Last month Samsung Foundry quietly announced that it was set to begin producing chips using its 3GAE (3 nm-class, gate-all-around transistors, early) process technology in the second quarter. While the industry’s first 3 nm-class node with GAA transistors is a noteworthy achievement by itself, one thing that is particularly important is that to make GAA transistors efficiently, fabs have to be equipped with new production tools. Fittingly, Applied Materials has recently outlined their next generation of tools that will be used to enable Samsung (and other fabs) to build their first GAA chips.

Gate-All-Around Transistors: Solving Many Problems at Once

New process technologies should enable higher performance, lower power, and higher transistor density to meet requirements of chip designers. But this combination has been particularly hard to achieve in the recent years as shrinking transistor sizes lead to negative effects like leakage current. To keep scaling performance and voltages while minimizing transistor sizes, the industry began transition from planar to FinFET transistors (which increased the contact area between the channel of the transistor and the gate by making the gate taller) in 2012 and this type of transistors will stay with us for a long time. But the pace of FinFET-based nodes innovation is slowing down as negative aspects of transistors shrinking starting to kick in.

Ever since Intel introduced its 22 nm FinFET-based technology over a decade ago, chipmakers began to talk about the next inevitable step in transistor design, which is gate-all-around. As the name suggests, in GAAFETs channels are horizontal and are surrounded by gates around all four sides of the channels, which solves many problems associated with the leakage current. But this is not the only advantage: in nanosheet/nanoribbon-based GAAFETs it is possible to adjust channel width to get higher performance or reduce power consumption. Since Samsung’s 3GAE and 3GAP technologies use nanosheets, Samsung calls its GAAFETs multi-bridge channel field-effect transistors (MBCFETs) to emphasize that it does not use nanowires. Furthermore, Applied Materials claims that GAA architecture reduces transistor variability, which means improved yields and faster time-to-yield (at least as far as transistor performance is concerned) and reduces cell area by 20% to 30%.

While all chipmakers have been talking about GAAFETs and their advantages over FinFETs at various academia industry events, Samsung was the first company to announce a transition to the new type of transistors with its 4 nm node sometimes in 2022 ~ 2023. Eventually, the company’s plans changed a bit and in 2019 it unveiled its GAAFET-based 3GAE and 3GAP nodes with high volume manufacturing due in 2022 and 2023, respectively. Last year the company re-emphasized plans to kick off production using its 3GAE in 2022 and this quarter it said that volume production was set to start in Q2.

But Manufacturing Is Challenging

But production of chips using any leading-edge process technology is challenging and making products using Samsung’s MBCFET-based 3GAE node certainly poses some risks not only because we are talking about a 3 nm-class manufacturing process, but because manufacturing of GAA transistors is different than making FinFETs.

Channels of GAA transistors are shaped using lithography, epitaxy and selective materials removal, processes that enable chipmakers to fine-tune width and uniformity for optimum power and performance. Those epitaxial steps are considerably more complex when compared to those with FinFETs as manufacturers need to deposit the multi-layer gate oxide and metal gate stacks around the tiny 10 nm channels, which is a key challenge with GAAFETs. This is where Applied’s high-vacuum Integrated Materials Solution (IMS) systems come into play. 

Firstly, Applied’s Producer Selectra Selective Etch IMS tool removes unnecessary SiGe to isolate the gate from the source/drain and define channel width without damaging surrounding materials. Secondly, the Applied Centura Prime Epi IMS tool can deposit an ultra-thin gate oxide layer using integrated atomic layer deposition (ALD), thermal, plasma treatment, and metrology steps. Applied’s IMS can reduce gate oxide thickness by 1.5 angstroms compared to competing solution, which allows to make gates thicker and enable higher drive current and therefore transistor performance without increasing current leakage, which is a consequence of thin gates.

Applied introduced its first Selectra etching system in 2016 and since then has shipped over 1,000 chambers that are currently in use by its customers, so the industry is quite aware how to use them. While etching is a widely used process, its usage and importance will grow as the industry begins to transit to newer process technologies, including those based on GAA transistors. 

It is necessary to note that numbers shared by Applied Materials come from various companies (including TSMC and Qualcomm) and should not be applied exclusively to Samsung Foundry’s 3GAE and 3GAP nodes. Meanwhile, the general advantages of GAA/MBC/nanoribbon/nanosheet transistors, as well as challenges with their manufacturing, are similar for the whole industry.

Samsung’s 3GAE: +30% Performance, -45% Area

Speaking of advantages, Samsung expects its 3GAE process to offer a 30% performance increase or a 50% power consumption reduction, and a 45% area reduction when compared to 7LPP. Such tangible benefits may be enough to justify challenges associated with this transition in the short term.

For now, the use of 3GAE manufacturing technology that relies a brand-new transistor structure is challenging in general, as in addition to new electronic design automation (EDA) tools, Samsung will need to invent new IP, and adopt brand-new design rules. All of these things, when combined with new lithography, new etching, and new deposit steps are going to make Samsung’s transition difficult. 

But in exchange, it will start gaining experience with GAA transistors years ahead of Intel and TSMC, which may bring certain gains in the long-term future. Furthermore, experience with GAAFETs will be beneficial for Samsung’s DRAM manufacturing as memory greatly benefits from smaller cell sizes and higher transistor densities. 

Read MoreAnandTech

Western Digital is announcing the sampling of its new 22TB CMR and 26TB SMR hard drives today at its What’s Next Western Digital Event. As usual, the hyperscale cloud customers will get first dibs on these drives. The key takeaway from today’s presentation is that Western Digital doesn’t yet feel the need to bring heat-assisted magnetic recording (HAMR) into the picture. In fact, WD is doubling down on energy-assisted PMR (ePMR) technology and OptiNAND (introduced first in the 20TB CMR drives). WD is also continuing to use the triple-stage actuator that it started shipping in the first half of 2020 in the new drives. It goes without saying that the new high-capacity drives are helium-filled (HelioSeal technology). The main change common to both drives is the shift to a 10-stack design.

The SMR drives are getting an added capacity boost, thanks to WD’s new UltraSMR technology. This involves adoption of a new advanced error correction algorithm to go along with encoding of larger blocks. This allows improvement in the tracks-per-inch (TPI) metric, resulting in 2.6TB per platter. The new Ultrastar DC HC670 uses ten platters to provide 26TB of host-managed SMR storage for cloud service providers.

ArmorCache Technology

As part of the announcements, WD also provided additional details on the caching feature enabled by OptiNAND – ArmorCache. Last year’s announcement was quite light in terms of actual performance numbers, but the 20TB OptiNAND drives have been out in the market for a few quarters now.

The UFS-based iNAND package helps the OptiNAND-based HDDs deliver upwards of 80% improvement in IOPS for low-queue depth large-sized random writes (10% – 20% improvement for small-sized writes) in use-cases where the write cache is disabled. Since non-OptiNAND HDDs can only cache whatever can be flushed out to the serial flash during emergency power-off situations, the larger cache size afforded by the iNAND device relaxes this limitation considerably. As discussed in the OptiNAND announcement coverage last year, use-cases with write caching enabled benefit from the EPO data protection afforded by the iNAND device. Enabling write caching often requires hosts to send out cache flush commands to the HDDs. These commands require the HDD to stop accepting new commands until the completion of the flush, resulting in loss of performance. The ArmorCache makes these flush commands unnecessary, allowing the drive to be used with full performance with no risk of data loss due to EPO.

Concluding Remarks

The 22TB Ultrastar DC HC570 and 26TB Ultrastar DC HC670 are currently sampling to hyperscalers. Volume shipment of the CMR drive to the channel is set for the next quarter. The Ultrastar HC HC670 is a host-managed SMR drive, and will hence ship only to select customers around the same timeframe. Western Digital will be delivering variants of the CMR drive across its HDD portfolio – 22TB WD Purple Pro for surveillance NVRs, WD Red Pro for NAS systems, and WD Gold for SMB and enterprise customers in summer.

The updated capacity points – in particular, the jump in the SMR drive capacity – delivers clear TCO benefits to WD’s cloud customers. Crucially, WD believes it has enough trust in its ePMR setup to deliver 30TB+ HDDs without having to go the HAMR route. From a technology perspective, this will make the upcoming roadmap / product announcements from Seagate (HAMR deployment for publicly-available drives) and Toshiba (need for any technology beyond FC-MAMR) interesting to watch.

Read MoreAnandTech

The last few months have seen multiple vendors launch products in the Wi-Fi 7 (802.11be) space. Mediatek demonstrated working 802.11be-compliant silicon under the Filogic lineup earlier this year. Concrete technical details and part numbers were not announced during the event. Last month, Broadcom introduced a comprehensive Wi-Fi 7 portfolio, detailing multiple 802.11be radios for access points, a networking SoC built keeping 802.11be bandwidth in mind, and a client radio for mobile applications.

Wi-Fi 7 / 802.11be Background

The 802.11 Working Group’s focus with 802.11be has been on extremely high throughput. This has been achieved primarily through a combination of three different aspects:

Support for up to 16 spatial streams
Support for channel widths up to 320MHz (with operation in 2.4 GHz, 5 GHz, and 6 GHz bands)
Support for 4096-QAM (4K-QAM) resulting in better utilization of available spectrum (a faster modulation / coding scheme).

It must be noted that wider channels are available only in the 5 GHz and 6 GHz bands.

Theoretically, these aspects allow for up to around 46 Gbps of wireless throughput. 802.11be also aims to enable usage of Wi-Fi for real-time applications by including features for low-latency communications such as Multi-link operation (MLO). This allows a client and an access point to simultaneously communicate over multiple channels that might even belong to different bands.

The 802.11be specifications also allow for better performance in environments with heavy interference by permitting adaptive puncturing – the desired channel need not be necessarily contiguous. This improves on the preamble puncturing feature already available as part of the 802.11ax specifications.

Interference and co-existence with non-Wi-Fi users of the same spectrum is handled using automatic frequency coordination (AFC). Unlike the Open AFC initiative (of which Broadcom is a part), Qualcomm is opting to provide its own turnkey solution for its customers.

As expected, AFC will require the device agent to be connected to the Internet for channel configuration and power modulation purposes.

Qualcomm’s Wi-Fi 7 Portfolio

At MWC 2022, Qualcomm had provided details of their 802.11be client silicon targeting mobile devices. The FastConnect 7800 is expected to become available in H2 2022, and integrates Bluetooth 5.3 support with key Wi-Fi 7 features. Today, the company is introducing its Wi-Fi 7 Networking Pro Series access point platforms to expand its Wi-Fi 7 portfolio.

Qualcomm started marketing its Wi-Fi solutions under the Networking Pro and FastConnect monikers back in 2019, with the launch of its Wi-Fi 6 solutions. The 2nd generation products catered to the Wi-Fi 6E ecosystem, and it comes as no surprise that today’s 3rd Gen. launch focuses on Wi-Fi 7.

The introduction of the Networking Pro tag has contributed to opaqueness in the composition of the router / AP platforms from a end-consumer perspective. Unlike Broadcom’s public announcement of core details of the networking SoC, and characteristics of the various radio options for its reference designs, Qualcomm provides these details only to their customers. Eventually, these details do become public after market availability of the products. For example, the Networking Pro 800 platform’s breakdown is available here – we see a core WiSoC with integrated MAC and baseband coupled with discrete radios. Therefore, it is disappointing that Qualcomm has not given us much to analyze in today’s announcements beyond basic product specifications. These high-level feature specifications match the features in Broadcom’s products announced last month.

It must be noted that Qualcomm also provides the ‘Immersive Home’ platform for entry-level Wi-Fi routers. These are typically similar to the ‘Networking Pro’ platforms, except for cut-down stream counts and reduced WiSoC capabilities in terms of CPU core counts and frequencies. Given that Qualcomm refused to provide any details of CPU core counts or frequencies for the Networking Pro series being launched today, we need to wait for an official announcement related to their Wi-Fi 7 portfolio for the cost-sensitive segment.

The 3rd Gen. Networking Pro series consists of four members, with stream counts ranging from 6 in the Networking Pro 620 to 16 in the Networking Pro 1620. Radio options exist for two or four streams in each of the 2.4 GHz, 5 GHz, and 6 GHz bands.

Qualcomm claims peak system PHY rates of up to 33 Gbps, and per-channel wireless PHY rates of more than 10 Gbps, with the capability of supporting more than 500 users in each channel.

The availability of 6 GHz spectrum for Wi-Fi is not universal. Qualcomm indicated that different configurations of the Networking Pro 3rd Gen series are possible, based on the allocation.

Qualcomm’s Wi-Fi 7 Networking Pro series will allow its customers to create a wide range of enterprise-class and premium wireless access points / routers. The platforms are already sampling – we should be seeing market availability of BE10000, BE16000, BE21000, and BE33000 Wi-Fi routers based on them in a few quarters.

5G-based fixed wireless access is fast becoming an attractive option for many consumers – sometimes even as the primary broadband connection. Qualcomm’s strengths in 5G will act as an asset from a bundling perspective for integrated Wi-Fi 7 / 5G platforms. However, Broadcom has significant presence with service providers for 10G PON and DOCSIS 4.0. It is likely that those high-speed ISPs will find Broadcom’s possible bundled offerings attractive for their Wi-Fi 7-capable consumer gateways. While we wait for Mediatek and MaxLinear (to whom Intel sold their Home Gateway Platform Division) to reveal their Wi-Fi 7 cards fully, it appears that initial battle will be between products based on Qualcomm’s 3rd Gen. Networking Pro series and those using Broadcom’s Wi-Fi 7 portfolio.

Read MoreAnandTech