Story image

Focus on quality over quantity with server deployment, says exec

02 Feb 2017

Server memory is a component that’s either sufficient or insufficient and you likely don’t even think about RAM because other problems consume your attention.

That’s according to Crucial, global brand of Micron Technology located in both Australia and New Zealand.

The organisation adds that when you have insufficient memory, your servers and your organisation’s productivity slow to a crawl because DRAM feeds your CPUs.

Michael Moreland, marketing manager for Crucial Server DRAM, explains that commissioning a recent survey enabled them see and hear what challenges IT pros are experiencing in their data centres.

“Based on what they’ve said, it’s clear that they believe server memory is critical for improving system performance,” he says.

“Adding more memory to servers improves CPU performance and efficiency, which ultimately helps alleviate the top five workload constraints they mentioned: limited budget, unexpected or unpredictable workload demands, limited floor space, rapid growth in user base, and power or cooling costs.”

The survey asked organisations what the top challenges they were currently facing when overcoming server workload constraints were.

There were 353 respondents selected by Spiceworks that required to have purchase influence in their organisation and were required to have at least 30 physical servers and be using virtualisation software.

Overall, 23 industries were represented (ranging from technology to energy to manufacturing) and 74% of respondents were running 100 or more physical servers, with 41% running over 200 boxes.

“When we surveyed the 350-plus IT managers from around the world, they listed a limited budget as their top workload constraint. Making the most of scarce resources is a hallmark of modern IT, and that’s why it’s critical to keep the total cost of ownership (TCO) down,” adds Moreland.

“That’s where adding server memory can help – maxing out a server’s memory provides fuel for the CPU to run optimally, allowing you to use fewer servers to accomplish more. With each server functioning more efficiently, it limits the power, cooling, and burdensome licensing costs that come with having more servers in your server room.”

Moreland explains that memory actually helps maximize limited floor space and that scaling up almost always involves increasing a server’s installed memory capacity to get as much out of the box and feed as many VMs as possible

“Virtualised applications are heavily dependent on active data, and when there’s a spike in workload activity, available server memory resources are depleted and QoS drops. Filling your servers to their maximum RAM capacities reduces the strain on the systems when activity gets intense,” he says.

“Limited square footage in your server room places a premium on making the most of each individual server. When you don’t have the physical space to scale out, scaling up by maxing out the memory of each server can ultimately match the performance of multiple half-full ones.”

He explains that memory is like fuel for your CPUs – as long as they have enough of it, they’re OK.

“But there’s a significant difference between having enough RAM and truly improving workload efficiency. With just enough RAM, you’re certainly able to run applications, but with the maximum installed memory capacity, you’re often able to use fewer servers to get more done at a lower total TCO,” says Moreland.

“In the case of servers, more isn’t always better, especially when they can be maxed-out with memory and meet or surpass the same level of performance of half-full servers,” he adds.

“Focus on quality, not quantity, when it comes to your server deployment and reduce your power, cooling, and license costs.”

Why 'right to repair' legislation could be a new lease on life for broken devices
“These companies are profiting at the expense of our environment and our pocketbooks as we become a throw-away society that discards over 6 million tonnes of electronics every year.”
Attacks targeting Cisco Webex extension explode in popularity - WatchGuard
WatchGuard's Internet Security Report for Q4 2018 also finds growing use of a new sextortion phishing malware customised to individual victims.
SAS partners with NVIDIA on deep learning and computer vision
“By partnering with NVIDIA, we combine our strengths to augment human intelligence and realise the true potential of AI.” 
Why businesses must embrace automation to ensure success
“For many younger workers, the traditional view of a steady job at one company, perhaps for life, simply doesn’t reflect reality."
Dropbox invests in hosting data inside Australia
Global collaboration platform Dropbox has announced it will now host Australian customer files onshore to support its growing base in the country.
TYAN unveils new inference-optimised GPU platforms with NVIDIA T4 accelerators
“TYAN servers with NVIDIA T4 GPUs are designed to excel at all accelerated workloads, including machine learning, deep learning, and virtual desktops.”
Worldwide spending on security to reach $103.1bil in 2019 - IDC
Managed security services will be the largest technology category in 2019.
How Cognata and NVIDIA enable autonomous vehicle simulation
“Cognata and NVIDIA are creating a robust solution that will efficiently and safely accelerate autonomous vehicles’ market entry."