SK hynix validates LPDDR6 memory, 33% faster than LPDDR5X but AI devices get priority

SK hynix introduces turbocharged LPDDR6, 33% faster and 20% more power efficient than LPDDR5X -- 16Gb chips deliver 10.7 Gbps, uses 10nm node

LPDDR6 will likely be a big boon for the datacenter market, once this memory architecture debuts on SOCAMM modules. SK Hynix has announced the successful development of its first LPDDR6 DRAM, touting 33% greater speed and 20% better power efficiency compared to previous generation LPDDR5X memory. The memory manufacturer also announced that it has developed its LPDDR6 memory on its leading-edge 10nm-class (1c) process node that it announced in 2024. The new modules have a base operating speed of "over" 10.7 Gbps -- which exceeds the fastest outgoing LPDDR5X memory modules on the market today -- and a memory capacity of 16Gb per chip. To help with power consumption, SK Hynix implemented a new sub-channel structure and DVFS, or Dynamic Voltage and Frequency Scaling, which it says reduces power consumption by more than 20% compared to previous generation LPDDR memory products. The sub-channel structure accomplishes power reductions by powering only the data paths that are in use, while DVFS reduces clock speed and voltage when memory demands are light. SK Hynix has announced its LPDDR6 eight months after JEDEC finalized and published the LPDDR6 standard last July. However, SK Hynix is not the first manufacturer to make LPDDR6; Samsung already announced its first LPDDR6 product and showcased it at CES 2026, with speeds up to 10.7Gbps. SK Hynix has confirmed that its 1c LPDDR6 will be used in smartphones and tablets, but we can also expect LPDDR6 to be a huge boon in the datacenter market. LPDDRX has been very popular in AI servers that take advantage of SOCAMM/SOCAMM2 memory modules that only support LPDDR memory. For instance, Nvidia's GB300 Grace Blackwell Ultra Superchip uses SOCAMM, and Nvidia's latest Vera Rubin Superchip uses SOCAMM2 memory modules. Late last year, SK Hynix stated that it expects post-Vera Rubin Nvidia AI chip designs to take advantage of LPDDR6. This is just the beginning for LPDDR6; speeds well beyond 10.7Gbps are expected to become the norm as memory makers get to grips with optimizing and improving on their LPDDR6 designs. The JEDEC group believes that LPDDR6 will have data rates of up to 14,400 MT/s, which is well beyond what the fastest DDR5 overclocking record holds. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

DRAM maker SK hynix has validated its next-gen LPDDR6 which would be great news for handheld PCs, except that the targeted market is AI devices

Lots more performance and way better power efficiency, but destined to be rarer than pixie dust. There's one thing that handheld gaming PCs, some laptops, and most phones all have in common, and it's the use of LPDDR5X for the system memory. While fast in speed and low in power usage, it doesn't offer a lot of bandwidth, so the fact that SK hynix is set to start mass production of its next-gen LPDDR6 this year is good news. The bad news is that the DRAM maker seems hell-bent on selling this to the AI crowd. SK hynix's announcement that it has successfully validated its LPDDR6 design on a cutting-edge process node is genuinely good to hear, simply because the performance of most handheld gaming PCs is held back by the fact that the memory they currently use is a bit rubbish at offering lots of VRAM bandwidth. Let me explain. DDR5 memory, as used in millions of desktop PCs around the world, is pretty good stuff, thanks to high clock speeds and twin, independent 32-bit data channels. LPDDR5 and its higher clocked cousin, LPDDR5X, utilise a bunch of tricks to reduce power consumption, but they also have to use narrow data channels, either two 16-bit or one 32-bit. So while the likes of the Asus ROG Xbox Ally X have super-fast LPDDR5x-8000, the bandwidth on offer is equivalent to DDR5-4000. This is one of the things that LPDDR6 partly resolves, as it uses twin 24-bit data channels, and the technology supports higher clock speeds than LPDDR5X. With all the new updates combined, the new low-power DRAM is at least 33% faster than the best LPDDR5X, even at its base data rate of 10.7 Gbps. SK hynix says its LPDDR6 chips use 20% less power compared to the previous generation of DRAM modules, and together with the better performance, it's obvious that they're ideal for handheld gaming PCs, as well as certain laptops. Although there are no processors or APUs on the market that support LPDDR6 yet, there probably will be by the time the new memory is fully available to purchase in bulk. But that might be somewhat of a sticking point, because there's a certain phrase that festoons SK hynix's press release and it's AI, of course. I'm not suggesting that the South Korean firm is targeting the hulking data center market, although that will probably use the new memory at some point; instead, it's talking about "mobile products such as smartphones and tablets equipped with on-device AI." Such manufacturers will certainly be banging at SK's door for this memory, especially Apple, and given that all DRAM manufacturers are currently focusing production for the AI data center market, that means there will be scant offerings left for everyone else. So no matter if AMD or Intel's next series of LPDDR6-capable gaming APUs are really good, there's unlikely to be enough memory to go around to meet the demand of every market. Just as it is now, sadly.

SK Hynix develops 1c-Node 16Gb LPDDR6 DRAM for on-device AI

SK hynix has announced the successful development of its 16-gigabit (Gb) LPDDR6 DRAM, utilizing its sixth-generation 10nm-class (1c) process technology. Following the product's initial unveiling at CES last January, the company has now completed the world's first validation of 1c LPDDR6 development. The company aims to finalize preparations for mass production within the first half of the year, with product supply scheduled to begin in the second half. This move is intended to expand SK hynix's lineup of conventional DRAM optimized for artificial intelligence (AI) applications. The 1c LPDDR6 is primarily designed for mobile products, such as smartphones and tablets, that utilize on-device AI. To support the demanding workloads of on-device AI implementations, SK hynix has focused on improving both data processing speeds and power efficiency relative to existing LPDDR5X models. Key performance metrics include: The enhanced power efficiency is driven by two primary structural and technological implementations: Through these architectural advancements, SK hynix anticipates that end-users will experience tangible benefits, including extended battery life and optimized multitasking performance on mobile devices. The company stated its intention to prepare for the needs of global mobile customers in alignment with market demand. SK hynix emphasized its ongoing commitment to delivering timely AI memory solutions in collaboration with clients, aiming to provide distinct value to on-device AI users.

SK hynix develops 1c LPDDR6 low-power memory - The Korea Times

SK hynix said Tuesday it has completed developing low-power double data rate 6 (LPDDR6) memory using the sixth-generation 10-nanometer-class dynamic random access memory technology, called 1c process. LPDDR is commonly used in smartphones and other portable devices and is designed primarily to minimize power consumption. LPDDR6 represents the eighth generation of the LPDDR standard. SK hynix said it will complete preparations for mass production within the first half of the year and begin shipping it to clients in the second half, so that it can "establish a general memory product lineup optimized for artificial intelligence (AI) applications." The company said its 1c LPDDR6 memory will be used for smartphones and tablet PCs featuring on-device AI functions, leveraging its improved data processing speed and power efficiency. Compared with the previous LPDDR5X, the product improves data processing speed by 33 percent by expanding bandwidth to increase the amount of data transferred per unit time. Its operating speed is at least 10.7 gigabits per second, exceeding the maximum speed of the previous generation. Power consumption was reduced by more than 20 percent compared with the previous generation by adopting a sub-channel architecture as well as dynamic voltage and frequency scaling (DVFS) technology. DVFS is a power management technology that optimizes power consumption and performance by adjusting voltage and frequency depending on the chip's operating conditions. Through this, the chip is designed to operate only the data paths required for a certain task and adjust frequency and voltage depending on the mobile environment. In high-performance situations such as gaming, DVFS is increased to enable maximum bandwidth operation, while during normal use the frequency and voltage are lowered to reduce power consumption. SK hynix said LPDDR6 will enable longer battery life and an optimized multitasking experience, adding that it will prepare manufacturing depending on global mobile clients.