study compared the reliability and availability of over one dozen of the most widely deployed mainstream server platforms and one dozen operating system (OS) distributions. ITIC’s latest study updated a select subset of the survey questions from its annual 2019 Global Server Hardware, Server OS Reliability poll. The poll also tracked the impact of pivotal issues like security, human error, software flaws and aging server hardware on corporate server reliability. To obtain the most accurate and unbiased results, ITIC accepted no vendor sponsorship.
Organizations conduct business 24 x 7 irrespective of time or location, 365 days a year. Corporations continue to expand their operations into the cloud and connect people, applications and devices via the Internet of Things (IoT). Applications like Analytics, AI and Business Intelligence (BI) are complex and compute intensive. They place greater demands on the server hardware. The corporate workforce is increasingly mobile. Users access data from myriad devices. Companies require fast, efficient processing and throughput. It must be secure by design, secure in use, secure in transmission and secure at rest.
To reiterate, all of the high end mission critical servers maintained their top ranked positions from ITIC’s earlier 2019 Global Server Hardware Server OS Reliability Survey published in the first calendar quarter of this year.
The IBM Z mainframe system is in a class of its own, delivering true fault tolerance – “six nines” – 99.9999% uptime to 89% of enterprise users. It delivered imperceptible instances of inherent server failure – 0.74 seconds per/server due to any inherent flaws in the server hardware.
Among the mainstream server distributions, IBM’s Power Systems topped the poll, registering a record low of 1.75 minutes per server downtime followed very closely by the Lenovo Think System servers with 1.88 minutes of per server downtime due to any flaws in the server hardware. Hewlett Packard Enterprise’s (HPE) Superdome X, Huawei’s KunLun FusionServer x86 platforms each recorded 2 minutes of server downtime due to any underlying problems with the server hardware.
Each of these distributions delivered a solid “five nines,” 99.999% of inherent hardware reliability. These leading edge server platforms experienced minimal amounts of unplanned downtime due to flaws in the server hardware or any of its component parts.
ITIC’s 2019 Reliability Mid-Year Update Survey did deliver a few surprises. Cisco Systems’ Unified Computing System (UCS) servers – which are frequently deployed at the network edge – showed a marked improvement in reliability. The Cisco UCS servers reduced per server/per annum downtime by nearly 50% from the 4.1 minutes in ITIC’s prior first quarter reliability survey to 2.3 minutes in the latest poll.
ITIC’s Mid-Year Update survey for the first time also included uptime statistics for Inspur Systems, headquartered in Jinan, China as one of the top five server vendors worldwide in terms of shipments. Inspur server offerings scored in the middle range of hardware platforms with 9.1 minutes of unplanned downtime.
Metrics of three, four and five nines of uptime – 99.9%, 99.99% and 99.999%, – equate to 8.76 hours; 4.38 hours, 52.56 and 5.26 minutes of per server/per annum downtime, respectively.