Expanding the system's computing capability with an additional accelerator like a GPU is common. However, expanding the system's memory capacity with room for more DIMM is something new. Thanks to ServeTheHome, we see that at the OCP Summit 2024, Lenovo showcased its ThinkSystem SR860 V3 server, leveraging CXL technology and Astera Labs Leo memory controllers to accommodate a staggering 16 TB of DDR5 memory across 128 DIMM slots. Traditional four-socket servers face limitations due to the memory channels supported by Intel Xeon processors. With each CPU supporting up to 16 DDR5 DIMMs, a four-socket configuration maxes out at 64 DIMMs, equating to 8 TB when using 128 GB RDIMMs. Lenovo's new approach expands this ceiling significantly by incorporating an additional 64 DIMM slots through CXL memory expansion.

The ThinkSystem SR860 V3 integrates Astera Labs Leo controllers to enable the CXL-connected DIMMs. These controllers manage up to four DDR5 DIMMs each, resulting in a layered memory design. The chassis base houses four Xeon processors, each linked to 16 directly connected DIMMs, while the upper section—called the "memory forest"—houses the additional CXL-enabled DIMMs. Beyond memory capabilities, the server supports up to four double-width GPUs, making it also a solution for high-performance computing and AI workloads. This design caters to scale-up applications requiring vast memory resources, such as large-scale database management, and allows the resources to stay in memory instead of waiting on storage. CXL-based memory architectures are expected to become more common next year. Future developments may see even larger systems with shared memory pools, enabling dynamic allocation across multiple servers. For more pictures and video walkthrough, check out ServeTheHome's post.

G.SKILL International Enterprise Co., Ltd., the world's leading brand of performance overclock memory and PC components, is excited to announce the launch of Trident Z5 CK and Trident Z5 CK RGB series extreme overclock DDR5 CU-DIMM memory, featuring a built-in clock driver (CKD) and available up to a blistering overclock speed of DDR5-9600. Designed for use with the latest Intel Core Ultra 200 K-series desktop processors and Intel Z890 chipset motherboards, the Trident Z5 CK and Trident Z5 CK RGB series comes with mirrored black finish heatspreaders and pushes the boundaries of memory overclock to new heights. The Trident Z5 CK and Trident Z5 CK RGB series DDR5-9600 memory kits are currently listed on Newegg.com.

DDR5 CU-DIMM - The New Generation of DDR5 Memory
The Trident Z5 CK and Trident Z5 CK RGB series is built on the new CU-DIMM standard, which introduces a built-in clock driver (CKD) chip on the memory module. Designed to strengthen signals between the CPU and memory IC chips, the CKD helps in improving stability in high-speed memory operations.

Team Group Industrial has officially introduced the industry's first industrial-grade DDR5 6400 MHz CU-DIMM/CSO-DIMM memory modules. The innovative product addresses challenges posed by high-frequency operation on signal stability by incorporating a specialized component known as the Client Clock Driver (CKD), which effectively buffers and drives clock signals, ensuring the signals remain stable and complete even under high-frequency conditions. This advancement sets a new benchmark in industrial storage technology, enhancing reliability and performance for demanding applications.

Team Group utilizes DDR5 Clocked Unbuffered Dual Inline Memory Module technology, which enables frequency and voltage adjustments based on system load and operating conditions, facilitating exceptional data transmission speeds and optimized power consumption performance, fully supporting the dynamic load requirements prevalent in industrial applications, and delivering solutions with outstanding performance and reliability.

TrendForce's latest findings reveal that weaker consumer demand has persisted through 3Q24, leaving AI servers as the primary driver of memory demand. This dynamic, combined with HBM production displacing conventional DRAM capacity, has led suppliers to maintain a firm stance on contract price hikes.

Smartphone brands continue to remain cautious despite some server OEMs continuing to show purchasing momentum. Consequently, TrendForce forecasts that Q4 memory prices will see a significant slowdown in growth, with conventional DRAM expected to increase by only 0-5%. However, benefiting from the rising share of HBM, the average price of overall DRAM is projected to rise 8-13%—a marked deceleration compared to the previous quarter.

Intel today announced that its lead products on Intel 18A, Panther Lake (AI PC client processor) and Clearwater Forest (server processor), are out of the fab and have powered-on and booted operating systems. These milestones were achieved less than two quarters after tape-out, with both products on track to start production in 2025. The company also announced that the first external customer is expected to tape out on Intel 18A in the first half of next year.

"We are pioneering multiple systems foundry technologies for the AI era and delivering a full stack of innovation that's essential to the next generation of products for Intel and our foundry customers. We are encouraged by our progress and are working closely with customers to bring Intel 18A to market in 2025." -Kevin O'Buckley, Intel senior vice president and general manager of Foundry Services

TrendForce's latest report on the memory industry reveals that DRAM and NAND Flash revenues are expected to see significant increases of 75% and 77%, respectively, in 2024, driven by increased bit demand, an improved supply-demand structure, and the rise of high-value products like HBM.

Furthermore, industry revenues are projected to continue growing in 2025, with DRAM expected to increase by 51% and NAND Flash by 29%, reaching record highs. This growth is anticipated to revive capital expenditures and boost demand for upstream raw materials, although it will also increase cost pressure for memory buyers.

As AMD prepares to roll out its next-generation Ryzen 9000 series of CPUs based on Zen 5 architecture, we are starting to see some systems being tested by third-party OEMs and system integrators. Today, we have Geekbench 6 scores of the Ryzen 9 9900X CPU, and the 12-core, 24-thread processor that has demonstrated impressive performance gains. Boasting a base clock of 4.4 GHz and a boost clock of up to 5.6 GHz, the CPU features only 120 W TDP, a significant reduction from the previous 170 W of the previous generation. In Geekbench 6 tests, the Ryzen 9 9900X achieved a single-core score of 3,401 and a multicore score of 19,756.

These results place it ahead of Intel's current flagship Core i9-14900KS, which scored 3,189 points in single-core performance. Regarding multicore tasks, the i9-14900K scored 21,890 points, still higher than AMD's upcoming 12-core SKU. The benchmark of AMD's CPU was conducted on an ASUS ROG Crosshair X670E Gene motherboard with 32 GB of DDR5 memory. As anticipation builds for the official release, these early benchmarks suggest that AMD will deliver a compelling product that balances high performance with improved energy efficiency. The top tier models will still carry a 170 W TDP, while some high-end and middle-end SKUs get a TDP reduction like the Ryzen 7 9700X and Ryzen 5 9600X dial down to 65 W, decreased from 105 W in their previous iterations.

TrendForce reports that a recovery in demand for general servers—coupled with an increased production share of HBM by DRAM suppliers—has led suppliers to maintain their stance on hiking prices. As a result, the ASP of DRAM in the third quarter is expected to continue rising, with an anticipated increase of 8-13%. The price of conventional DRAM is expected to rise by 5-10%, showing a slight contraction compared to the increase in the second quarter.

TrendForce notes that buyers were more conservative about restocking in the second, and inventory levels on both the supplier and buyer sides did not show significant changes. Looking ahead to the third quarter, there is still room for inventory replenishment for smartphones and CSPs, and the peak season for production is soon to commence. Consequently, it is expected that smartphones and servers will drive an increase in memory shipments in the third quarter.

SK hynix has returned to Las Vegas to showcase its leading AI memory solutions at HPE Discover 2024, Hewlett Packard Enterprise's (HPE) annual technology conference. Held from June 17-20, HPE Discover 2024 features a packed schedule with more than 150 live demonstrations, as well as technical sessions, exhibitions, and more. This year, attendees can also benefit from three new curated programs on edge computing and networking, hybrid cloud technology, and AI. Under the slogan "Memory, The Power of AI," SK hynix is displaying its latest memory solutions at the event including those supplied to HPE. The company is also taking advantage of the numerous networking opportunities to strengthen its relationship with the host company and its other partners.

The World's Leading Memory Solutions Driving AI
SK hynix's booth at HPE Discover 2024 consists of three product sections and a demonstration zone which showcase the unprecedented capabilities of its AI memory solutions. The first section features the company's groundbreaking memory solutions for AI, including HBM solutions. In particular, the industry-leading HBM3E has emerged as a core product to meet the growing demands of AI systems due to its exceptional processing speed, capacity, and heat dissipation. A key solution from the company's CXL lineup, CXL Memory Module-DDR5 (CMM-DDR5), is also on display in this section. In the AI era where high performance and capacity are vital, CMM-DDR5 has gained attention for its ability to expand system bandwidth by up to 50% and capacity by up to 100% compared to systems only equipped with DDR5 DRAM.

SK hynix presented its leading AI memory solutions at COMPUTEX Taipei 2024 from June 4-7. As one of Asia's premier IT shows, COMPUTEX Taipei 2024 welcomed around 1,500 global participants including tech companies, venture capitalists, and accelerators under the theme "Connecting AI". Making its debut at the event, SK hynix underlined its position as a first mover and leading AI memory provider through its lineup of next-generation products.

"Connecting AI" With the Industry's Finest AI Memory Solutions
Themed "Memory, The Power of AI," SK hynix's booth featured its advanced AI server solutions, groundbreaking technologies for on-device AI PCs, and outstanding consumer SSD products. HBM3E, the fifth generation of HBM1, was among the AI server solutions on display. Offering industry-leading data processing speeds of 1.18 terabytes (TB) per second, vast capacity, and advanced heat dissipation capability, HBM3E is optimized to meet the requirements of AI servers and other applications. Another technology which has become crucial for AI servers is CXL as it can increase system bandwidth and processing capacity. SK hynix highlighted the strength of its CXL portfolio by presenting its CXL Memory Module-DDR5 (CMM-DDR5), which significantly expands system bandwidth and capacity compared to systems only equipped with DDR5. Other AI server solutions on display included the server DRAM products DDR5 RDIMM and MCR DIMM. In particular, SK hynix showcased its tall 128-gigabyte (GB) MCR DIMM for the first time at an exhibition.

Yesterday we reported on DDR6 memory hitting new heights of performance and it looks like LPDDR6 will follow suit, at least based on details in a JEDEC presentation. LPDDR6 will just like LPDDR5 be available as solder down memory, but it will also be available in a new LPCAMM2 module. The bus speed of LPDDR5 on LPCAMM2 modules is expected to peak at 9.2 GT/s based on JEDEC specifications, but LPDDR6 will extend this to 14.4 GT/s or roughly a 50 percent increase. However, today the fastest and only LPCAMM2 modules on the retail market which are using LPDDR5X, comes in at 7.5 GT/s, which suggests that launch speeds of LPDDR6 will end up being quite far from the peak speeds.

There will be some other interesting changes to LPDDR6 CAMM2 modules as there will be a move from 128-bit per module to 192-bit per module and each channel will go from 32-bits to 48-bits. Part of the reason for this is that LPDDR6 is moving to a 24-bit channel width, consisting of two 12-bit sub channels, as mentioned in yesterday's news post. This might seem odd at first, but in reality is fairly simple, LPDDR6 will have native ECC (Error Correction Code) or EDC (Error Detection Code) support, but it's currently not entirely clear how this will be implemented on a system level. JEDEC is also looking at developing a screwless solution for the CAMM2 and LPCAMM2 memory modules, but at the moment there's no clear solution in sight. We might also get to see LPDDR6 via LPCAMM2 modules on the desktop, although the presentation only mentions CAMM2 for the desktop, something we've already seen that MSI is working on.

TrendForce reports that the three largest DRAM suppliers are increasing wafer input for advanced processes. Following a rise in memory contract prices, companies have boosted their capital investments, with capacity expansion focusing on the second half of this year. It is expected that wafer input for 1alpha nm and above processes will account for approximately 40% of total DRAM wafer input by the end of the year.

HBM production will be prioritized due to its profitability and increasing demand. However, limited yields of around 50-60% and a wafer area 60% larger than DRAM products mean a higher proportion of wafer input is required. Based on the TSV capacity of each company, HBM is expected to account for 35% of advanced process wafer input by the end of this year, with the remaining wafer capacity used for LPDDR5(X) and DDR5 products.

EK, a leader in premium liquid cooling solutions, proudly unveils the EK-Quantum Momentum² Quad RAM Module Set D-RGB, an advanced cooling solution meticulously designed for effective thermal management of DDR4 and DDR5 memory modules. This comprehensive set includes a water block capable of covering four RAM sticks with aluminium heatsinks, integrating high-quality cooling efficiency with dynamic RGB lighting to enhance both the performance and visual appeal of any high-end computing setup. The system is specifically engineered to support configurations with four RAM sticks, optimizing cooling for modern high-performance memory arrays.

Enhanced Cooling and Sophisticated Design
The EK-Quantum Momentum² Quad RAM Module Set D-RGB accommodates both single and double-sided DDR4 and DDR5 memory modules, ensuring versatility across different RAM configurations. It features four sleek, black, anodized aluminium heatsinks with an elox finish for enhanced durability and corrosion resistance. This treatment not only improves the heatsinks' appearance but also boosts their thermal conductivity and cooling effectiveness.

G.SKILL International Enterprise Co., Ltd., the world's leading brand of performance overclock memory and PC components, is announcing the release of Ripjaws M5 RGB, a high-performance RGB DDR5 memory series designed for the latest DDR5-enabled Intel platforms. At launch, initial specifications will be offered up to DDR5-6400 CL32-39-39-102 96 GB (2x48GB). Available in matte black or matte white aluminium heatspreaders and features customizable RGB lighting, the Ripjaws M5 RGB is an ideal memory kit for a wide variety of PC builds and themes.

Designed for Stylish Performance
Available in matte white or matte black aluminium heatspreaders and standing at 41 mm tall, the Ripjaws M5 RGB series is designed for a minimalistic look and features customizable RGB lighting to match a wide variety of PC build themes.

The rollout of SiPearl's much-anticipated Rhea processor for European supercomputers has been pushed back by a year to 2025, but the delay comes with a silver lining - a significant upgrade in core count and potential performance. Originally slated to arrive in 2024 with 72 cores, the homegrown high-performance chip will now pack 80 cores when it eventually launches. This decisive move by SiPearl and its partners is a strategic choice to ensure the utmost quality and capabilities for the flagship European processor. The additional 12 months will allow the engineering teams to further refine the chip's architecture, carry out extensive testing, and optimize software stacks to take full advantage of Rhea's computing power. Now called the Rhea1, the chip is a crucial component of the European Processor Initiative's mission to develop domestic high-performance computing technologies and reduce reliance on foreign processors. Supercomputer-scale simulations spanning climate science, drug discovery, energy research and more all require astonishing amounts of raw compute grunt.

By scaling up to 80 cores based on the latest Arm Neoverse V1, Rhea1 aims to go toe-to-toe with the world's most powerful processors optimized for supercomputing workloads. The SiPearl wants to utilize TSCM's N6 manufacturing process. The CPU will have 256-bit DDR5 memory connections, 104 PCIe 5.0 lanes, and four stacks of HBM2E memory. The roadmap shift also provides more time for the expansive European supercomputing ecosystem to prepare robust software stacks tailored for the upgraded Rhea silicon. Ensuring a smooth deployment with existing models and enabling future breakthroughs are top priorities. While the delay is a setback for SiPearl's launch schedule, the substantial upgrade could pay significant dividends for Europe's ambitions to join the elite ranks of worldwide supercomputer power. All eyes will be on Rhea's delivery in 2025, mainly from Europe's governments, which are funding the project.

Patriot Memory, a renowned name in performance memory and storage solutions, has teamed up with MSI, an innovator in gaming hardware, to unleash the Viper Xtreme 5 RGB DDR5 MPOWER Series - an extreme-speed DDR5 memory kit crafted for gamers, overclockers, and power users. This cutting-edge collaboration between the two brands combines blistering data rates up to 8000MT/s with massive 32 GB (16 GB x2) or 48 GB (24 GB x2) capacities. The Viper Xtreme 5 RGB DDR5 MPOWER series unleashes a torrent of memory bandwidth, turbocharging gaming rigs, content creation workstations, and other high-intensity applications demanding extreme responsiveness and throughput.

Optimized for Maximum Overclocking Potential
Intel XMP 3.0 certified for seamless plug-and-play overclocking on the latest Intel platforms, the Xtreme 5 RGB DDR5 MPOWER kit also integrates with MSI's user-friendly EZ Dashboard utility. This enables granular control over advanced timings and voltages to extract every last drop of performance.

The battle is over. Microsoft is finally changing the measuring methodology in its Task Manager from Mega Hertz (MHz) to Mega Transfers per second (MT/s). This comes amid the industry push for more technical correctness in RAM measuring, where the MHz nomenclature does not technically represent the speed at which the memory is actually running. While DRAM manufacturers list both MHz and MT/s, the advertised MHz number is much higher than the effective speed at which the DRAM is running, resulting in confusion and arguments in the industry about choosing the correct labeling of DRAM. A little history lesson teaches us that when single data rate (SDR) RAM was introduced, 100 MHz memory would perform 100 MT/s. However, when double data rate (DDR) memory appeared, it would allow for two memory transfers per clock cycle.

This would introduce some confusion where the MHz speed is often mixed up with MT/s. Hence, Microsoft is trying to repair the damage and list memory speeds in MT/s. Modern DDR5 memory makers are advertising DDR5 kits with "DDR5-4800" or "DDR5-6000," without any suffix like MHz or MT/s. This is because, for example, a DDR5-6000 kit runs at 6,000 MT/s, the effective speed is only 3,000 MHz. The actual clock of the memory is only half of what is advertised. The MT/s terminology would be more accurate and describe memory better. This Task Manager update is in the Windows 11 Insider Preview Build 22635.3570 in the Beta Channel, which will trickle down to stable Windows 11 updates for everyone soon.

TrendForce's latest forecasts reveal contract prices for DRAM in the second quarter are expected to increase by 13-18%, while NAND Flash contract prices have been adjusted to a 15-20% Only eMMC/UFS will be seeing a smaller price increase of about 10%.

Before the 4/03 earthquake, TrendForce had initially predicted that DRAM contract prices would see a seasonal rise of 3-8% and NAND Flash 13-18%, significantly tapering from Q1 as seen from spot price indicators which showed weakening price momentum and reduced transaction volumes. This was primarily due to subdued demand outside of AI applications, particularly with no signs of recovery in demand for notebooks and smartphones. Inventory levels were gradually increasing, especially among PC OEMs. Additionally, with DRAM and NAND Flash prices having risen for 2-3 consecutive quarters, the willingness of buyers to accept further substantial price increases had diminished.

Thanks to recent developments in the AI field, and following in the wake of the world's recovery from COVID-19, the transition of factories to partial or full automation proceeds with unstoppable momentum. And the best place to learn about the latest technologies that aim to make this transition as painless as possible is at Automate 2024. This is North America's largest robotics and automation event, and it will be held in Chicago, Illinois from May 6 to 9.

Automate attracts professionals from around the world, and Apacer is no exception. Apacer team will be on hand to discuss the latest technological developments created by our experienced R&D team. Many of these developments were specifically created to reduce the pain points commonly experienced by fully automated facilities. Take CoreSnapshot, for example. This backup and recovery technology can restore a crashed system to full operation in just a few seconds, reducing downtime and associated maintenance costs. Apacer recently updated CoreSnapshot, creating CoreRescue ASR and CoreRescue USR. The name of CoreRescue ASR refers to Auto Self Recovery. This technology will harness AI to learn the system booting process and analyze how long a boot should take. If this average boot time is significantly longer than usual, the system will trigger the self-recovery process and revert to an earlier, uncorrupted version of the drive's content. CoreRescue USR offers similar functionality, except the self-recovery process is triggered by connecting a small USB stick drive.

SK hynix showcased its next-generation technologies and strengthened key partnerships at the TSMC 2024 Technology Symposium held in Santa Clara, California on April 24. At the event, the company displayed its industry-leading HBM AI memory solutions and highlighted its collaboration with TSMC involving the host's CoWoS advanced packaging technology.

TSMC, a global semiconductor foundry, invites its major partners to this annual conference in the first half of each year so they can share their new products and technologies. Attending the event under the slogan "Memory, the Power of AI," SK hynix received significant attention for presenting the industry's most powerful AI memory solution, HBM3E. The product has recently demonstrated industry-leading performance, achieving input/output (I/O) transfer speed of up to 10 gigabits per second (Gbps) in an AI system during a performance validation evaluation.

Today, Lenovo launched its latest mobile workstation offerings meticulously crafted to deliver the exceptional power and performance essential for handling complex workloads. Lenovo's ThinkPad P1 Gen 7, P16v i Gen 2, P16s i Gen 3, and P14s i Gen 5, with their cutting-edge AI technologies, are set to transform the way professionals engage with AI workflows. By collaborating with industry partners, Intel, NVIDIA, and Micron, Lenovo has introduced powerful and performance-packed AI PCs that meet the demands of modern-day AI-intensive tasks. The inclusion of the Intel Core Ultra processors with their integrated neural processing unit (NPU) and NVIDIA RTX Ada Generation GPUs signifies a major advancement in AI technology, boosting overall performance and productivity capabilities.

The latest ThinkPad P series mobile workstations powered by Intel Core Ultra processors and NVIDIA RTX Ada Generation GPUs deliver flexible, high-performance, and energy-efficient AI-ready PCs. The integrated NPU is dedicated to handling light, continuous AI tasks, while the NVIDIA GPU runs more demanding day-to-day AI processing. This combination enables smooth and reliable functioning of AI technologies, serving professionals engaged in diverse tasks ranging from 3D modeling and scene development to AI inferencing and training.

Following in the wake of an earthquake that struck on April 3rd, TrendForce undertook an in-depth analysis of its effects on the DRAM industry, uncovering a sector that has shown remarkable resilience and faced minimal interruptions. Despite some damage and the necessity for inspections or disposal of wafers among suppliers, the facilities' strong earthquake preparedness of the facilities has kept the overall impact to a minimum.

Leading DRAM producers, including Micron, Nanya, PSMC, and Winbond had all returned to full operational status by April 8th. In particular, Micron's progression to cutting-edge processes—specifically the 1alpha and 1beta nm technologies—is anticipated to significantly alter the landscape of DRAM bit production. In contrast, other Taiwanese DRAM manufacturers are still working with 38 and 25 nm processes, contributing less to total output. TrendForce estimates that the earthquake's effect on DRAM production for the second quarter will be limited to a manageable 1%.

TrendForce's latest report reveals that despite DRAM suppliers' efforts to trim inventories, they have yet to reach healthy ranges. As they continue to improve their lose situations by boosting capacity utilization rates, the overall demand outlook for this year remains tepid. Additionally, significant price increases by suppliers since 4Q23 are expected to further diminish the momentum for inventory restocking. As a result, DRAM contract prices for the second quarter are projected to see a modest increase of 3-8%.

The shift toward DDR5-compatible CPUs is set to drive an increase in PC DRAM demand in the second quarter. As manufacturers move toward more advanced, cost-efficient production processes for DDR5, their profitability is expected to rise significantly. This anticipation of higher DRAM prices in 1H24 has led to suppliers to aim for price increases in Q2, targeting a 3-8% hike in PC DRAM contract prices. Notably, even though DDR5 prices have already seen a notable rise in Q1—exceeding the average increase for other products—the expected emergence of AI PC demand may lead to a slight moderation in DDR5 price increases in Q2.

During its 55th annual shareholders' meeting, Samsung Electronics announced its entry into the AI processor market with the upcoming launch of its Mach-1 AI accelerator chips in early 2025. The South Korean tech giant revealed its plans to compete with established players like NVIDIA in the rapidly growing AI hardware sector. The Mach-1 generation of chips is an application-specific integrated circuit (ASIC) design equipped with LPDDR memory that is envisioned to excel in edge computing applications. While Samsung does not aim to directly rival NVIDIA's ultra-high-end AI solutions like the H100, B100, or B200, the company's strategy focuses on carving out a niche in the market by offering unique features and performance enhancements at the edge, where low power and efficient computing is what matters the most.

According to SeDaily, the Mach-1 chips boast a groundbreaking feature that significantly reduces memory bandwidth requirements for inference to approximately 0.125x compared to existing designs, which is an 87.5% reduction. This innovation could give Samsung a competitive edge in terms of efficiency and cost-effectiveness. As the demand for AI-powered devices and services continues to soar, Samsung's foray into the AI chip market is expected to intensify competition and drive innovation in the industry. While NVIDIA currently holds a dominant position, Samsung's cutting-edge technology and access to advanced semiconductor manufacturing nodes could make it a formidable contender. The Mach-1 has been field-verified on an FPGA, while the final design is currently going through a physical design for SoC, which includes placement, routing, and other layout optimizations.

TrendForce reports that significant capital investments have occurred in the memory sector due to the high ASP and profitability of HBM. Senior Vice President Avril Wu notes that by the end of 2024, the DRAM industry is expected to allocate approximately 250K/m (14%) of total capacity to producing HBM TSV, with an estimated annual supply bit growth of around 260%. Additionally, HBM's revenue share within the DRAM industry—around 8.4% in 2023—is projected to increase to 20.1% by the end of 2024.

HBM supply tightens with order volumes rising continuously into 2024
Wu explains that in terms of production differences between HBM and DDR5, the die size of HBM is generally 35-45% larger than DDR5 of the same process and capacity (for example, 24Gb compared to 24Gb). The yield rate (including TSV packaging) for HBM is approximately 20-30% lower than that of DDR5, and the production cycle (including TSV) is 1.5 to 2 months longer than DDR5.