IT seems to be growing again. Gartner forecasts worldwide IT spending will increase 4.5% this year to $3.68 trillion, driven by artificial intelligence, big data analytics, blockchain technology, and the IoT.
Digital transformations are fashionable. You won’t find an enterprise that isn’t leveraging some combination of cloud, analytics, artificial intelligence and machine learning to better serve customers or streamline operations. But here’s a hard truth about digital transformations: Many are failing outright or are in danger of failing. Typical reasons for failing are not understanding what is digital transformation (different people understand it differently), lack of CEO sponsorship, talent deficiency, resistance to change. Usually a technology-first approach to digital transformation is a recipe for disaster. Truing to just push trough technically unfeasible transformation idea is another way to fail.
The digital era requires businesses to move with speed, and that is causing IT organizations to rethink how they work. A lot of IT is moving off premises to SaaS providers and the public cloud. Research outfit 451 standout finding was that 60 per cent of the surveyed enterprises say they will run the majority of their IT outside the confines of enterprise data centres by the end of 2019. From cost containment to hybrid strategies, CIOs are getting more creative in taking advantage of the latest offerings and the cloud’s economies of scale.
In 2018 there seems to be a growing Software Engineering Talent Shortage in both quantity and quality. For the past nine years, software engineers have been at the top of the hardest to fill jobs in the United States. And same applies to many other countries including Finland. Forrester projects that firms will pay 20% above market for quality engineering talent in 2018. Particularly in-demand skills are data scientists, high-end software developers and information security analysts. There is real need for well-studied, experienced engineers with a formal and deep understanding of software engineering. Recruiting and retaining tech talent remains IT’s biggest challenge today. Most CIOs are migrating applications to public cloud services, offloading operations and maintenance of computing, storage and other capabilities so they can reallocate staff to focus on what’s strategic to their business.
The enterprise no longer is at the center of the IT universe. It seems that reports of the PC’s demise have been greatly exaggerated and the long and painful decline in PC sales of the last half-decade as tailed off, at least momentarily. As the sales of smartphones and tablets have risen, consumers had not stopped using PCs, but merely replaced them less often. FT reports that PC is set to stage a comeback in 2018, after the rise of smartphones sent sales of desktop and laptop computers into decline in recent years. If that does not happen, then PC market could return to growth in 2019. But the end result is that PC is no longer seen as the biggest growth driver for chip makers. An extreme economic shift has chipmakers focused on hyperscale clouds.
Microservices are talked about a lot. Software built using microservices is easier to deliver and maintain than the big and brittle architectures or old; these were difficult to scale and might take years to build and deliver. Microservices are small and self-contained, so therefore easy to wrap up in a virtual machine or a container (but don’t have to live in containers). Public cloud providers increasingly differentiate themselves through the features and services they provide. But it turns out that microservices are far from being one-size-fit-for-all silver bullet for IT challenges.
Containers will try to make break-trough again in 2018. Year 2017 was supposed to be the year of containers! It wasn’t? Oops. Maybe year 2018 is better. Immature tech still has a bunch of growing up to do. Linux Foundation’s Open Containers Initiative (OCI) finally dropped two specifications that standardise how containers operate at a low level. The needle in 2018 will move towards containers running separately from VMs, or entirely in place of VMs. Kubernates gains traction. It seems that the containers are still at the point where the enterprise is waiting to embrace them.
Serverless will be talked about. Serverless computing is a cloud computing execution model in which the cloud provider dynamically manages the allocation of machine resources. Serverless architectures refer to applications that significantly depend on third-party services (knows as Backend as a Service or “BaaS”) or on custom code that’s run in ephemeral containers (Function as a Service or “FaaS”), the best known vendor host of which currently is AWS Lambda.
Automation is what everybody with many computers wants. Infrastructure automation creates and destroys basic IT resources such as compute instances, storage, networking, DNS, and so forth. Security automation helps keeping systems secure. It bosses want to create self-driving private clouds. The journey to self-driving clouds needs to be gradual. The vision of the self-driving cloud makes sense, but the task of getting from here to there can seem daunting. DevOps automation with customer control: Automatic installation and configuration, Integration that brings together AWS and VMWare, workflows migration controlled by users, Self-service provisioning based on templates defined by users, Advanced machine learning to automate processes, and Automated upgrades.
Linux is center of many cloud operations: Google and Facebook started building their own gear and loading it with their own software. Google has it’s own Linux called gLinux. Facebook networking uses Linux-based FBOSS operating system. Even Microsoft has developed its own Linux for cloud operations. Software-defined networking (SDN) is a very fine idea.
Memory business boomed in 2017 for both NAND and DRAM. The drivers for DRAM are smartphones and servers. Solid-state drives (SSDs) and smartphones are fueling the demand for NAND. NAND Market Expected to Cool in Q1 from the crazy year 2017, but it is still growing well because there is increasing demand. Memory — particular DRAM — was largely considered a commodity business.
Lots of 3D NAND will go to solid state drives in 2018. IDC forecasts strong growth for the solid-state drive (SSD) industry as it transitions to 3D NAND. SSD industry revenue is expected to reach $33.6 billion in 2021, growing at a CAGR of 14.8%. Sizes of memory chips increase as number of layer in 3D NAND are added. The traditional mechanical hard disk based on magnetic storage is in hard place in competition, as the speed of flash-based SSDs is so superior
There is search for faster memory because modern computers, especially data-center servers that skew heavily toward in-memory databases, data-intensive analytics, and increasingly toward machine-learning and deep-neural-network training functions, depend on large amounts of high-speed, high capacity memory to keep the wheels turning. The memory speed has not increased as fast as the capacity. The access bandwidth of DRAM-based computer memory has improved by a factor of 20x over the past two decades. Capacity increased 128x during the same period. For year 2018 DRAM remains a near-universal choice when performance is the priority. There is search going on for a viable replacement for DRAM. Whether it’s STT-RAM or phase-change memory or resistive RAM, none of them can match the speed or endurance of DRAM.
PCI Express 4.0 is ramping up. PCI-standards consortium PCI-SIG (Special Interest Group) has ratified and released specifications for PCIe 4.0 Specification Version 1. Doubling PCIe 3.0’s 8 GT/s (~1 GB/s) of bandwidth per lane, PCIe 4.0 offers a transfer rate of 16 GT/s. The newest version of PCI Express will start appearing on motherboards soon. PCI-SIG has targeted Q2 2019 for releasing the finalized PCIe 5.0 specification, so PCIe 4.0 won’t be quite as long-lived as PCIe 3.0 has been. So we’ll See PCIe 4.0 this year in use and PCIe 5.0 in 2019.
USB type C is on the way to becoming the most common PC and peripheral interface. The USB C connector has become faster more commonplace than any other earlier interface. USB C is very common on smartphones, but the interface is also widespread on laptops. Sure, it will take some time before it is the most common. In 2021, the C-type USB connector has almost five billion units, IHS estimates.
It seems that the after-shocks of Meltdown/Spectre vulnerabilities on processors will be haunting us for quite long time this year. It is now three weeks since The Register revealed the chip design flaws that Google later confirmed and the world still awaits certainty about what it will take to get over the silicon slip-ups. Last pieces of farce has been that Intel Halts Spectre, Meltdown CPU Patches Over Unstable Code and Linux creator Linus Torvalds criticises Intel’s ‘garbage’ patches. Computer security will not be the same after all this has been sorted out.
What’s Next With Computing? IBM discusses AI, neural nets and quantum computing. Many can agree that those technologies will be important. Public cloud providers increasingly provide sophisticated flavours of data analysis and increasingly Machine Learning (ML) and Artificial Intelligence (AI). Central Banks Are Using Big Data to Help Shape Policy. Over the past few years, machine learning (ML) has evolved from an interesting new approach that allows computers to beat champions at chess and Go, into one that is touted as a panacea for almost everything. 2018 will be the start of what could be a longstanding battle between chipmakers to determine who creates the hardware that artificial intelligence lives on.
ARM processor based PCs are coming. As Microsoft and Qualcomm jointly announced in early December that the first Windows 10 notebooks with ARM-based Snapdragon 835 processors will be officially launched in early 2018, there will be more and more PCs with ARM processor architecture hitting the market. Digitimes Research expects that ARM-based models may dominate lower-end PC market, but don’t hold your breath on this. It is rumoured that “wireless LTE connectivity” function will be incorporated into all the entry-level Window 10 notebooks with ARM processors, branded by Microsoft as the “always-connected devices.” HP and Asustek have released some ARM-based notebooks with Windows 10S.
Sources:
Ohjelmistoalan osaajapula pahenee – kasvu jatkuu
PC market set to return to growth in 2018
PC market could return to growth in 2019
PC sales grow for the first time in five years
PCI-SIG Finalizes and Releases PCIe 4.0, Version 1 Specification: 2x PCIe Bandwidth and More
Hot Chips 2017: We’ll See PCIe 4.0 This Year, PCIe 5.0 In 2019
Outsourcing remains strategic in the digital era
8 hot IT hiring trends — and 8 going cold
EDA Challenges Machine Learning
The Battle of AI Processors Begins in 2018
How to create self-driving private clouds
ZeroStack Lays Out Vision for Five-Step Journey to Self-Driving Cloud
2017 – the year of containers! It wasn’t? Oops. Maybe next year
2018′s Software Engineering Talent Shortage— It’s quality, not just quantity
How Central Banks Are Using Big Data to Help Shape Policy
Digitimes Research: ARM-based models may dominate lower-end PC market
Intel Halts Spectre, Meltdown CPU Patches Over Unstable Code
Spectre and Meltdown: Linux creator Linus Torvalds criticises Intel’s ‘garbage’ patches
Meltdown/Spectre week three: World still knee-deep in something nasty
What’s Next With Computing? IBM discusses AI, neural nets and quantum computing.
PCI Express 4.0 as Fast As Possible
Microsoft has developed its own Linux!
Microsoft Built Its Own Linux Because Everyone Else Did
Facebook has built its own switch. And it looks a lot like a server
Googlella on oma sisäinen linux
Is the writing on the wall for on-premises IT? This survey seems to say so
12 reasons why digital transformations fail
7 habits of highly effective digital transformations
857 Comments
Tomi Engdahl says:
The application software market to over $ 30 billion
Digitalisation of business, Internet, analytics and the use of artificial intelligence are now rapidly expanding the application software market. Last year, the market grew 12 percent to $ 28.4 billion and this year go beyond the $ 30 billion limit, estimates Gartner.
According to Gartner’s figures, last year, IBM made over $ 6.1 billion in net sales. Oracle is equally clear on the market with its $ 3.1 billion sales.
Third, the largest application developer is Salesforce, a sales and customer management solution that has invested heavily in artificial intelligence.
Next on the list came from Microsoft ,which sold apps for cloud services by $ 1.3 billion
Next Amazon, which sold apps for cloud services by $ 839 million last year.
Source: http://www.etn.fi/index.php/13-news/8116-sovellusohjelmistojen-markkinat-yli-30-miljardiin-dollariin
Tomi Engdahl says:
The servers are now going to sell at a wild rate
Companies have renewed their own data center machines while building more capacity in the clouds. The trend is clearly visible in the service trade, which, according to Gartner, accelerated in the first quarter of the year to one third higher than last year.
In terms of money, servers were sold in January-March by nearly $ 16.7 billion. At the same time, server volume sales increased by more than three million machine lines, which is 17.3 percent higher than a year earlier.
There is a regional difference in the market. In North America, growth was 34 percent and in Asia 47.8 percent. In Europe, server sales grew by 32.1 per cent, although the number of equipment deliveries declined by two per cent. At least in Europe, nowadays, more expensive and more powerful servers are being purchased.
Of the manufacturers, Dell grew 51.4 percent in the first quarter. Its net sales increased to $ 3.6 billion, and it went to the manufacturers list at the end of the HPE. Dell’s market share is now 21.5 percent and HPE’s 19.9 percent.
x86 sales continued to grow. In the early part of the year, growth was 35.7 per cent. At the same time, the decline in sales of UNIX machines continued
Source: http://www.etn.fi/index.php/13-news/8123-palvelimia-menee-nyt-kaupaksi-hurjaa-vauhtia
Tomi Engdahl says:
Future of the internet could depend on submarine datacenters
https://mashable.com/2018/06/11/microsoft-underwater-datacenter/?europe=true#PUhTp3a5vsqA
Tomi Engdahl says:
Microsoft tests underwater datacenter in the Scottish sea
https://mashable.com/2018/06/06/microsoft-underwater-datacenter-orkney-islands/?europe=true#O79IrQqucsqx
On Tuesday, Microsoft announced that it had deployed the second phase of Project Natick: a “moonshot” research project to test the viability of underwater datacenters. It placed the datacenter off the coast of the Orkney Islands, which are at the northernmost tip of Scotland. Microsoft will monitor the device for the next 12 months to see how what was an idea on paper just four years ago functions in the real world.
First, Microsoft placed the datacenter next to the European Marine Energy Center. The center captures energy generated by the movement of seawater through tidal turbines and wave energy converters. That renewable energy powers the datacenter — no fossil fuels required.
Second, placement under water provides a naturally cool environment. Microsoft brought on an experienced marine engineering firm for the project. It found that it could use many of the same techniques that maintain the climate of submarine — such as pipes that channel seawater around the vessel — to keep the datacenter cool, too.
Tomi Engdahl says:
Inside the Summit Supercomputer
https://www.eetimes.com/author.asp?section_id=36&doc_id=1333374
Interconnects have long been key to supercomputer performance, but the recent Summit system bucks the trend to proprietary designs by leveraging standards and off-the-shelf components.
On June 8, America’s Summit supercomputer was announced as up and running at an impressive 200 petaflops maximum theoretical performance. (Public benchmarks are expected later this month.) It became the fastest system in the world, retaking the lead from China which had claimed dominance for several years.
Competition remains stuff. China has multiple exaflop projects expected to be running a year or more before the U.S. has a system at that level.
The Summit supercomputer at the Department of Energy’s Oak Ridge National Laboratory consists of 4,608 compute nodes, containing a total of 9,216 IBM Power 9 processors and 27,648 Nvidia Tesla V100 GPU modules. The Power 9 and V100 chips talk over NVLink, a high speed, high bandwidth mesh interconnect.
The proprietary NVLink is fine for building individual high-performance compute nodes. But scaling thousands of nodes into a high-performance cluster requires a state-of-the-art network. Summit’s full fat-tree network is built using InfiniBand EDR cards from Mellanox.
Competitors will find that Summit’s network performance and overall performance scalability would not be possible without PCI Express 4.0. The supercomputer is the first public high-performance cluster to support PCIe 4.0 at a scale of thousands of nodes.
Summit is comprised of 256 59kW racks of compute nodes and 40 38kW racks of IBM Spectrum Scale storage. Each rack includes two top of rack switches. Eighteen racks of core switches implement the fat-tree network between compute and storage racks. Overall, the massive system claims a cross-sectional network bandwidth of nearly a petabit/second.
Tomi Engdahl says:
Steve Lohr / New York Times:
US now has the world’s fastest supercomputer, capable of 200 quadrillion calculations per second, called Summit, which cost $200M to develop
https://www.nytimes.com/2018/06/08/technology/supercomputer-china-us.html
Tomi Engdahl says:
Intel confirms it’ll release GPUs in 2020
They sell like hot cakes so why wouldn’t Chipzilla want in?
https://www.theregister.co.uk/2018/06/13/intel_gpus_2020/
Intel has confirmed it will start to sell discrete GPUs in the year 2020.
News of Chipzilla’s plans appeared in a post by analyst Ryan Shrout, who said that Intel CEO Brian Krasnich last week told an analyst event about the company’s plans.
Intel confirmed Shrout’s piece, telling us “We’re pleased to confirm our first discrete GPU is coming in 2020. As we’ve previously stated, our intent is to expand our leading position in integrated graphics for the PC market with high-end discrete graphics solutions for a broad range of computing segments.”
Details remain scanty, but a move into GPUs would be entirely sensible for a few reasons.
Firstly, they’re selling like hot cakes and demand is so high they can sometimes be hard to find. Intel would be mad not to offer buyers another source given AMD and NVIDIA can hardly shove their kit out of a FAB fast enough to keep up with demand.
Second, Intel recently discontinued its Xeon Phi co-processor line. GPUs would be a more-than-handy replacement for the Phi.
Third, GPUs are Just Becoming The Way Stuff Gets Done, especially in markets like AI, HPC and visualization in which Intel already plays.
Lastly, Intel needs new revenue sources. It missed mobile and its IoT efforts have not gone well.
Meanwhile PC sales are slumping, Optane isn’t yet a big money-spinner and AMD is nibbling at its data centre business .
Tomi Engdahl says:
Qualcomm to keep server CPUs but avoids head-on Intel battle
Plans to target greenfield hyperscalers, skip boring old servers
https://www.theregister.co.uk/2018/06/13/qualcomm_to_keep_server_cpus_but_avoids_headon_intel_battle/
Qualcomm president Cristiano Amon has said the company has no plans to dispose of its Arm-powered server CPU unit.
Speaking to Reuters, Amon rubbished whispers that Qualcomm wants out of the server CPU business.
But he also explained that the company doesn’t really want in, either, because Qualcomm won’t go head-to-head with Intel for general purpose servers.
“It’s very clear to us that the ARM opportunity is focused on a few players where you don’t have the software x86 barrier to entry,” Amon told the newswire. The company will therefore target hyperscalers who roll their own stacks and haven’t already done so on Intel or AMD silicon.
Such buyers could be huge customers: internet giants buy vast numbers of servers. Telcos building for 5G and network function virtualization will do likewise. Selling such customers one architecture for all the pieces of the data centres and networks they’ll need could help such customers differentiate. Selling direct to operators could also give Qualcomm a chance in deals like AT&T’s plan to replace 100,000 routers with white boxes.
Tomi Engdahl says:
Stephen Nellis / Reuters:
Qualcomm president says company has no plans to sell its server chip unit but is making staff reductions and folding it into Qualcomm’s CDMA Technologies unit
Qualcomm will not exit data center business, chip chief says
https://www.reuters.com/article/us-qualcomm-chips/qualcomm-will-not-exit-data-center-business-chip-chief-says-idUSKBN1J902Z
Tomi Engdahl says:
“Enterprise developers seem to be mostly OK — if not somewhat wary — with the deal; the rest of open source, not so much.”
Why Open Source Software is Moving to GitLab After Microsoft-GitHub Deal
http://www.itprotoday.com/open-source/why-open-source-software-moving-gitlab-after-microsoft-github-deal
Much open source software is seemingly on the move, as the also-ran GitLab repository benefits from GitHub’s sale to Microsoft.
Tomi Engdahl says:
Frederic Lardinois / TechCrunch:
Google says App Maker, its tool for quickly building and deploying business apps on the web, is now generally available as part of G Suite
App Maker, Google’s low-code tool for building business apps, comes out of beta
https://techcrunch.com/2018/06/14/app-maker-googles-low-code-tool-for-building-business-apps-comes-out-of-beta/
It’s been a year and a half since Google announced App Maker, its online tool for quickly building and deploying business apps on the web. The company has mostly remained quiet about App Maker ever since and kept it in a private preview mode, but today, it announced that the service is now generally available and open to all developers who want to give it a try.
Access to App Maker comes with any G Suite Business and Enterprise subscription, as well as the G Suite for Education edition. The overall idea here is to help virtually anybody in an organization — including those with little to no coding experience — to build their own line-of-business apps based on data that’s already stored in G Suite, Google’s Cloud SQL database or any other database that supports JDBC or that offers a REST API (that that’s obviously a bit more of an advanced operation).
To do this, App Maker provides users with a low-code application development environment that lets you build applications through a straightforward drag and drop environment.
Tomi Engdahl says:
Tech Talk: Data-Driven Design
https://semiengineering.com/tech-talk-data-driven-design/
How more data is shifting memory architectures.
Tomi Engdahl says:
What’s all the C Plus Fuss? Bjarne Stroustrup warns of dangerous future plans for his C++
Language creator calls proposals ‘insanity’
https://www.theregister.co.uk/2018/06/18/bjarne_stroustrup_c_plus_plus/
Earlier this year, Bjarne Stroustrup, creator of C++, managing director in the technology division of Morgan Stanley, and a visiting professor of computer science at Columbia University in the US, wrote a letter inviting those overseeing the evolution of the programming language to “Remember the Vasa!”
The formidable Vasa suffered from a design flaw: it was top-heavy, so much so that it was undone by a gust of wind. By invoking the memory of the capsized ship, Stroustrup served up a cautionary tale about the risks facing C++ as more and more features get added to the language.
Quite a few such features have been suggested. Stroustrup cited 43 proposals in his letter. He contends those participating in the evolution of the ISO standard language, a group known as WG21, are working to advance the language but not together.
Is C++ too challenging for newcomers, and if so, what features do you believe would make the language more accessible?
Stroustrup: Some parts of C++ are too challenging for newcomers.
On the other hand, there are parts of C++ that makes it far more accessible to newcomers than C or 1990s C++. The difficulty is to get the larger community to focus on those parts and help beginners and casual C++ users to avoid the parts that are there to support implementers of advanced libraries.
I recommend the C++ Core Guidelines as an aide for that.
Register: How would you characterize the current state of the language?
Stroustrup: C++11 was a major improvement of C++ and C++14 completed that work. C++17 added quite a few features without offering much support for novel techniques. C++20 looks like it might become a major improvement. The state of compilers and standard-library implementations are excellent and very close to the latest standards. C++17 is already usable. The tool support is improving steadily. There are lots of third-party libraries and many new tools. Unfortunately, those can be hard to find.
For C++20, we recommend to focus on:
Concepts
Modules (offering proper modularity and dramatic compile-time improvements)
Ranges (incl. some of the infinite sequence extensions)
Networking Concepts in the standard library
https://github.com/isocpp/CppCoreGuidelines/blob/master/CppCoreGuidelines.md
Tomi Engdahl says:
Linux literally loses its Lustre – HPC filesystem ditched in new kernel
Version 4.18 rc1 also swats Spectre, cuddles Chromebooks
https://www.theregister.co.uk/2018/06/18/linux_4_18_rc_1_removes_lustre_filesystem/
Linux has literally lost its Lustre – the filesystem favoured by HPC types has vanished in the first release candidate of version 4.18 of the Linux kernel.
Linus Torvalds’ announcement of the new release lauds the fact it’s shrunk markedly, much of which can be attributed to the removal of Lustre.
Lustre developers appear to be unfussed by the news: our scan of the project’s mailing lists didn’t find any comment.
Linux is not alone in losing interest in Lustre: Intel hired more than a handful of its prominent developers and added support for Hadoop but in April 2017 decided to stop offering its own version of filesystem and instead offered its code to the open-source community.
When Lustre emerged in the year 2003 it had little competition for creation of large-scale filesystems. Nearly 15 years on and Red Hat offers Gluster, IBM’s Spectrum Scale (aka the GPFS General Parallel File System) and scale-out NFS can all do plenty of what made Lustre useful. HDFS has emerged, too, for big data workloads.
Torvalds’ post on the new release lauds 4.18’s new features, among them support for AMD GPUs, fixes for Spectre V4 – aka Speculative Store Bypass – on Arm CPUs, and drivers to make two-in-one Chromebooks better at switching between desktop and tablet mode.
Tomi Engdahl says:
US Reclaims Title Of Fastest Supercomputer In The World – And It’s Insane
http://www.iflscience.com/technology/us-reclaims-title-of-fastest-supercomputer-in-the-world-and-its-insane/
The US has reclaimed the title of having the fastest supercomputer in the world, with the unveiling of a machine known as Summit.
Revealed last week at the Department of Energy’s Oak Ridge National Laboratory in Tennessee, it is capable of performing 200,000 trillion calculations per second, known as 200 petaflops. This is more than twice the previous record holder, China’s Sunway TaihuLight, capable of 93 petaflops.
It is not a single machine but rather a system of 4,608 computer servers, each containing two 22-core IBM Power9 processors and six NVIDIA Tesla V100 graphics processing unit accelerators.
Tomi Engdahl says:
Why Skylake CPUs Are Sometimes 50% Slower – How Intel Has Broken Existing Code
https://aloiskraus.wordpress.com/2018/06/16/why-skylakex-cpus-are-sometimes-50-slower-how-intel-has-broken-existing-code/
I got a call that on newer hardware some performance regression tests have become slower. Not a big deal. Usually it is a bad configuration somewhere in Windows or some BIOS settings were set to non optimal values. But this time we were not able to find a setting that did bring performance back to normal.
Same OS, Same Hardware, Different CPU – 2 Times Slower
But both hot instructions are preceded by the same common instruction named pause. Different methods execute the same CPU instruction which is for some reason very time consuming.
Pause on the new Skylake CPUs is an order of magnitude slower. Sure things can get faster and sometimes a bit slower. But over 10 times slower?
No this is not a bug, it is a documented feature.
That means when I execute heavily multithreaded applications on .NET on latest hardware things can become much slower.
A context switch is somewhere in the microsecond region and becomes much slower when many threads are waiting on the same kernel object.
Conclusions
This is not a .NET issue. It affects all Spinlock implementations which use the pause instruction. I have done a quick check into the Windows Kernel of Server 2016 but there is no issue like that visible. Looks like Intel was kind enough to give time a hint that some changes in the spinning strategy are needed.
When the issue was reported to .NET Core in August 2017 in September 2017 it was already fixed and pushed out with .NET Core 2.0.3
If I had a wish I would Microsoft make to port the ETW infrastructure on Linux because the current performance tooling still sucks at Linux.
The tool will report an issue only if you are running on a not fixed .NET Framework on a Skylake CPU.
Tomi Engdahl says:
Abner Li / 9to5Google:
Google Cloud and Unity to build managed services and tools for creating connected games; Unity moving infrastructure for its current services to Google Cloud — Unity’s engine and technology is playing a big role in the current wave of 3D and virtual reality games.
Google Cloud announces Unity strategic alliance to power real-time multiplayer games
https://9to5google.com/2018/06/19/google-cloud-unity-strategic-alliance-games/
Unity’s engine and technology is playing a big role in the current wave of 3D and virtual reality games. Google’s Cloud division today announced a strategic alliance with Unity to simplify game development that also sees the big developer migrate over its infrastructure to Google.
The strategic alliance is initially focussed on building real-time multiplayer gaming experiences. A suite of managed services and tools will help developers build, test, and launch connected games while offloading the hard work to Google and Unity.
Developers will be able to leverage the Google Cloud right from Unity’s development environment and tools. This allows for low-latency experiences, scaling to multiple servers, and multi-terabyte memory nodes for massive 3D environments.
Powering up connected game development through our alliance with Unity
https://www.blog.google/topics/google-cloud/powering-up-connected-game-development-through-our-alliance-with-unity/
Tomi Engdahl says:
Intel’s CEO resigns as information about a ‘past consensual relationship’ surfaces
https://techcrunch.com/2018/06/21/intels-ceo-resigns-as-information-about-a-past-consensual-relationship-surfaces/?sr_share=facebook&utm_source=tcfbpage
In a press release touting “another record year,” Intel dropped a bombshell, announcing that CEO Brian Krzanich is resigning, amid revelations of a “past consensual relationship” with an employee.
“Intel was recently informed that Mr. Krzanich had a past consensual relationship with an Intel employee,” the company notes in the release. “An ongoing investigation by internal and external counsel has confirmed a violation of Intel’s non-fraternization policy, which applies to all managers.
https://newsroom.intel.com/news-releases/intel-ceo-brian-krzanich-resigns-board-appoints-bob-swan-interim-ceo/
Tomi Engdahl says:
From here on, Red Hat’s new GPLv2 software
https://www.theregister.co.uk/2018/06/18/red_hat_gpl_violation/
projects will have GPLv3 cure for license violators
We’ll forgive you if you change your ways, says Linux giant
Red Hat on Monday said all of its newly initiated open-source projects that adopt GPLv2 or LGPLv2.1 licenses will be expected to include the GPLv3 “cure” provision.
The move follows Red Hat’s announcement last November, in conjunction with Facebook, Google and IBM, that the four companies intended to extend the GPLv3 violation remediation language to existing projects under GPLv2, LGPLv2.1 and LGPLv2, except when defending against lawsuits.
make the licenses more friendly to accidental violators.
The GPLv2 and LGPLv2.x licenses include a provision that automatically terminates the license upon noncompliance.
The GPLv3 cure clause gives open-source license violators grace periods of 30 or 60 days, depending on the circumstances, to remedy license violations.
Tomi Engdahl says:
AMD Tackles Coming “Chiplet” Revolution With New Chip Network Scheme
https://spectrum.ieee.org/tech-talk/semiconductors/design/amd-tackles-coming-chiplet-revolution-with-new-chip-network-scheme
The time may be coming when computers and other systems are made not from individually packaged chips attached to a printed circuit board but from bare ICs interconnected on a larger slice of silicon. Researchers have been developing this concept called “chiplets” with the idea that it will let data move faster and freer to make smaller, cheaper, and more tightly integrated computer systems.
4 Strange New Ways to Compute
https://spectrum.ieee.org/nanoclast/computing/hardware/4-strange-new-ways-to-make-a-computer
Tomi Engdahl says:
Where in the DevOps cycle do you do security?
https://opensource.com/article/18/6/where-cycle-security-devops?sc_cid=7016000000127ECAAY
The short answer: everywhere. Here’s how to implement a strategy.
Tomi Engdahl says:
The Science Behind Why Small Teams Work More Productively: Jeff Bezos’ 2 Pizza Rule
https://blog.bufferapp.com/small-teams-why-startups-often-win-against-google-and-facebook-the-science-behind-why-smaller-teams-get-more-done
Bigger doesn’t mean better when it comes to work. Jeff Bezos, the CEO of Amazon famously coined this with the 2 Pizza rule:
One former executive recalled that, at an offsite retreat where some managers suggested that employees should start communicating more with each other, Mr. Bezos stood up and declared, “No, communication is terrible!” He wanted a decentralized, even disorganized company where independent ideas would prevail over groupthink.
According to Bezos, the ideal is the “two pizza team:” if a team couldn’t be fed with two pizzas, it was too big.
Tomi Engdahl says:
React Native at Airbnb
https://medium.com/airbnb-engineering/react-native-at-airbnb-f95aa460be1c
In 2016, we took a big bet on React Native. Two years later, we’re ready to share our experience with the world and show what’s next.
Tomi Engdahl says:
5 big myths about introverts in IT
https://enterprisersproject.com/article/2018/6/5-big-myths-about-introverts-it?sc_cid=7016000000127ECAAY
Does IT have more introverts than any other department? We bust the most common myths about IT introverts – to help leaders tap the full potential of all team members
Tomi Engdahl says:
Software release life cycle
https://en.m.wikipedia.org/wiki/Software_release_life_cycle
Tomi Engdahl says:
3 reasons cost isn’t the best motivator for moving to the cloud
https://opensource.com/article/18/6/reasons-move-to-cloud?sc_cid=7016000000127ECAAY
Learn why cost savings shouldn’t be the primary driver for migrating to the cloud.
Tomi Engdahl says:
Do-it-yourself vs. Red Hat OpenStack Platform
https://verticalindustriesblog.redhat.com/do-it-yourself-vs-red-hat-openstack-platform/?sc_cid=7016000000127ECAAY
Tomi Engdahl says:
Why (some) agile teams fail
https://opensource.com/article/18/6/agile-vision-consider
A simple explanation of why an agile team might fail.
Tomi Engdahl says:
Getting started with React Native animations
https://opensource.com/article/18/6/getting-started-react-native-animations?sc_cid=7016000000127ECAAY
Here are the tools you need to overcome performance challenges when implementing React Native animations.
Tomi Engdahl says:
5 traits of good systems architecture
If you don’t know what data is in your system, you don’t know what it does.
https://opensource.com/article/17/10/systems-architect?sc_cid=7016000000127ECAAY
Tomi Engdahl says:
Open Source Web Design Editor Brackets 1.13 Released
https://www.linuxuprising.com/2018/06/open-source-web-design-editor-brackets.html?m=1
Tomi Engdahl says:
Peter Bright / Ars Technica:
Details of Qualcomm SDM1000, likely Snapdragon 1000, leak showing bigger, 20×15mm sized chip for Windows 10 on ARM PCs to compete with Intel Core Y and U CPUs
More details leak on “Snapdragon 1000,” Qualcomm’s chip for Windows 10 laptops
Design goes head to head with Intel’s Y and U series Core-branded processors.
https://arstechnica.com/gadgets/2018/06/more-details-leak-on-snapdragon-1000-qualcomms-chip-for-windows-10-laptops/
Details of the SDM1000, tentatively named Snapdragon 1000, a new Qualcomm chip built for Windows 10 laptops, have started to trickle out.
Microsoft’s development of Windows 10 for ARM has seen the company partner with chip company Qualcomm. The first Windows 10 on ARM machines use the Snapdragon 835 processor, with designs based on the Snapdragon 850 (a higher clocked Snapdragon 845 intended for laptops) expected later this year. Snapdragon 1000 will be the follow-up to the 850.
The Snapdragon 1000 is believed to be an even more powerful laptop chip intended to go head to head with Intel’s Y- and U-series Core processors. These have a 4.5W and 15W power envelope, respectively, and are used in a wide range of tablets and Ultrabook-type laptops. The Snapdragon 1000 is reported to have a 6.5W power draw for the CPU itself, with a total power draw of 12W for the entire SoC. The Snapdragon 1000 test platform has 16GB of LPDDR4X RAM and two 128GB UFS flash drives. It also has 802.11ad gigabit Wi-Fi, gigabit LTE, and a new power management controller.
Tomi Engdahl says:
Tom Warren / The Verge:
As Intel looks for its next CEO, its traditional chips business is troubled by delays and the firm is chasing competitors in GPUs ideal for AI computing
Intel now faces a fight for its future
Brian Krzanich’s surprise departure sets Intel on a race to find a new CEO
https://www.theverge.com/2018/6/22/17492184/intel-future-ceo-brian-krzanich-resignation-2018
Intel is facing a turning point in its nearly 50-year history. Intel CEO Brian Krzanich resigned yesterday, following an ongoing investigation into a past consensual relationship with an Intel employee that violated the company’s non-fraternization policy. It’s a surprise end for Krzanich, who first joined Intel more than 35 years ago and spent most of his time at the company on the operations side.
Krzanich was appointed Intel CEO five years ago, and was left with the messy task of fleshing out Intel’s mobile strategy and driving the company forward in new markets. Known for PCs and servers, Intel’s business has been disrupted by smartphones and the cloud, and the company was caught seemingly unaware by the rise of AI and autonomous vehicles.
Intel was originally planning to release its 10-nanometer processors back in late 2016, but the company recently delayed that once again to 2019 due to yield issues. It’s the first major stutter in Intel co-founder Gordon Moore’s incredibly accurate Moore’s Law prediction of roughly doubling the number of transistors in its processors every couple of years.
Intel is also redesigning its processors to protect against the Spectre security flaws that were uncovered earlier this year. Once Intel gets production back on track, it faces competition that it can’t pay off PC makers to stem.
Tomi Engdahl says:
Ubuntu reports 67% of users opt in to on-by-default PC specs slurp
Early data reveals most users run a single CPU, 4GB of RAM, one 1080P monitor
https://www.theregister.co.uk/2018/06/25/ubuntu_user_data_early_release/
Ubuntu has reported on data collected using the new user-profiling “feature” in version 18.04 of its GNU/Linux distribution.
Canonical, the company that backs Ubuntu, said it wanted to collect information about users’ PCs, attached hardware and location to help it focus future development efforts for the desktop version of the OS. Even with that noble intention, and sternly-worded pledges not to invade users’ privacy,
The data also revealed the following about Ubuntu desktop users:
They overwhelmingly use one CPU and one monitor;
That monitor is mostly 1920 x 1080 at 120dpi, but there’s a lot of Ubuntu users suffering at 1366 x 768;
The new “minimum install” option has attracted “a little over” 15 per cent of users;
Most users have four or eight gigabytes of RAM;
Just over half of users wipe the disk on which they install Ubuntu, but just 3.65 per cent bother to create an encrypted volume.
Tomi Engdahl says:
‘Snapdragon 1000′ Chip May Be Designed For PCs From the Ground Up
https://hardware.slashdot.org/story/18/06/24/0322204/snapdragon-1000-chip-may-be-designed-for-pcs-from-the-ground-up?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+Slashdot%2Fslashdot%2Fto+%28%28Title%29Slashdot+%28rdf%29%29
Qualcomm’s Snapdragon 850 processor may be intended for PCs, but it’s still a half step — it’s really a higher-clocked version of the same processor you’d find in your phone. The company may be more adventurous the next time, though. From a report:
WinFuture says it has obtained details surrounding SDM1000 (possibly Snapdragon 1000), a previously hinted-at CPU that would be designed from the start for PCs.
Tomi Engdahl says:
‘Snapdragon 1000′ chip may be designed for PCs from the ground up
It wouldn’t just be a spruced up mobile processor.
https://www.engadget.com/2018/06/23/snapdragon-1000-leak/
Tomi Engdahl says:
Linus Torvalds tells kernel devs to fix their regressive fixing
And get their timing right so that fixes aren’t features
https://www.theregister.co.uk/2018/06/25/linux_4_18_rc2/
Linus Torvalds has given the Linux kernel development community a bit of a touch-up, after finding some contributions to Linux 4.18 complicated the kernel development process.
In his post announcing release candidate 2 of Linux kernel 4.18, Torvalds mentioned “some noticeable filesystem updates, particularly to cifs.”
Tomi Engdahl says:
AMD Tackles Coming “Chiplet” Revolution With New Chip Network Scheme
Active silicon interposers could make for smaller, better computers but the networks need to mesh
https://spectrum.ieee.org/tech-talk/semiconductors/design/amd-tackles-coming-chiplet-revolution-with-new-chip-network-scheme
The time may be coming when computers and other systems are made not from individually packaged chips attached to a printed circuit board but from bare ICs interconnected on a larger slice of silicon. Researchers have been developing this concept called “chiplets” with the idea that it will let data move faster and freer to make smaller, cheaper, and more tightly integrated computer systems. The idea is that individual CPUs, memory, and other key systems can all be mounted onto a relatively large slice of silicon, called an active interposer, which is thick with interconnects and routing circuits.
There’s (at least) one problem: Though each chiplet’s own on-chip routing system can work perfectly, when they’re all connected together on the interposer’s network a situation can arise where a network tries to route data in such a way that a traffic jam occurs that winds up seizing up the computer. “A deadlock can happen basically where you have a circle or a cycle of different messages all trying to compete for same sorts of resources causing everyone to wait for everyone else,” Loh explains.
“Each of those individual [chiplets] could be designed so that they never have deadlocks,” says Loh. “But once I put them together, there are now new paths and new routes that no individual had planned for ahead of time.” Trying to avoid these new deadlocks by designing all the chiplets together with a particular interposer network in mind would defeat the advantages of the technique: Chiplets, then, couldn’t be designed and optimized easily by separate teams, and they couldn’t easily be mixed and matched to quickly form new systems.
The AMD team found that deadlocks on active interposers basically disappear if you follow a few simple rules when designing on-chip networks. These rules govern where data is allowed to enter and leave the chip and also restricts which directions it can go when it first enters the chip. Amazingly, if you follow those rules you can pretend everything else on the interposer—all the other logic chiplets, memory, the interposer’s own network, everything—is just one node on the network. Knowing that, separate teams of engineers can design chiplets without having to worry about how the networks on other chiplets work or even how the network on the active interposer works.
Tomi Engdahl says:
http://www.etn.fi/index.php/13-news/8154-maailman-pienin-tietokone
Tomi Engdahl says:
U.S., China Spar in Supercomputers
U.S. wins in petaflops, China has more systems
https://www.eetimes.com/document.asp?doc_id=1333414
The U.S. retook a lead in supercomputer performance on the latest Top 500 list, but for how long remains to be seen. China not only extended its lead in the number of high-performance systems but it is racing to retake a performance lead with an early exaflop-class computer.
The latest results give the U.S. bragging rights but show that China continues to grow in its technological sophistication. It comes at a time of trade tensions between the countries in part due to friction over semiconductor policies.
The list also shows that accelerators, mainly Nvidia GPUs, continue to drive many of the most muscular systems in the world. However, Intel and Pezy Computing, a fast-rising accelerator maker based in Japan, also made significant showings.
Tomi Engdahl says:
Ubisoft CEO: Cloud Gaming Will Replace Consoles After the Next Generation
https://m.slashdot.org/story/341936
Tomi Engdahl says:
Why Go is skyrocketing in popularity
https://opensource.com/article/17/11/why-go-grows?sc_cid=7016000000127ECAAY
In only two years, Golang leaped from the 65th most popular programming language to #17. Here’s what’s behind its rapid growth
Tomi Engdahl says:
Can ARM-based Thin Clients Provide a Secure Alternative to Windows Desktops?
https://www.electropages.com/2018/06/arm-based-thin-clients-provide-secure-alternative-to-windows-desktops/?utm_campaign=&utm_source=newsletter&utm_medium=email&utm_term=article&utm_content=Can+ARM-based+Thin+Clients+Provide+a+Secure+Alternative+to+Windows+Desktops%3F
Do the challenges presented by the disclosure of the Meltdown and Spectre exploits remain locked into the usual bug-patch-repeat dynamic? Do businesses have a choice beyond putting up with PCs slowed by patching the bugs?
It turns out that there could be a hardware solution that doesn’t demand a massive increase to the hardware budget: ARM-based thin clients on every desktop.
Security guru Bruce Schneier suggests that these vulnerabilities represent “the future of security – and it doesn’t look good for the defenders… attacks against hardware, as opposed to software, will become more common.”
Using hardware less susceptible to attack seems a better solution than patching mistakes.
Inadequate Solutions Call for New Approaches
Now widely rolled-out, these patches have so far been successful in mitigating the risks of both Spectre and Meltdown’s vulnerabilities. In short, they should now be protected from exploitation. But the solution comes with a cost: speed.
By tightening up the security holes in the affected CPUs, operating systems have slowed down. In some cases (especially on older CPUs) this reduction in server and PC performance has had frustrating results.
Not all hardware is affected by these vulnerabilities, however.
many current Intel Atom laptops and tablets are immune to the Meltdown and Spectre bugs.
ARM Computers with Thin Client Support
Although some ARM-based systems are affected by Meltdown and Spectre, others are not. Many Android smartphones, for instance, require patching. Some models of the Raspberry Pi, on the other hand, do not.
Back in 2016, desktop virtualization publisher Citrix discussed how the Raspberry Pi 2 could be used as a thin client, noting that “Typical business users don’t care if they have a PC with a 2.0 GHz CPU, or 3.0 or 4.0 as long as it works well and looks good.”
Significantly, both the Raspberry Pi 2 and 3 models are unaffected by Meltdown and Spectre. Could these devices prove the necessary jumping off point for a new generation of office-ready hardware that doesn’t have the weaknesses of its predecessors?
At this point, it’s probably worth revisiting Bruce Schneier’s words concerning computer security of the future: “attacks against hardware, as opposed to software, will become more common.” Sooner or later
Tomi Engdahl says:
Micro Focus sells Suse for $2.5B
https://techcrunch.com/2018/07/02/micro-focus-sells-suse-for-2-5b/?sr_share=facebook&utm_source=tcfbpage
Suse, one of the longest-running commercial Linux distributors and, these days, a major player in the open-source infrastructure and management space, has been through a few ownership changes in recent years. Micro Focus acquired Suse from The Attachmate Group back in 2014, which itself had acquired Novell, the then-owner of Suse, in 2010.
What was once a solid Linux distribution for the enterprise is now a player in various open-source fields, with a focus on software-defined infrastructure and application delivery solutions, as well as other managed cloud services.
Tomi Engdahl says:
20 questions DevOps job candidates should be prepared to answer
https://opensource.com/article/18/3/questions-devops-employees-should-answer?sc_cid=7016000000127ECAAY
Want to build a positive, productive work environment? Focus on finding a mutual fit during the hiring process.
Tomi Engdahl says:
https://www.tekniikkatalous.fi/tyoelama/koodaajapula-nakyy-epatoivoisinakin-rekrytointiyrityksina-hilavitkuttimia-ja-muita-etuja-yritetaan-6731414?utm_source=facebook&utm_medium=paidsocial&utm_campaign=TT_FB_NA_07
Tomi Engdahl says:
https://www.tivi.fi/Kaikki_uutiset/no-nyt-on-todellinen-lapimurto-muisteissa-pikkuruiseen-korttiin-mahtuu-pian-enemman-kuin-tyypilliselle-palvelimelle-6731645?utm_source=Facebook&utm_medium=Social&utm_campaign=TV_NA_06&utm_content=Ad+-+Post%3A+%2FTivi%2Fposts%2F10155558243732267
SD Express card capacity up to 128 terabytes
transfer rate up to 985 Mb in second
https://www.businesswire.com/news/home/20180626006664/en/SD-Express-%E2%80%93-Revolutionary-Innovation-SD-Memory
Tomi Engdahl says:
On Labeled data
https://medium.com/@TalPerry/on-labeled-data-85fbaf1bdf89
It turns out that a large portion of real-world problems have the property that it is significantly easier to collect the data (…) than to explicitly write the program. A large portion of programmers of tomorrow do not maintain complex software repositories, write intricate programs, or analyze their running times. They collect, clean, manipulate, label, analyze and visualize data that feeds neural networks.
Tomi Engdahl says:
The life cycle of a software bug
https://opensource.com/article/18/6/life-cycle-software-bug?sc_cid=7016000000127ECAAY
From finding software glitches to resolving them, here’s how development teams squash bugs.
Tomi Engdahl says:
Red Hat OpenStack Platform 13 is here!
https://redhatstackblog.redhat.com/2018/06/27/red-hat-openstack-platform-13-is-here/?sc_cid=7016000000127ECAAY
Red Hat OpenStack Platform lets you build an on-premise cloud environment designed to accelerate your business, innovate faster, and empower your IT teams.