Network World Data Center

Author Archives: Network World Data Center

HPE launches supercomputers for the enterprise

Supercomputers are super expensive, but Hewlett Packard Enterprise has announced plans to make supercomputing accessible for more enterprises by offering scaled down, more affordable versions of its Cray supercomputers.The new portfolio includes HPE Cray EX and HPE Cray XD supercomputers, which are based on the Frontier exascale supercomputer at Oak Ridge National Labs. These servers come with the full array of hardware, including compute, accelerated compute, interconnect, storage, software, and flexible power and cooling options.To read this article in full, please click here

World’s fastest supercomputer is still Frontier, 2.5X faster than #2

Frontier, which became the first exascale supercomputer in June and ranked number one among the fastest in the world, retained that title in the new TOP500 semiannual list of the world’s fastest.Without any increase in its speed—1.102EFLOP/s—Frontier still managed to score 2.5 times faster that the number two finisher, Fugaku, which also came in second in the June rankings. An exascale computer is one that can perform 1018 (one quintillion) floating point operations per second (1 exaFLOP/s).Despite doubling its maximum speed since it was ranked number three in June, the Lumi supercomputer remained in third-place.There was just one new member of the top-ten list, and that was Leonardo, which came in fourth after finishing a distant 150th in the TOP500 rankings in June.To read this article in full, please click here

Looking at user login time with the ac command

While not a very well known Linux command, ac can provide very useful stats on user login time. In its simplest form, it will show you how much time users have spent on the system in the time period covered by the wtmp file. All you have to type is “ac” to get a figure showing overall login time for all users.$ ac total 8360.60 The figure above indicates that users spent a total of 8,360.6 hours on the system. Looking at the wtmp file with the who command, we can see that the saved logins started on June 6th – a little more than 6 months earlier.$ who /var/log/wtmp | head -2 shs tty2 2022-06-06 16:00 (tty2) shs pts/1 2022-06-06 16:23 (192.168.0.12) To look at the times by user, add the -p (people) argument.To read this article in full, please click here

Intel announces CPUs and GPUs for high-performance computing

Intel has announced new processors with high-bandwidth memory (HBM) geared toward high-performance computing (HPC), supercomputing, and artificial intelligence (AI).The products are known as the Xeon CPU Max series and GPU Max series. The chips are based on existing technology; the CPU is 4th Generation Xeon Scalable, aka Sapphire Rapids, and the GPU is Ponte Vecchio, the data center version of Intel's Xe GPU technology.To read this article in full, please click here

Bash: A primer for more effective use of the Linux bash shell

Bash is not just one of the most popular shells on Linux systems, it actually predates Linux by a couple of years. An acronym for the “GNU Bourne-Again Shell”, bash not only provides a comfortable and flexible command line, it delivers a large suite of scripting tools—if/then commands, case statements, functions, etc.—that allow users to build complex and powerful scripts.This post contains a collection of articles about important aspects of bash that will help you make better use of this versatile shell.Commands vs bash builtins While Linux systems install with thousands of commands, bash also supplies a large number of “built-ins”—commands that are not sitting in the file system as separate files, but are part of bash itself. To get a list of the bash builtins, just type “help” when you’re on the bash command line. For more about built-ins, refer to “How to tell if you’re using a bash builtin”.To read this article in full, please click here

Intel is shipping the next generation of Xeon Scalable processors

After almost a year and a half of delays, Intel has begun to ship its 4th Generation Xeon Scalable processors, code-named Sapphire Rapids, to customers, and it has set January 10, 2023 as the formal launch date.The launch is a formality because, according to an Intel spokesperson, the new Xeons are already shipping to customers—OEMs—now, but it falls to those OEMs to announce their product release plans.CEO Pat Gelsinger said during the company’s earnings call last week that the company was ramping up production for launch and that he expected the new Xeons to see the fastest ramp to one million units ever.The challenge for Intel wasn’t in design, it was manufacturing. This will be the first generation of chips using Intel 7 fabrication, an advanced 10nm design that took years to get right.To read this article in full, please click here

Nvidia tests: DPUs can cut power needed by servers

The chip maker says tests of its BlueField-2 data-processing units (DPU) in servers results in significant power savings over servers that don’t use the specialized chips to offload tasks from the CPUs.The DPUs, or SmartNICs, take on certain workloads—packet routing, encryption, real-time data analysis—leaving the CPU free to process data. But Nvidia says they can also reduce power consumption.The four tests involved running similar workloads on servers with and without DPUs, and Nvidia concluded that even with the additional power draw by the DPUs, overall power consumption by the servers dropped.To read this article in full, please click here

Nvidia tests: DPUs can cut the power servers use

The chip maker says tests of its BlueField-2 data-processing units (DPU) in servers results in significant power savings over servers that don’t use the specialized chips to offload tasks from the CPUs.The DPUs, or SmartNICs, take on certain workloads—packet routing, encryption, real-time data analysis—leaving the CPU free to process data. But Nvidia says they can also reduce power consumption.The four tests involved running similar workloads on servers with and without DPUs, and Nvidia concluded that even with the additional power draw by the DPUs, overall power consumption by the servers dropped.To read this article in full, please click here

Kyndryl’s first year yields revenue challenges and a plan to make more

In the year since it spun out of IBM, Kyndryl has made a number of big strides in establishing itself as a core infrastructure service management player, but challenges remain.For example, while it has substantially expanded its partnerships and technology, its financial situation hasn’t shown a similar bump. Just this week the company reported  second quarter revenues of $4.2 billion, a year-over-year drop of 9%. The company has reported similar results in other recent quarters. “Currency and energy-costs impacts are superseding the operational progress we’re making,” Kyndryl CFO David Wyshner told Wall Street analysts during the firm’s 2Q 2023 call this week. “And while the risk of a global recession has clearly increased, we continue to see broadbase demand for digital transformation in infrastructure services.”To read this article in full, please click here

AMD posts operating loss, but solid growth for data center, embedded segments

AMD announced third quarter results this week, and while it posted a $64 million loss in terms of overall operating income—mainly due to its acquisition of Xilinx—but large gains in the company’s data center, embedded and gaming segments provided an encouraging note.Total revenue rose by 29% for the third quarter of 2022, to $5.56 billion from $4.31 billion one year ago. Gross profit also rose in year on year terms, from $2.08 billion in last year’s third quarter to $2.35 billion for the past three months. The decline in operating income was caused by much higher operating expenses, which more than doubled in the third quarter, rising from $1.14 billion a year ago to $2.42 billion in the most recent figures.To read this article in full, please click here

HPE launches 11th generation ProLiant servers

Hewlett Packard Enterprise has introduced the 11th generation of its ProLiant servers designed for a range of modern workloads, including AI, analytics, cloud-native applications, graphic-intensive applications, machine learning, Virtual Desktop Infrastructure (VDI), and virtualization.The new ProLiants have three processor options: AMD Epyc “Genoa” generation processors, Intel Xeon Scalable “Sapphire Rapids” generation processors, and Ampere Altra and Altra Max cloud-native processors.Compared to the previous server generation, the new HPE ProLiant Gen11 servers support twice as much I/O bandwidth and 33% more high-performance GPU density per server to support AI and graphic-intensive workloads than the prior generation.To read this article in full, please click here

NetApp unifies its storage offerings under a new BlueXP roof

NetApp announced Tuesday that its on-premises and cloud storage offerings are now unified under the umbrella of a single platform, called BlueXP, which serves as a control plane for each of its products and simplifies the management of enterprise storage for organizations.BlueXP—which is a free upgrade for its customers—is a reaction to the reality that more and more companies’ storage environments are hybrids these days, combining cloud and on-premises storage, according to NetApp. Businesses of almost any size that have been in operation for more than a decade or so are, more often than not, involved in digital transformation efforts that move at various paces, said company senior vice president and general manager for cloud storage Ronen Schwartz.To read this article in full, please click here

Broadcom CEO outlines what combined Broadcom and VMware might look like

There has been much teeth-gnashing, mixed with a little obfuscation and concern, about what the merged VMware and Broadcom might look like and what it will mean to customers.   On Thursday, Broadcom President and CEO Hock Tan took to his blog to offer some details about what he expects the VMware buy will mean to Broadcom and try to ease some of the concerns customers are having.One of the apprehensions for all customers is cost of products going forward. “Following the purchases of CA and Symantec, Broadcom raised prices, decreased support, and stopped investing in innovation,” Tracy Woo, senior analyst for Forrester told Network World in a recent article. “VMware customers would be wise to have an exit plan,” she cautioned.To read this article in full, please click here

Broadcom CEO: What the VMware merger will look like

There has been much teeth-gnashing, mixed with a little obfuscation and concern, about what the merged VMware and Broadcom might look like and what it will mean to customers.   Broadcom President and CEO Hock Tan has taken to his blog to offer some details about what he expects the deal will mean to Broadcom and try to ease some customer concerns.One worry: cost of products going forward. “Following the purchases of CA and Symantec, Broadcom raised prices, decreased support, and stopped investing in innovation,” Tracy Woo, senior analyst for Forrester told Network World in a recent article. “VMware customers would be wise to have an exit plan,” she cautioned.To read this article in full, please click here

Counting individual characters on Linux

Determining how many characters are in a file is easy on the Linux command line: use the ls -l command.On the other hand, if you want to get a count of how many times each character appears in your file, you’re going to need a considerably more complicated command or a script. This post covers several different options.Counting how many times each character appears in a file To count how many of each character are included in a file, you need to string together a series of commands that will consider each character and use a sort command before it counts how many of each character are included.To do that, you can use a command like this one:To read this article in full, please click here

Finding and fixing typos on Linux

If you want to check a text file for typos, Linux can help.It has a couple of tools and a number of commands that can point out the errors including aspell and enchant, and I’ll share a script that I put together recently that looks for typos using the system's words file.Using aspell aspell is very clever tool that will point out typos and make it surprisingly easy to fix them. When used to make changes to a single file, it reverses the text and background colors to highlight misspelled words. You would start it with a command like this:$ aspell check myfile If aspell detects no typos, it simply exits. Otherwise, it will open with a display that contains the file text (or just the top lines depending on the length of the file) followed by a list of suggested replacement words and, below that, a list of the commands that you can run. The first typo (or suspected typo) will be displayed with the text and background colors reversed as shown below.To read this article in full, please click here

AMD promises faster and more efficient networking, eventually

This time last year, AMD had no networking products. Today, it has three, thanks to two separate acquisitions. It's a shift that's strengthening the company’s position against competitors like Intel and Nvidia by providing a full suite of silicon, including for the enterprise.To fill out its portfolio and stay competitive, AMD had to build up its offerings of workload accelerators and its networking technologies. It's a very active market, with Nvidia (BlueField), Intel (FPGA-based smartNICs), Marvell Technology (Octeon), and Broadcom (Stingray) all competing for the smartNIC market. AMD risked being left behind.To read this article in full, please click here

IBM sales jump shows the mainframe is not dead, with hybrid cloud alive and well

At a time when most enterprises are planning cloud deployments and many are reportedly sharpening their mainframe exit strategy, IBM is seeing double-digit growth in its big iron business for the quarter ended September.The company, which declared its third quarter results on Wednesday, reported a 98% jump in revenue for its z line of mainframe computer in terms of constant currency (that is, eliminating the effect of currency fluctuations).  IBM, which buckets mainframes under its infrastructure line of business, released the z16 mainframe in April before beginning to sell it in the second quarter.To read this article in full, please click here

Startup says its chips handle HPC workloads better than GPUs

A semiconductor startup is targeting the high-performance computing (HPC), claiming that in some instances, GPUs aren’t the best fit for the task.The chip is known for now as Thunderbird but will get a formal name when it launches in Q1 of 2023, according to Doug Norton, vice president of business development at InspireSemi. Thunderbird is mounted on a PCI Express card that plugs into a server, just as GPU accelerators from Nvidia and AMD do.The Thunderbird chip contains 2,560 RISC-V cores, and there are two chips per card. GPUs also come with thousands of cores but CPUs have less than 100, except for the Ampere Altra Max, with 128 cores.To read this article in full, please click here

1 … 13 14 15 16 17 … 20