Following on from our recent article discussing some of the changes afoot for the 5th and 6th Gen Intel NUC and its absence of a mini-PCIe adapter, we’re pleased to announce we have our very own M.2 Single and Dual Port Gigabit Ethernet card for 5th and 6th Gen machines.
We’re really excited about the product as it promises all of the functionality our customers have been asking us for including:
Single and Dual Port options
Gigabit Ethernet Speed
Reliable Intel i210 Chipset Controller
Integration into our unique 1U NUC case
PCI express bus comms standard – not USB!
For the first time, owners of 5th and 6th Gen machines will be able to add a second (or third) Gigabit Ethernet port to attach a NUC to multiple networks. We believe this is potentially another world-first after we were the first company ever to rack-mount the Intel NUC in 2014. To know more, please get in touch and speak with one of the team.
We are also able to sell the Ethernet cards as a standalone item because they are compatible with all 5/6th Gen NUCs or in fact any M.2 2280 size slot. However we do stress that the standard Intel NUC retail cases do not allow space for expansion using this module. Please contact us for a quote if you wish to use it as a standalone item.
Last year we spoke in detail about the gap between traditional storage and SSD closing significantly. A lot can change in a year, a phrase which appears particularly apt to the world of SSD. Being more affordable, more powerful and with greater capacity, it’s fair to say SSD is playing a major part in ushering out traditional spinning disk storage.
Previously, we suggested that SSD manufacturers were looking at new ways to increase capacities almost exponentially and that is now becoming a reality. By adopting Micron’s 3D NAND flash chips and stacking them vertically, Intel has created an affordable 10TB SSD, which they recently announced will be making its way to market later this year. With each of these new advancements we are increasingly seeing a Moore’s Law for storage memory.
Intel’s new solid state drive promises a huge jump in capacity and performance, yet retaining the same physical footprint – meaning positive impacts on both device size and data centre real estate. The arrival of the new 3D NAND technology should also signal a lower price per/GB, whilst delivering significant advancements over standard Planar NAND. And if all goes to plan you should be able to reap these cost and productivity gains in the coming months.
It’s more than just capacity
If anything was holding people back from SSD, some might argue it was the capacity, but really that’s a non-argument now, especially with the arrival of 3D NAND. Similarly, SSD uses an embedded controller to read and write data and in the early days of their development, write speeds were lower than read speeds leading to gripes compared to traditional disk alternatives. However, the current crop of SSDs offer massive jumps in performance too, offering a far more even performance symmetry – with read and write speeds potentially averaging 580 mps and 500 mps respectively, meaning less time required to store and pull data when you need it.
The prospect of 3D NAND technology promises even greater read/write bandwidth, I/O speeds and power savings all within the next 12 months. Consequently, it appears 3D NAND will smash previous ceilings on both capacity and performance. Perhaps its arrival will mark the final demise of spinning disk.
If you’d like to know more about SSD and its inclusion in our products or believe the technology could lend itself to a potential challenge you face, please get in touch to talk further.
It is with much disappointment that we have discovered that Intel will not be manufacturing board-only versions of their 6th generation NUC. Little has been communicated by Intel, the variant mysteriously dropping off the roadmap for the product. Despite attempts to qualify with Intel why this decision has been reached, we are yet to receive any good reason for the decision.
As the leading innovator in the field of NUC, G2 broke new ground when we manufactured the world’s first rack-mount NUC in 2014. Since then, we have continued to develop what has been a highly successful product for us, which provides a solution to many real-world deployment scenarios.
Unfortunately, the ramifications of Intel’s decision will be some level of feature loss between 5th and 6th generation products. Whilst not vast, the primary difference is the ability to use higher performance, more efficient DDR4 memory and the inclusion of Intel Iris graphics which was not present in the previous generation. Whilst small differences now, if this decision from Intel sticks and board-only NUC options are dropped permanently then the gulf in features will become larger over time. Inevitably, this will impact many customers who want to use NUC in non-standard configurations and outside of Intel’s preferred NUC kit alternative. It’s especially frustrating when removing boards from NUC kits to potentially achieve a board-only result actually invalidates the warranty as well. In the meantime, OEMs like G2 will be forced to continue using 4th and 5th generation board-only options to provide innovation outside of the NUC kits Intel is now focussing its efforts on.
It’s a real shame, but rest assured we will continue to put pressure on Intel in every way we can to hopefully overturn this decision.
If you’d like to know more about our family of NUC products, why not take a further look here? Alternatively, contact us to speak further with one of the team.
The delays Intel experienced getting Haswell to market and subsequent availability issues has inevitably led many PC manufacturers to stick with the microarchitecture once it ‘arrived’ and skip Broadwell – it’s more powerful and efficient sibling. Now, most PC manufacturers (ourselves included) are eagerly anticipating the arrival of Skylake and the significant jumps this promises – claimed by Intel to be the most significant processor upgrade for a decade. Importantly, we hope Intel has learnt some lessons from Haswell as Skylake holds much promise and it would be a shame for adoption to suffer at the hands of early availability issues like its forefather.
Life after Haswell…
What the arrival of Skylake does spell however, is troubled times ahead for users of Haswell. We are already finding it difficult to source Haswell motherboards and whilst we will be able to offer legacy support for some time yet on Haswell, G2 is still at the mercy of component availability and motherboards are becoming like proverbial hen’s teeth. Disappointingly, this will have a knock-on effect for customers whose software is dependent on this architecture. We’d therefore urge any organisations who expect to be tied into Haswell for some time to consider roadmaps and options for their software as early as possible.
It’s not all bad news though! 6th generation Skylake is built to use DDR4 RAM and until its advent, this next-gen memory was both expensive and challenging to get hold of. The Skylake-effect is driving down prices and significantly improving availability as adoption gathers momentum. The result means changing to the latest systems and running DDR4 RAM can now be achieved with only a modest increase over DDR3. For the geeks out there, what DDR4 brings is greater chip density thanks to larger DIMM capacities, so more memory can be attached to a single motherboard (theoretically 512GB or four times DDR3) improved data transfer rates (no ceiling has been found on this yet!) and significantly lower power consumption, which is important if the machine is part of a larger deployment when power savings start to mount up and memory stability is critical. In layman’s terms this means greater data capacity and higher data transfer rates, all at lower voltage and therefore greatly reducing the chances of memory failure, particularly caused by taxing applications.
If you’re interested in exploring the possibilities of Skylake and DDR4 memory further or want some advice on life after Haswell, please get in touch and speak to one of the team.
With Windows 10 first year anniversary just round the corner (an anniversary update is expected sometime in July), it’s likely to prompt people to start looking more closely at the operating systems (OS) they have deployed and any application dependencies tied to them. In particular, it brings into focus what’s ahead for Windows 7. The End of Sales for PCs pre-installed with Windows 7 Professional is actually October 2016! In our opinion, it’s not been as well publicised as it otherwise might have been because of all the hype surrounding Windows 10. However, before, you come out in a cold sweat, this doesn’t mean the end of support – Microsoft will continue to patch and update the OS until January 2020. What it does mean though is that if you want to buy new PCs pre-installed with the OS you’re not going to be able to through conventional routes.
You work with a partner like G2 who operates under an embedded licensing agreement. This means we can continue to sell Rack and Mini PC systems pre-installed with the OS (and in fact even older OS like XP) beyond the end of sale date.
For us it’s critical as in our experience there are still many organisations that rely on Windows 7 for critical applications – having proved itself as a robust and reliable OS. In fact, 52% of all machines globally run Windows 7 proving exactly that, and it wasn’t that long ago people panicked about getting off of XP.
Using our industrialised motherboards we will be able to build and support Windows 7 PCs for a LONG time to come.
If you’re interested in understanding more about our Rack and Mini PCs and the supply of Windows 7 pre-installed machines then get in touch. Alternatively,explore our products to find something suitable for your needs.
The phrase ‘system on chip’ has become a buzz phrase among tech enthusiasts in recent years. Essentially, it refers to a processor embedded onto a board which may be integrated with all of the components required to function a computer or electrical device. Recently there has been some commercial success with easy-entry platforms like Raspberry Pi and Intel’s Galileo, but do these devices, also known as Integrated Circuit PCs have any place in enterprise computing?
First out the blocks
As you’d expect, Intel has been one the first out of the blocks in product development and building the hype around system on chip. Their virgin product is an Integrated Circuit PC on a HDMI input device called Compute Stick. Although just 4.5 inches in length, the Compute Stick does come with an impressive spec. The device has a choice of Core M processors and up to 4 GB of memory, which will be sure to spark interest and inspire innovation within the small computing industry.
Whilst Intel has developed an out-of-the-box product that is at this moment quite unique, it will arguably still only appeal to those looking for a tiny device where performance and necessity to configure is not king. For this reason, in our opinion it’s still really a consumer device for now and is unlikely to stand up to the demands of enterprise computing, where users need to connect to multiple network or require more powerful compute to run graphics intensive software. But, as we’ve seen with other consumer computing phenomenon like the Apple iPad, it doesn’t take long to establish traction and see new consumer tech start to play a disruptive role in the enterprise.
The new old
We believe the trend is set and manufacturers will continue to create products marketed as system on chip, albeit primarily aimed at the consumer market. Despite the buzz however, system on chip isn’t actually something new. Integrated Circuits have been around in small computing for a while now. In fact, most of today’s rack PCs are built using components embedded onto a chip. Indeed, G2 harnesses a variation of system on chip technology in our rack-mount NUC, but complement this with other functionality that better meets the performance and expandability requirements of the enterprise.
Trickle down benefits
It’s practically guaranteed that as system on chip technology develops, with that will come greater performance options and evermore innovation. And that makes us excited, as undoubtedly there will be trickle-down benefits for our own product development roadmap, helping us to continue creating better performing, smaller product packages that pack-in as much functionality as possible.
As ever, we’ll be keeping a keen eye on this market and will of course share our observations and developments.
If you’d like any help understanding system on chip or would like to talk to us about any of our products, please get in touch.
Anyone familiar with Intel’s 4th generation NUC will probably be aware that if you needed the product to connect to multiple networks via Ethernet then you needed a mini-PCIe adapter. This is something G2 has provided to our NUC customers for some time, meeting a pretty commonplace requirement.
All change for 5th & 6th generation NUC
Unfortunately, the ability to add a second Ethernet port to the latest NUC is more problematic, causing issues for those who need to bridge connections to multiple networks. As Intel has phased out the Mini PCIe interfaces on their motherboards in favour of M.2 slots to enable increasingly smaller package footprints, there is no longer a compatible Ethernet adapter. Until now….
At last an M.2 Ethernet adapter for the latest NUC!
To meet this need we’re currently developing an M.2 Ethernet card for the 6th generation Intel NUC. Here’s what it promises:
Two Intel i210 Ethernet ports
Interface with the smallest 2230 M.2 Interface
Dual Gigabit Ethernet Ports from one PCB – your NUC can be now configured with three Ethernet Ports instead of one
Utilises superior PCI express BUS interface and NOT USB adaption
It’s an exciting development and one we believe satisfies a real market need, whilst opening up new options for where and how the latest NUC technology can be used.