Category: Know How

High-end graphics card: NVIDIA’s Pascal architecture

Demanding graphics applications in the fields of digital signage and virtual reality require powerful hardware. NVIDIA is currently setting the standard with its Pascal architecture, the Geforce 10 series, which has already triggered a storm of enthusiasm among gamers. We will most likely be able to answer the question of how these top performers can be integrated into professional mini PC solutions in the first quarter of 2017.

Mobile networks: more and more 4G

The availability of 4G networks in Europe is constantly increasing. The successor to the current 3G standard is characterized by significantly higher data rates. Suitable LTE modules will soon be available for many spo-comm systems.

Apollo Lake SOCs: Now also for IoT and embedded

Following the presentation of the Intel Apollo Lake CPUs for notebooks and tablets, the CPUs for embedded systems have now been presented. These are listed in a table below. It is not yet clear which of these will be used in spo-comm systems in the coming year. However, our product development team has already taken a closer look at these small power packs.

Discover our professional mini PCs

The abbreviation RAID stands for “Redundant Array of Independent Disks”, which means “redundant arrangement of independent hard disks”. In a technical context, redundancy means that additional functionally identical or comparable parts of a system are present that are not required in normal operation. If a failure occurs, they can step in or, if they are already running in parallel, do the work on their own from then on.

A RAID is a network of at least two mass storage devices(hard disks or SSDs) that work together. The size and type of storage medium should be identical. The different RAID levels (e.g. RAID 0 or RAID 1) differ in the size of the storage capacity and the security of the stored data.

 

Hardware or software RAID

 

The RAID can be created in different ways. Hardware RAID requires an extra microprocessor, a so-called RAID controller. The chip is often located near the memory and organizes the data distribution. The main processor is not loaded. In larger network environments, such as data centers, external RAID systems are often used.

 

A software RAID is completely organized on the software side. The RAID can already be implemented in the BIOS in order to subsequently install the operating system on it. So this is secured. Alternatively, the RAID array can also be set up in the operating system first, but then the operating system does not benefit from the RAID. As there is no special RAID controller, all calculations are carried out on the main processor, which, depending on the application, is subject to a higher load.

 

RAID 0

 

With RAID 0, two or more storage media are combined to form a logical drive. These can work at an increased speed. For this purpose, the hard disks are divided into blocks of the same size, which are then arranged alternately, as in the zipper method. This technique, which allows all disks to be accessed in parallel, is known as “striping”. The advantage of RAID 0 is the acceleration of data transfer. One disadvantage is the lack of security: if a disk fails, data can no longer be fully restored. It is therefore only recommended if reliability is not an issue. As there is no redundancy in RAID 0, strictly speaking it is not a RAID, but only an “array of independent disks”.

 

RAID 1

 

RAID 1 ensures a high level of reliability with mirroring technology. The data is written simultaneously to all the hard disks involved (usually two). Thanks to two identical disks, there is now full redundancy. In plain language, this means that if one of the two components fails, the remaining hard disk can continue to deliver all data. However, a RAID 1 is no substitute for a data backup. Errors, such as viruses or accidental deletion of files, are immediately transferred to the mirror disk. The total storage capacity of RAID 1 is as large as the smallest hard disk involved.

 

RAID 5

 

At least three storage media are required for a RAID 5. It offers increased speed and some safety. As with RAID 0, the data is distributed to all drives in blocks. In addition, the information required to restore the data (so-called parity data) of another drive is stored on each disk. If only one hard disk fails, the data can be restored. The possible storage capacity is calculated from the number of disks minus one, multiplied by the capacity of the smallest disk. With three disks, each with 1 terabyte of memory, this would be 2 TB, for example.

 

RAID 10

 

RAID 10 is a combination of 1 and 0, so it can combine both advantages and offers both more speed and a high level of security. At least four drives are required. Two hard disks are connected to form RAID 1. These sets are then combined to form RAID 0. This variant is very safe, as one plate may fail in each set. However, only half of the total storage capacity is available for this.

 

RAID with spo-comm Mini-PCs

As at least two hard disks are required for a RAID array, this is only possible with certain spo-comm PCs. These include the spo-book EXPANDED Q170, spo-book NINETEEN Q170 and spo-book NOVA CUBE Q87. We will be happy to advise you if you have any questions. Otherwise you can simply specify when ordering if RAID 0 or 1 is required.

Get in touch with us

Green IT: Federal Environment Agency recommends mini PCs

The Federal Environment Agency has presented a 6-point plan for the environmentally conscious and resource-saving use of computers in the workplace. A key component is the recommendation to equip purely stationary workstations with mini PCs. These are characterized by lower electricity consumption and therefore also have a better greenhouse gas balance. In addition, the service life is much longer than that of laptops. Long-term availability makes it easier to install and set up the systems, as the same platform can be used for years.

Current Atom chips: Apollo Lake

Following the news about the high-end processors of the Kaby Lake generation, Intel has now also introduced more cost-effective models, which are also manufactured using the 14nm process. Celeron and Pentium processors with two or four Goldmont CPU cores and a TDP of 10 watts are referred to internally as Apollo Lake. They are considerably faster than their predecessors and are suitable for use in mini PCs. The first spo-comm models with the  new Atom technology are planned for 2017.

Even smaller and even faster: M.2 SSDs

Solid state disks in M.2 format are also increasingly being used in the embedded and industrial PC sector. One advantage over the standard SSDs in 2.5″ format is the much smaller design. Furthermore, the integrated SATA 3 port ensures faster transfer rates than previous card formats. M.2 SSDs will also be part of some spo-comm PCs from 2017.

Windows 10 IoT: New license areas

Microsoft has now provided specific details regarding the licenses for Windows 10 IoT. The former Windows Enterprise area is now divided into three licenses: Entry, Value and High-End. These depend on the hardware used.

More about Windows IoT

On PCs with Skylake CPUs, Windows 7 cannot be installed via a Windows CD as there is no driver support. Intel has removed support for eHCI (USB 2.0) in the chipset for these platforms and now only supports the new xHCI standard (USB 3.0). However, as the drivers for xHCI are not yet available in the Windows 7 installation media, the USB interfaces cannot be addressed.

What is xHCI and eHCI?

xHCI (eXtensible Host Controller Interface) is the specification of the new USB 3.0 interface and thus the successor to eHCI (enhanced Host Controller Interface), which defined the USB 2.0 standard. The xHCI module is located on the mainboard and controls the USB interfaces (3.0 and 2.0). If, as with the installation of Windows 7, the drivers for xHCI are missing, all USB ports (including USB 2.0) will not work. Accordingly, the mouse and keyboard connected via USB cannot be operated.

Extended support period

Microsoft originally limited the support period (provision of updates and security patches) for Windows 7 on Skylake systems until 2017. However, according to a recent announcement, this has now been extended until the end of 2020.

Windows 7 for spo-comm Mini-PCs with Skylake platform

spo-comm has found a solution to these difficulties and delivers its Skylake systems with pre-installed Windows 7 Professional 64-bit and Windows 7 Embedded on request. Each delivery includes a USB recovery medium for any necessary reinstallation (recovery process of a system in 5-10 minutes).

Discover our mini PCs with the latest Windows versions

New Intel platform: Kaby Lake

With Kaby Lake, Intel has introduced the latest and now seventh processor generation. As with Skylake, production is based on the 14 nm process. The platform is designed to increase performance by up to 12 percent. There may also be completely new CPU types, such as a quad-core processor with 15 watt TDP (U series). The first spo-comm systems with Kaby Lake processors will be available from mid-2017.

Internet of Things for everyone: Windows 10 IoT

With the introduction of Windows IoT, the successor to Windows Embedded, Microsoft has launched a new license model that is dependent on the CPU installed. Depending on the hardware and applications used, Windows IoT is often the cheaper alternative to the standard Windows 10 operating system.

Vehicle Computing: News from the MOVE series

An update for the MOVE series systems will be launched on the market shortly. The tried and tested features are brought up to date with the latest technology. Furthermore, special current transformers and the optional battery unit ensure smooth operation of the Mini-PCs in vehicles of all kinds.

Windows 7: Now also for Skylake

spo-comm has found a solution to run Windows 7 Professional 64-bit on industrial PCs with the latest Skylake platform. New systems with sixth-generation CPUs, such as the spo-book CORE and the OPS player, can now be ordered with the popular Microsoft operating system.

More spo-comm Mini-PCs

A MAR license is quickly explained:

Microsoft Authorized Refurbisher (MAR) is a program in which previously used Windows licenses have been deactivated and are now being used again.
Some of these are already labelled. However, compared to the new versions, no compromises are made in terms of functionality and the like.
The MAR variant is a low-cost alternative to conventional licenses and is available on request.

spo-comm and WindowsIoT

In another article, we have already looked at the difference between HDD and SSD as well as the advantages and disadvantages. Now let’s go into more detail and introduce different types of SSD.

More about the HDDs and SSDs

The terms SLC, MLC and TLC refer to the type of SSD memory. They indicate how many bits are written per memory cell on an SSD.

What do the abbreviations stand for?

SLC stands for “Single Level Cell”. As the name suggests, an SLC cell stores exactly one bit. MLC stands for “Multi Level Cell” and means that more than one bit is stored per cell. This is usually two bits per cell, as TLC (“Triple Level Cell”) has now also been developed, with three bits per memory cell.

Of these three types, single level cells have the highest read and write speed. They also require less power, are more durable, more robust and suitable for an extended temperature range. However,  they only have a relatively small memory and are comparatively expensive. Multi-level cell SSDs, on the other hand, have a higher storage density and can therefore store a considerably larger volume of data.

How does storage density affect wear?

MLC and TLC SSDs are often criticized for not being durable enough. Since more information is stored per cell, wear is higher than with SLCs. For comparison: While manufacturers often specify a service life of 100,000 write cycles for single level cells, triple level cells only last for 1,000 write operations. MLCs are around 5,000 to 10,000 cycles.
This sounds comparatively low, but is usually absolutely sufficient. Thanks to technologies such as wear leveling, almost nobody has to worry about the durability of their SSD these days.

What is wear leveling?

Wear leveling (literally: “wear compensation”) is used to extend the service life of an SSD – a technology that is often integrated into SSDs. An algorithm is used to distribute the write operations so that all memory cells are written to equally often and the SSD wears out evenly.

This technique also distinguishes between dynamic and static wear leveling. The former only distributes dynamic data, i.e. data that changes when something is deleted and something else is saved. To prevent some cells from being written to only once while others are already failing, static wear leveling also moves static data to other blocks. The dynamic variant is often found in USB sticks, while static wear leveling is usually used for SSDs.

What kind of SSDs does spo-comm sell?

spo-comm offers Multi Level Cell SSDs in various sizes as standard, as these offer a high storage volume at a  suitable price-performance ratio. Some PCs (such as the WINDBOX or RUGGED series) can also be ordered with a 16 or 32 gigabyte SLC SSD. These are suitable for an extended temperature range of -40 to 85 degrees, but – despite the lower storage density –  are priced somewhat higher.

Small, smaller, smallest: mini PCs are shrinking by leaps and bounds

In the near future, mini PCs will be introduced that are considerably smaller while offering comparable performance. Thanks to new manufacturing technologies, not only is the size of important components such as the CPU reduced, they also consume less power. This enables smaller fans or heat sinks and consequently smaller housings. The disadvantage: there is less space for connections.

Now even sharper: 4K@60Hz

The HDMI 2.0, Display Port 1.2 and higher standards are increasingly being introduced in new systems. This means that there are numerous models, particularly among the mini PCs for the digital signage market, that can display UHD films and animations without any problems.

New memory technology: DDR4

From 2017, new systems will be equipped with DDR4 memory. In addition to the higher clock rate, the maximum memory capacity of the RAM modules is also increased. Mini-PCs in the medium to high performance class can then address up to 32 gigabytes of RAM.

Never wireless again: secure digital plug connections

Plug connections that are exposed to mechanical influences can easily come loose without additional protection. This is why the traditional DB9, DVI and VGA connections are still popular, as they can be secured with two screws. With newer standards, it is more difficult, as the respective design guidelines rarely include a backup option. This is changing with the USB connection, for which the first screw-on connectors have been defined in the Type C standard. s po-comm also tries to offer secure connections for as many systems and connections as possible. Various systems are now equipped with screw-on power supply units.

spo-book WINDBOX II Quad: Now with 4 GB RAM

The WINDBOX II Quad is the first spo-comm PC to be equipped with 4 gigabytes of RAM as standard. The price remains the same.

Configure your WINDBOX II Quad

Tablet support increases: Uniform Windows licenses

Tablets or handhelds with a fully-fledged Microsoft operating system are increasingly being offered not only in the area of consumer devices (e.g. Microsoft Surface Pro), but also in the commercial environment. The advantage is obvious: no independent apps need to be developed. Instead, the ERP, CRM or PPS systems used in the companies can be installed and used directly.

What does RAM actually mean?

The abbreviation “RAM” stands for “random access memory”, which is often also referred to as “direct access memory”. With this type of memory, access to each individual memory cell is possible, while other types, such as flash, only store in blocks. In contrast to the hard disk, RAM is a volatile (semiconductor) memory. This means that as soon as the power supply is switched off, all data is lost.

While hard disks can store large amounts of data over a long period of time, RAM stores small amounts for a short time. It works at a much higher speed: the read and write process is up to 250 times faster than with SSDs, and the access time is up to 120,000 times faster. Due to its speed, RAM is usually used as working memory. The two terms are therefore used synonymously.

The memory is located on the RAM modules in the form of chips. More gigabytes of RAM therefore means more memory chips on the RAM bar. Nowadays, the modules generally have 2 to 64 gigabytes of RAM.

What does the working memory do?

The working memory serves as a short-term buffer for the CPU. The operating system, Internet browser and other data on which changes or calculations are made are temporarily stored in it. For example, the processor also loads texts, images and videos into RAM so that they can be displayed and edited quickly. The data is only saved to the hard disk or SSD when it is saved.

More spo-comm knowledge

To find this out, you first need the serial number of the warranty seal beginning with “SB”, which can usually be found on the underside of the PC. This allows you to assign the system to the delivery bill and match the time of purchase with the warranty period (standard warranty is two years, extended warranties up to five years).

If you do not have the corresponding delivery bill to hand, you are welcome to contact us by e-mail or telephone. We will then check the warranty period and help you with the next steps.

Contact us!!!