A critical security flaw known as "Sinkclose" (CVE-2023-31315) has been identified in all AMD processors dating back to 2006, potentially affecting hundreds of millions of devices worldwide. This vulnerability allows malicious actors to exploit the chip architecture, leading to unauthorized access to sensitive data. Researchers Enrique Nissim and Krzysztof Okupski, researchers from the security firm IOActive, have revealed that the vulnerability can be exploited through various methods, enabling attackers to extract confidential information from affected systems, including passwords and personal data. The issue is especially concerning, given that it is present in all AMD CPUs made in the last 18 years and their widespread use in both consumer and enterprise environments. However, to exploit this vulnerability, an attacker must possess access to system's kernel. Downloading of malware-infused files can trigger it, so general safety measures are recommended.

The Sinkclose method exploits a little-known capability in AMD processors called TClose. This name is a blend of "TClose" and "Sinkhole," with the latter referring to a previous vulnerability found in Intel's System Management Mode in 2015. AMD chips employ a protective mechanism named TSeg, which blocks operating systems from accessing a specific memory area reserved for System Management Mode (SMM), known as System Management Random Access Memory (SMRAM). However, the TClose feature is designed to maintain backward compatibility with older hardware that might use the same memory addresses as SMRAM. It does this by remapping memory when activated. The security experts discovered that they could manipulate this TClose remapping function using only standard operating system permissions. By doing so, they could deceive the SMM into retrieving altered data, enabling them to redirect the processor and run their own instructions with the high-level privileges of SMM. This technique essentially allows attackers to bypass standard security measures and execute malicious code at one of the most privileged levels of the processor, potentially compromising the entire system.

The expansion of AI technology affects not only the production and demand for graphics cards but also the electricity grid that powers them. Data centers hosting thousands of GPUs are becoming more common, and the industry has been building new facilities for GPU-enhanced servers to serve the need for more AI. However, these powerful GPUs often consume over 500 Watts per single card, and NVIDIA's latest Blackwell B200 GPU has a TGP of 1000 Watts or a single kilowatt. These kilowatt GPUs will be present in data centers with 10s of thousands of cards, resulting in multi-megawatt facilities. To combat the load on the national electricity grid, US President Joe Biden's administration has been discussing with big tech to re-evaluate their power sources, possibly using smaller nuclear plants. According to an Axios interview with Energy Secretary Jennifer Granholm, she has noted that "AI itself isn't a problem because AI could help to solve the problem." However, the problem is the load-bearing of the national electricity grid, which can't sustain the rapid expansion of the AI data centers.

The Department of Energy (DOE) has been reportedly talking with firms, most notably hyperscalers like Microsoft, Google, and Amazon, to start considering nuclear fusion and fission power plants to satisfy the need for AI expansion. We have already discussed the plan by Microsoft to embed a nuclear reactor near its data center facility and help manage the load of thousands of GPUs running AI training/inference. However, this time, it is not just Microsoft. Other tech giants are reportedly thinking about nuclear as well. They all need to offload their AI expansion from the US national power grid and develop a nuclear solution. Nuclear power is a mere 20% of the US power sourcing, and DOE is currently financing a Holtec Palisades 800-MW electric nuclear generating station with $1.52 billion in funds for restoration and resumption of service. Microsoft is investing in a Small Modular Reactors (SMRs) microreactor energy strategy, which could be an example for other big tech companies to follow.

As new applications, use cases and connected devices multiply, Western Digital's enterprise-class hard disk drives (HDDs) remain critical for helping data center customers design more cost-efficient, scalable and sustainable infrastructure. When combined with the emergence of generative AI and the multitude of innovative applications it is enabling, the demand for high-capacity, low-power and reliable storage is poised to surge.

Western Digital leverages numerous technology innovations - ePMR, OptiNAND, UltraSMR and more - to deliver predictable capacity growth in scalable, highly reliable products. Western Digital is now shipping its new 10-disk 24 TB CMR HDD family for hyperscale, cloud and enterprise data center customers. Sampling since August, the new 28 TB SMR HDD is also ramping, adding to the company's SMR leadership position and momentum as its 26 TB SMR HDD exabyte shipments reached nearly half of its data center exabytes shipped in the first quarter fiscal year 2024.

Building on significant technology innovations cultivated over decades, Western Digital today announced that it is sampling its new industry-leading 22 TB and 26 TB UltraSMR HDDs to select hyperscale cloud customers, further expanding its areal density leadership and delivering customer value by driving TCO lower. Western Digital's HDD technology portfolio and areal density leadership puts it at the center of storage innovation. Leveraging its unique OptiNAND technology, energy-assisted PMR (ePMR), triple-stage actuator (TSA), HelioSeal and now UltraSMR technologies, Western Digital is on a clear path to delivering 30+ TB with ePMR.

"As a longstanding partner of the industry's leading cloud providers, we understand their unique requirements in building next-generation cloud infrastructure and invested in several HDD innovations we developed alongside our areal density technology," said Ashley Gorakhpurwalla, EVP and GM, HDD Business Unit, Western Digital. "Our intent with taking this development strategy was not only to address the capacity demands of the world's largest cloud titans but to deliver on a roadmap that would also support the evolving economics of their data centers for decades to come. With ePMR, OptiNAND and now UltraSMR as the foundation of Western Digital's HDD roadmap, our cloud customers can continue rapidly scaling their business by taking advantage of our innovations to lower their storage TCO."

Samsung and Western Digital today signed a memorandum of understanding (MOU) that implies that the two companies will join their efforts to create next-generation data placement, processing, and fabrics (D2PF) storage technologies. The MOU states that the two companies will collaborate on creating Zoned Storage devices, including ZNS (Zoned Namespaces) SSDs and SMR (Shingled Magnetic Recording) HDDs. The ZNS SSD initiative is a challenge of creating entirely new SSDs that work on the ZNS principle instead of traditional block-based drives. Opposite to data in traditional SSDs written in blocks, ZNS SSDs will place data into zones and have software understand where each bit of information is written without excessive read/write action. Another benefit of ZNS is that managing data allows less garbage collection, thus higher efficiency.

Software adoption for ZNS makes it hard, and Samsung's conjoined efforts with Western Digital aim to simplify it. Western Digital notes that "In addition, this collaboration is expected to serve as a starting point to expand zone-based (e.g. ZNS, SMR) device interfaces, as well as future-generation, high-capacity storage devices with enhanced data placement and processing technologies. At a later stage, these initiatives will be expanded to include other emerging D2PF technologies such as computational storage and storage fabrics including NVMe over Fabrics (NVMe-oF)." So, with further development of ZNS, Samsung and Western Digital will collab on more advanced storage options as well.

Empowering the world's most essential data infrastructures, Western Digital Corp. today announced that Dropbox Inc., a leading global collaboration platform, is one of the first to qualify the Ultrastar DC HC650 20 TB, host-managed, shingled magnetic recording (SMR) hard disk drives (HDD). With Western Digital SMR HDDs serving as the storage foundation for its custom-built, multi-exabyte storage platforms, Dropbox continues its strategic path, taking advantage of the highest storage densities with the lowest TCO without sacrificing data durability and availability for its 600 million+ online customers.

"We've been working with Western Digital for many years, and throughout several phases of our storage platform development," said Andrew Fong, vice president of engineering, Dropbox. "Online tools and services are more important than ever before, so we need to be able to quickly innovate, expand functionality and scale to help stay competitive. We look forward to deploying these higher capacity 20 TB SMR hard drives to further our cost savings and to provide our customers with even more value. Western Digital is an important strategic partner and we look forward to continuously pushing cloud infrastructure boundaries together."

Western Digital has been hit by a class-action lawsuit over alleged false advertising over some of its WD Red hard drives featuring undocumented DM-SMR (drive-managed shingled magnetic recording), a physical layer data writing technique that maximizes capacity at heavy costs of random write performance, that effectively render the drives unfit for RAID applications that are common with NAS setups. Trouble brewed for Western Digital in April, as a Blocks & Files report exposed presence of SMR on certain popular WD Red drives as an explanation as to why the drives couldn't be added to RAID volumes.

Western Digital initially tried to defend its position by explaining what DM-SMR is, that the standard WD Red has been tested on most SOHO NAS devices, that they're not meant for serious (> 8-drive) NAS setups; and pointing people to their pricier WD Red Pro, or enterprise-segment Ultrastar HDDs, leading to more community backlash. Days later, Western Digital finally came out with a list of its HDDs that use SMR. Bellevue Washington-based Hattis & Lukacs, class-action specialists, are leading the charge against Western Digital, and are inviting aggrieved U.S. residents to join the class.

The introduction of Shingled Magnetic Recording (SMR) technology has enabled HDD manufacturers, such as Toshiba, to increase the capacity of their spinning platter drives beyond that of existing approaches. SMR technology is recognized as having an impact on write-speeds in drives where this technology is used, especially in the case of continuous random writing. For this reason, Toshiba products are carefully tailored to specific workloads and use cases.

For example, in use cases such as network-attached storage (NAS), where continuous random writing regularly occurs, Toshiba's current product line for consumers features the N300, which does not use SMR.

Western Digital on Thursday (23/04) released an updated statement about the company's stand on SMR (shingled magnetic recording) being used on certain internal hard drives, including certain client-segment WD Red models recommended by the company for NAS applications, without proper disclosure on the product's marketing materials or data-sheets. It's surprising to note that SMR is being used not in some of the higher-capacity models (8 TB or higher) as previously thought, but rather lower-capacity ones, 6 TB or lower, including a 2 TB 3.5-inch drive, and a 1 TB 2.5-incher.

Perhaps SMR is being used in these lower-capacity drives to reduce the number of platters, by cramming in more data per recording surface. As on 22nd April (when the list was internally compiled by Western Digital), there are 9 client-segment hard drives that use SMR, four of which are from the company's WD Red family, two each from the WD Blue 3.5-inch and WD Blue 2.5-inch families; and one from the 2.5-inch WD Black family. Among the WD Red series drives with SMR are the 2 TB WD20EFAX, the 3 TB WD30EFAX, the 4 TB WD40EFAX, and the 6 TB WD60EFAX.

Earlier this week, Western Digital was in the line of fire when it emerged that several of its WD Red family of "NAS-optimized" hard drives allegedly employ some form of shingled magnetic recording (SMR), a physical-layer data writing technique that maximizes density at a heavy cost of random write performance that effectively makes the HDDs unfit for use in RAID volumes, and in turn most NAS applications that commonly employ RAID and encourage end-users to build RAID volumes for data redundancy. A new report by Blocks & Files finds that the issue is more widespread than previously thought, and that even Seagate employs it without disclosure on certain models.

Several of Seagate's higher capacity Barracuda desktop internal hard drives use SMR without disclosing it in their data-sheets. These include the 8 TB ST8000DM004, and 5 TB ST5000DM000. Both these drives are sold under the Barracuda Compute brand, which markets "home servers, entry-level DAS, and desktop computers" among its use-cases. Seagate does market its Archive and Exos lines of HDDs are employing SMR, but mention of the technique is buried in their data-sheets, and not prominently in product marketing or packaging. Archive and Exos are targeted at bulk cold storage setups where capacity is more important than performance. Meanwhile, Toshiba has confirmed that its Desktop HDDs too employ SMR. The company's MQ04 2.5-inch and DT02 3.5-inch HDDs employ "managed SMR" (i.e. use conventional recording and switch to SMR as the platters are running out of space).

Seagate announced its roadmap for the coming years, and the company is naturally fighting tooth and nail for the relevance of HDD technology in the market. While the benefits of SSDs are already well understood by the entire industry, in some scenarios, it makes more sense to make use of high-density HDDs - particularly where deployment space is at a premium, and in scenarios where seek times for information stored on the media aren't all that important. This is why the company is aggressively pushing its new HAMR technology as a way to increase areal density on traditional platter-based media.

Plans to achieve 18 TB and 20 TB density HDDs in the first semester of 2020 seem to be well within reason, considering the company has recently shipped 16 TB HDDs. These HDD solutions will still make use of older technologies such as CMR (Conventional Magnetic Recording), for the 18 TB drives) and SMR (Shingled Magnetic Recording) for their 20 TB 2020 products. The company will later make use of their proprietary HAMR (Heat Assisted Magnetic Recording) technology in order to upgrade their 20 TB, and 20 TB+ HDDs with higher performance (and density) than can be achieved with the tried and true SMR. When it comes to performance improvements, a latent disadvantage in HDDs compared to solid state solutions, the company will eventually deploy HDDs which make use of two sets of read/write heads instead of a single one.

As computing systems advance and performance is increased, there is also increased need for sufficient storage space to store the results of data processing. Emerging trends and developments in storage technology have seen the promise of new, magnetic-based technologies such as Western Digital's own MAMR technology, which is expected to bring up to 40 TB HDDs in the near future. Until those days are with us, however, improvements on current platter-based technologies are consistent and dependable, with increasing storage space capacities and densities having been delivered by companies in the last few years.

Western Digital has now begun shipping up to 14 TB HDDs, on their UltraStar HS14 family of Enterprise drives, which bring storage areal density up to a staggering 1034 Gb/in². This level of areal storage density was achieved with recourse to SMR (Shingled Magnetic Recording) technology, though, which means performance in rewriting already written-to sectors will be somewhat decreased when compared to other HDDs that don't employ this particular technology. These helium-filled HDDs have a 7200 RPM spindle speed and a 512 MB cache to aid offsetting this performance penalty, though. The UltraStar HS14 HDDs also offer numerous technologies to increase reliability and durability of the drive, which, with its Enterprise classification, promise to deliver 2.5 million hours MTBF (Mean Time Between Failures) accompanied by a five-year warranty. Available interfaces are SATA 6 Gb/s and SAS 12 Gb/s, with the drives being able to sustain transfer rates in the order of 233 MB/s.