Reviewers of computer hardware and pretty much all the latest tech, Anand Tech has all the latest news and reviews regarding VR headsets, their applications, and their producers. Join the community of tech geeks and stay updated VR and lots more.
Pink is a color that is rarely used for PC hardware, however over the last couple of months we spotted three manufacturers adopting pink designs for desktops, laptops, peripherals, and even an all-in-one cooling system. This week ASUS ROG has followed, and started sales of multiple gaming peripherals featuring pink with a grey color scheme.
The ASUS ROG PNK family of products includes the ROG Gladius II Origin PNK mouse, the ROG Sheath PNK mousepad, the ROG Strix Fusion 300 PNK headset, as well as the ROG Strix Flare PNK keyboard with Cherry’s MX Red or MX Brown switches. Instead of choosing bright pink color it once used for a ZenBook UX laptop, the company decided to go with rather soft accents of pink, which will be suitable for anyone interested.
Since we are dealing with ASUS ROG-branded hardware, all the products are equipped with addressable RGB LEDs that can be controlled using the company’s Aura software.
ASUS ROG’s PNK peripherals are currently available directly from ASUS in the USA for $39.99 (mouse pad), $99.99 (mouse/headset), and $179.99 (keyboard). All the products are marked as limited edition, so we do not expect wide availability.
The biggest news of the annual Computex trade show came from AMD: the company is poised to launch its next generation Zen 2 microarchitecture, along with updates to its Ryzen and EPYC product lines. AMD is going all-in with its chiplet CPU architecture, as well as with its new RDNA graphics architecture for the upcoming Navi graphics product family. After AMD’s keynote, we joined a small roundtable of journalists to put questions to AMD’s CEO, Dr. Lisa Su.
While display interface standards are slow to move, at the same time their movement is inexorable: monitor resolutions continue to increase, as do refresh rates and color depths, requiring more and more bandwidth to carry signals for the next generation of monitors. Keeping pace with the demand for bandwidth, the DisplayPort standard, the cornerstone of PC display standards, has now been through several revisions since it was first launched over a decade ago. And now this morning the standard is taking its biggest leap yet with the release of the DisplayPort 2.0 specification. Set to offer nearly triple the available bandwidth of DisplayPort 1.4, the new revision of DisplayPort is almost moving a number of previously optional features into the core standard, creating what’s in many ways a new baseline for the interface.
The big news here, of course, is raw bandwidth. The current versions of DisplayPort – 1.3 & 1.4 – offer up to 32.4 Gbps of bandwidth – or 25.9 Gbps after overhead – which is enough for a standard 16.7 million color (24-bit) monitor at up to 120Hz, or up to 98Hz for 1 billion+ (30-bit) monitors. This is a lot of bandwidth, but it still isn’t enough for the coming generation of monitors, including the likes of Apple’s new 6K Pro Display XDR monitor, and of course, 8K monitors. As a result, the need for more display interface bandwidth continues to grow, with these next-generation monitors set to be the tipping point. And all of this is something that the rival HDMI Forum has already prepared for with their own HDMI 2.1 standard.
DisplayPort Signaling Standards
Target Monitor Resolutions
DP 1.0/1.1 (HBR1)
8K@60hz HDR >8K@60Hz SDR
DisplayPort 2.0, in turn, is shooting for 8K and above. Introducing not just one but a few different bitrate modes, the fastest mode in DisplayPort 2.0 will top out at 80 Gbps of raw bandwidth, about 2.5 times that of DisplayPort 1.3/1.4. Layered on that, DisplayPort 2.0 also introduces a more efficient coding scheme, resulting in much less coding overhead. As a result, the effective bandwidth of the new standard will peak at 77.4 Gbps, with at 2.98x the bandwidth of the previous standard is just a hair under a full trebling of available bandwidth.
Putting all of this in practical terms for a moment, for the VESA and its member manufacturers then, the new standard opens the door to higher resolution, higher refresh rate, and wider color gamut monitors. DisplayPort 2.0 is fast enough to not just support an 8K monitor without any kind of compression (including chroma subsampling), but it’s enough to do so at 30-bit color, allowing for HDR support even at that high of a resolution. Similarly, 10K monitors at 24-bit color are now possible without compression, and 16K monitors with compression. Overall the applications are about as varied as manufacturers want to go, with options ranging from ensuring there’s sufficient bandwidth for next-gen VR to enabling new daisy chaining setups (daisy chained 5K monitors, anyone?), not to mention even more interesting setups such as mixing USB data with high resolution DisplayPort monitors. Many things which the VESA is happy to point out that even HDMI 2.1 can’t do, due to the former’s significant bandwidth advantage (all of this solidly putting DisplayPort back on top in terms of total bandwidth).
DisplayPort 2.0 Under the Hood: Thunderbolt 3, UHBR, & Passive Cables
Diving a bit deeper into today’s announcement, let’s talk about the DisplayPort 2.0 physical layer. For the last half decade or so, VESA members have commented here and there about where the standard might go in the future, and what direction the physical ports would take. Developing the next generation of high bandwidth external interfaces only gets harder and more expensive with each generation, which has increasingly caused the various standard bodies to coalesce around a handful of physical layers and data transmission technologies. At the same time, the physical DisplayPort, which was designed over a decade ago, wasn’t originally designed to scale up to the amount of bandwidth DisplayPort 2.0 will be pushing. As a result every option has been on the table to some degree, including disregarding most of the DisplayPort standard, DisplayPort and all.
The end result then is an interesting compromise, and importantly, one that delivers more bandwidth while retaining backwards compatibility with existing DisplayPort gear. The DisplayPort itself is staying: it and the USB-C connector (via DP alt mode) are both official ports for the new DisplayPort 2.0 standard. And because of this, the number of pins and resulting high speed data lanes is remaining unchanged as well, with DisplayPort continuing to operate over 4 lanes. Finally, the DisplayPort 2.0 standard also retains the technology’s packet-based approach to communications, which means that image data continues to be sent as packets over a fixed bandwidth link, as opposed to pixel-centric pixel clock approaches.
So what has changed to enable DisplayPort 2.0? While the titular DisplayPort itself has stayed, the rest of the physical layer has been almost entirely replaced… with Thunderbolt 3.
Rather than attempting to reinvent the wheel, for DisplayPort 2.0 the VESA decided to take advantage of Intel’s existing Thunderbolt 3 technology, which already hits the data rates that the VESA was looking for. While initially a proprietary Intel technology, Intel released the technology to the wider industry as a royalty-free standard earlier this year. This allowed third parties to not only create pure Thunderbolt 3 devices without paying Intel, but also allowed Thunderbolt 3 technology to be repurposed for other standards. So whereas USB4 is a more straightforward rebranding of Thunderbolt 3, for DisplayPort 2.0 takes it in a different direction by essentially creating a one-way Thunderbolt 3 connection.
Under the hood, Thunderbolt 3 operates fairly similarly to DisplayPort, with 4 high-speed each lanes carrying packets of information at 20 Gbps. However while TB3 is a true bi-directional, full-duplex link with 2 lanes allocated for each direction, DisplayPort is focused on sending large volumes of data in just one direction: out. As a result, DisplayPort 2.0 reverses the two inbound lanes to outbound lanes, allowing the four total lanes to be combined into a single 80 Gbps link.
Thunderbolt 3 vs. DisplayPort 2.0
Max Cable Bandwidth
Max Channel Bandwidth
(Full Duplex, Bidirectional)
2x DP 1.4 Streams
1x DP 2.0 Streams
Passive Cable Option
Speaking of the link itself, the move to Thunderbolt 3 technology also means that DisplayPort inherits Thunderbolt 3’s signal encoding scheme. Whereas DisplayPort 1.x has always used relatively inefficient 8/10b encoding – resulting in 20% overhead – DisplayPort 2.0 will offer 128/132b encoding, which has just 3% overhead. This is why the practical bandwidth gains for DisplayPort 2.0 are more than just the raw bandwidth gains; the standard doesn’t just get more bandwidth, but it uses it more efficiently. Consequently, at its highest data rate, DisplayPort 2.0 will be able to offer 77.37 Gbps of bandwidth.
But what of cables? Here’s where things get a bit trickier, both for the VESA and for users. Thunderbolt 3 pushed the limits of copper cabling, and as a result for all but the shortest runs it requires active cabling, with transceivers at each end of a cable. While effective, this drove up the cost of Thunderbolt 3 cables versus relatively cheap all-copper commodity USB 3 and DisplayPort 1.x cables. By using Thunderbolt 3 as the basis of their new standard, the VESA has inherited the cable technology limits of the standard as well.
The answer to the cable question then is that the VESA hasn’t really answered it. Instead, they’re focusing on what they can do now with passive cables. All told, the DisplayPort 2.0 actually introduces not one, but three new data rates: 10 Gbps per lane, 13.5 Gbps per lane, and 20 Gbps per lane. Dubbed Ultra High Bit Rate (UHBR), the for free-standing monitors the VESA right now is focusing on 10 Gbps per lane (UHBR 10), which will deliver a total of 40 Gbps of bandwidth.
At just half the data rate of full-fat DisplayPort 2.0 (and Thunderbolt 3), UHBR 10 is resilient enough that it can operate over standard passive copper cabling, and cables should have little issue reaching 2-3 meters. The VESA has actually been preparing for this for some time now, and UHBR 10 aligns with their previously-launched DisplayPort 8K cable certification program; 8K-certified cables will be able to meet the signal integrity requirements for UHBR 10.
DisplayPort 2.0: UHBR Modes
Past that, however, the VESA isn’t currently exploring (or at least not focused on) passive cables for the higher bitrate modes. Instead, the group envisions UHBR 13.5 and UHBR 20 being tethered setups: manufacturers would ship devices with an appropriate port/cable already attached. These can potentially be passive cables for very short runs (think laptop docks), or integrated active cables for longer runs. I should note that the group hasn’t closed the door entirely to more traditional passive cable setups for these higher bitrates, but at least for the moment the group doesn’t see very many non-tethered use cases coming to market in the near future.
Tangentially, here, there is one more signal-related changed to the DisplayPort standard. Forward Error Correction (FEC), which was introduced to DisplayPort 1.4 as part of the Display Stream Compression (DSC) standard, is now a core part of DisplayPort 2.0. So on a 2.0 link, FEC will be in use at all times, reflecting the challenge in getting these high speed interfaces to constantly transmit data in an error-free manner.
Shifting gears, along with the significant physical layer changes being introduced in DisplayPort 2.0, the standard is also introducing some much more modest feature changes.
First and foremost, Display Stream Compression support is now mandatory for DisplayPort 2.0 devices. Previously introduced as part of DisplayPort 1.4 – and not really hammered out entirely until a couple of years after that – DSC is the group’s standard for “visually lossless” image compression. Operating on small groups of pixels, DSC offers modest compression ratios of around 3:1, with the goal of compressing images just enough to save power and bandwidth without introducing visual artifacts and without adding significant latency.
At any rate, starting with DisplayPort 2.0, DSC is now a core part of the DisplayPort standard. To be clear, 2.0 devices do not have to use DSC – the preference is clearly towards uncompressed images when the bandwidth allows for it – however 2.0 devices must be able to encode, pass, and decode DSC compressed data. This will, over time, lay the groundwork for manufacturers to develop and release monitors that require DSC (at least in certain modes), as they’ll be able to sell monitors knowing that all 2.0 devices can drive them.
Speaking of efficiency, the DisplayPort 2.0 standard is also introducing another feature focused on power efficiency, and that’s Panel Replay. Derived from earlier Panel Self Refresh technologies that are part of the embedded DisplayPort standard, Panel Replay is a partial self-update mechanism that allows a system to only transmit and update the portion of an image that has changed since the previous video frame. Like PSR in eDP, this feature is primarily intended for laptops and other mobile devices, where power consumption and the resulting impact to battery runtimes are important qualities. Transmitting less data reduces not only the amount of energy used chauffeuring bits around, but it also reduces the amount of processing required in a display controller.
Last but not least, DisplayPort 2.0 is also updating how “branch devices” work in the standard. Essentially the splitters in a Multi Stream Transport setup, DisplayPort 1.x required that the branch device be capable of decoding a DisplayPort bitstream, which is not an easy feat with 20 Gbps+ of data. So instead, for 2.0, branch devices are being simplified some, and now will just be able to forward data rather than having to decode it. This should make MST (and daisy chaining) a bit easier to implement overall, as branch devices won’t need to be as complex.
On a final note, ahead of today’s specification release I also asked about the state of variable refresh support on DisplayPort. VESA Adaptive Sync is an optional feature for monitors under DisplayPort 1.x, and it will remain so under DisplayPort 2.0. So manufacturers can continue adding it as a useful feature for their monitors, but there are no plans to make it mandatory.
Coming In Late 2020
Wrapping things up, the latest version of the DisplayPort standard is easily the biggest update to the PC display standard since it launched in 2007. By replacing the DisplayPort physical layer with Thunderbolt 3, the VESA has greatly increased DisplayPort’s bandwidth potential, laying the groundwork for 8K monitors and beyond. This update doesn’t come for free, and the VESA’s member companies will have to tackle the same kind of high-bandwidth cabling issues that Thunderbolt 3 itself had to address over the last few years, but ultimately it’s a situation that gives the display standard a major shot in the arm in terms of bandwidth, while pushing the PC industry ever so closer towards using a handful of common standards for all high bandwidth I/O.
As for the first retail products, the good news is that this will show up sooner than later. Because the new standard is based on the Thunderbolt 3 physical layer, member companies can hit the ground running on development and testing. As a result, the VESA expects the first retail devices to show up in the latter part of 2020, less than 18 months from now.
SK Hynix has announced it has finished development of its 128-layer 1 terabit 3D TLC NAND flash. The new memory features the company’s charge trap flash (CTF) design, along with the peripheral under cells (PUC) architecture that the company calls ‘4D’ NAND, announced some time ago. The new 128-layer TLC NAND flash devices will ship to interested parties in the second half of this year, and SK Hynix intends to offer products based on the new chips in 2020.
1 Tb 128-Layer ‘4D’ TLC NAND
SK Hynix’s 1 Tb 128-layer TLC NAND chip features four planes as well as a 1400 MT/s interface at 1.2 Volts. The quad-plane architecture along with a 1400 MT/s I/O bus will make the new TLC NAND devices not only significantly denser (in terms of Gb per mm2) than previous-generation products, but also at least 16% faster. In fact, real-world performance increase could be even higher as SK Hynix once said that its CTF design would enable a faster program time (tPROG) as well as a faster read time (tR).
To stack 128 layers inside its 6th Generation 3D NAND chips, the company had to use a multi-stacked design along with numerous new technologies, including ultra-homogeneous vertical etching technology as well as high-reliability multi-layer thin-film cell formation technology. Meanwhile, to increase performance of the I/O bus without increasing power consumption of the device, SK Hynix implemented its ultra-fast low-power circuit design.
It is noteworthy that transition to CTF + PUC architecture, along with various optimizations, enabled SK Hynix to reduce the number of process steps by 5% as well as increase bit productivity per wafer by 40% when compared to 96-layer TLC NAND. As a result, we should expect to see the new NAND run through the entire product lineup in due course.
SK Hynix and its partners will use the company’s 1 Tb 128-layer TLC NAND chip for a variety of applications.
“SK Hynix has secured the fundamental competitiveness of its NAND business with this 128-Layer 4D NAND,” said Executive Vice President Jong Hoon Oh, head of global sales & marketing. “With this product, with the industry’s best stacking and density, we will provide customers with a variety of solutions at the right time.”
Initial 1 Tb 128-layer TLC NAND chips that SK Hynix will start shipping in the second half of the year will be used primarily for mobile storage devices, such as USB drives and memory cards.
In the first half of next year SK Hynix promises to roll out its UFS 3.1 storage products based on the new 1 Tb devices. The company plans to offer 1 TB UFS 3.1 chips that will consume up to 20% less when compared to similar products that use 512 Gb ICs.
Later in 2020, SK Hynix intends to offer 2 TB client SSDs based on its own controller, as well as 16 and 32 TB server SSDs for datacenters ,.
176-Layer 4D NAND Incoming
String stacking technology, as well as the multi-stacked design, will enable SK Hynix to keep increasing the number of layers. SK Hynix says that it is currently developing 176-layer 4D NAND flash, but does not disclose when it is expected to become available.
One of the key elements of building a processor is that designing a secure product involves reducing the ‘attack surface’ as much as possible: the fewer ways an attack can get in, the safer your product is. For the white knights of the security world, when a vulnerability is found, the process usually goes through a period of reasonable disclosure, i.e. the issue is presented to the company, and they are often given a certain time to fix the issue (to help customers) before the full disclosure is made public (in case it might be swept under the rug). Using this method, a researcher at Google found a vulnerability in the way AMD’s EPYC processors provide Secure Encrypted Virtualization (SEV) which would allow an attacker to recover a secure key that would provide access between previously isolated VMs on a system. AMD has since released an update to the firmware which patches this issue.
Ever since Huawei was placed on the United States’ Entity List – effectively banning them from receiving US technology – the fate of support for their existing products has been in limbo. Would shipping firmware updates, security updates, and new OSes to these products violate the requirements of the list? At the time the ban went into effect, this is a question that no one really had a good answer for, either with suppliers or Huawei itself.
But now we’re finally getting some clarity on the issue. On the mobile side of matters, earlier this month Huawei formally announced that its latest two generations of smartphones will get Android Q and other important updates. Meanwhile on the PC side of matters, this week Microsoft and Intel are delivering their own update on the situation. In short (and like the smartphone situation), the two companies will continue providing updates for Huawei’s laptops.
In statements given to PCWorld, the two companies confirmed that they will still be able to offer support to customers with Huawei laptops. Intel, for example, will offer UEFI/BIOS firmware updates that will patch potential security risks of its CPUs (e.g., exploits with Zombieload) as well as drivers for its platforms (e.g., GPU drivers, etc.). Meanwhile Microsoft will naturally offer ongoing updates for its Windows operating system.
“We remain committed to providing exceptional customer experiences. Our initial evaluation of the U.S. Department of Commerce’s decision on Huawei has indicated we may continue to offer Microsoft software updates to customers with Huawei devices.”
Ultimately this resolution takes care of Huawei’s existing laptops, including those currently in use and remaining inventory still to be sold. Which is incredibly important, as it ensures these devices receive necessary security updates over their lifetimes and don't become easy targets for malware and other attacks. However it doesn't do anything to address the company's inability to build new computers; once the company exhausts its stockpile of CPUs and other components, Huawei will still effectively be on pause as a PC maker so long as the export ban stands.
Back in the late 1980s and early 1990s, both tower cases and lay flat desktop chassis were fairly common. Eventually tower cases prevailed, whereas desktops that could be used as pedestals for monitors remained on niche markets. The flat desktop case design has a lot of advantages for enthusiasts according to Antec, who has decided to go this route with its new Blazer GT design.
Stylized after jet engines used on fighters according to Antec, the Blazer GT is a semi open case that can be positioned in three possible ways: horizontally, vertically on the left side, and vertically on the right side. The chassis has multiple windows made of tempered glass and can fit it up to eight fans to provide air cooling.
When it comes to compatibility, the Antec Blazer GT can house an ATX, a Micro-ATX, or a Mini-ITX motherboard, as well as an air cooler that is up to 165 mm in height. Meanwhile, despite its dimensions, the case can only fit two 2.5-inch or 3.5-inch storage devices.
The key advantage of Antec’s Blazer GT is easy access to components, something that is very important for enthusiasts. Of course, it takes a lot of desktop space when positioned horizontally. Antec yet has to decide about pricing of the Blazer GT chassis. In fact, to some degree this device is considered as a concept and it remains to be seen when and if it set to become available and for how much.
Want to keep up to date with all of our Computex 2019 Coverage?
Low-profile graphics cards are primarily used by owners of small form-factor desktops who want to have a more or less decent graphics that can run casual video games. GPU manufacturers tend to release new low-profile products with every generation of GPUs, and in the recent weeks we encountered two cards based on NVIDIA’s GeForce GTX 1650: one from MSI and one from ZOTAC.
Specification wise, MSI’s GeForce GTX 1650 4GT LP and ZOTAC’s Gaming GeForce GTX 1650 Low Profile are very similar. Both cards are based on NVIDIA’s TU117 GPU featuring 896 CUDA cores clocked at up to 1665 MHz that is accompanied by 4 GB of GDDR5 memory. Both cards consume up to 75 W of power and therefore do not need any auxiliary PCIe power connectors, which makes them compatible with entry-level desktops from large OEMs that usually do not have any spare power cables inside.
The key difference between low-profile GeForce GTX 1650 graphics cards from MSI and ZOTAC is configuration of their display outputs. The MSI unit has a DVI-D and an HDMI 2.0b, whereas the Zotac comes with a DVI-D, a DisplayPort 1.4, and an HDMI 2.0b connector. Since the TU117 graphics processor supports hardware-accelerated decoding and encoding of HEVC (H.265) and VP9 video at 4Kp60 as well as HDR10, both cards can be used for HTPCs.
MSI’s GeForce GTX 1650 4GT LP and ZOTAC’s Gaming GeForce GTX 1650 Low Profile AIBs are already listed by their manufacturers on their websites, and so we expect the products to show up in retail shortly. In fact, since ZOTAC's unit was demonstrated at Computex, this board may be a little bit closer to release than MSI's device.
Neither MSI nor ZOTAC disclose MSRPs of their low-profile graphics cards, but it is logical to expect these entry-level units to be available at prices very close to MSRP recommended by NVIDIA, which is $149.
Amazon Japan has started to take pre-orders on Cosair’s upcoming PCIe 4.0 NVMe SSD. The MP600 SSDs will ship in mid-July, several days after AMD and its partners start to sell platforms that support the PCIe 4.0 interface.
At press time, Amazon Japan is offering Corsair MP600 M.2-2280 SSDs with a PCIe 4.0 x4 interface featuring 1 TB and 2 TB capacity for ￥36,936 ($320) and ￥66,852 ($579), respectively. Considering the fact that we are talking about ultra-high-end client SSDs aimed at expensive desktops, their prices are not exactly surprising. Meanwhile, keep in mind that these are not their official MSRPs.
According to Corsair, its MP600 SSDs will offer up to 4950 MB/s sequential write speed as well as up to 4250 MB/s sequential read speed when used with a PCIe 4.0 x4 interface, which is substantially faster when compared to modern PCIe 3.0 x4 drives.
Corsair’s MP600 drives are based on Phison’s PS5016-E16 controller as well as 3D TLC NAND memory. Considering that Phison usually sells controllers and NAND flash memory as a turnkey solution, expect other suppliers of SSDs to offer drives very similar to the MP600 family by Corsair at around the same timeframe. In the meantime, Corsair definitely deserves a credit for starting to offer its PCIe 4.0 SSDs ahead of competitors.
We saw a number of PCIe 4.0 drives at Computex this year, so we expect the market to have a sizeable number available in due course:
During Computex 2019, Noctua announced its plans to update its fabled D series of CPU coolers with a new design. The new, currently-unnamed coolers incorporate an extra heat pipe as compared to the current NH-D15 and NH-D15S, as well as increase performance and utilize an asymmetrical design for improved PCIe slot clearance.
Noctua of course is well known in the air cooling market for its high quality fans and coolers, which are aimed at the premium end of the market and deliver some of the best performance on the on the market today. Not hurting the company's visibility either is their highly distinctive beige and brown color scheme, which make their products easy to tell apart from more commodity products.
At Computex in addition to its fanless concept CPU cooler design, Noctua also had its next generation of D series CPU cooler on display. The new cooler features seven heat pipes – up from the traditional six of the previous NH-D15 models – and the heatsink itself has been tweaked to offer 10% more surface area for heat dissipation than previous models. One of the most interesting aspects is that Noctua revealed that the new D series coolers are designed to dissipate up to 400 W of heat, which makes them highly suitable for beefier sockets such as AMD's TR4 and SP3.
Showing off one of the prototypes in action, the company had one of the new coolers setup against one of its current NH-D15 models in a custom designed test chamber, using the two otherwise-identical test systems to show the differences in performance between the two models. And since this was designed to be a test of the heatsink and not the fans, only the heatsink itself was swapped here; both systems used the same NF-A15 PWM cooling fans.
The new D series cooler is on the left, the pre-existing NH-D15 is on the right
Overall we didn't see a massive difference – and to be fair, it was very hot (ed: Taipei in June), so ambient temperatures were running high – but none the less, the prototype cooler did come out consistently ahead of the NH-D15. In its custom chamber, the new D series cooler had one degree advantage over its predecessor.
Shifting gears to retail matters, Noctua is planning on releasing two versions of the new D series coolers, rolling out both a single fan model and a dual fan model. The new coolers will include a traditional multi-socket mounting kit, which will support AM4, LGA20xx, and LGA115x. Meanwhile there will be separate dedicated TR4 models due to the larger base plate needed to properly match the large IHS used on AMD's Threadripper processors. Each cooler will also be bundled with a tube of its new NT-H2 thermal compound.
At this point the company hasn't announced an official release date, but it's likely that the next generation of D-type coolers will be available by the end of the year, with a pricing structure similar to that of the current NH-D15 and NH-D15S models.
Want to keep up to date with all of our Computex 2019 Coverage?