Samsung Electronics the world leader in advanced memory technology, today showcased its latest advancements with processing-in-memory (PIM) technology at Hot Chips 33—a leading semiconductor conference where the most notable microprocessor and IC innovations are unveiled each year. Samsung's revelations include the first successful integration of its PIM-enabled High Bandwidth Memory (HBM-PIM) into a commercialized accelerator system, and broadened PIM applications to embrace DRAM modules and mobile memory, in accelerating the move toward the convergence of memory and logic.

In February, Samsung introduced the industry's first HBM-PIM (Aquabolt-XL), which incorporates the AI processing function into Samsung's HBM2 Aquabolt, to enhance high-speed data processing in supercomputers and AI applications. The HBM-PIM has since been tested in the Xilinx Virtex Ultrascale+ (Alveo) AI accelerator, where it delivered an almost 2.5X system performance gain as well as more than a 60% cut in energy consumption.

After DRAM prices made a rebound into an upward trajectory in 1Q21, buyers expanded their DRAM procurement activities in 2Q21 as they anticipated a further price hike and insufficient supply going forward, according to TrendForce's latest investigations. Not only was demand robust from clients in the notebook segment, which benefitted from ongoing WFH and distance learning applications, but CSPs also sought to gradually replenish their DRAM inventories. Furthermore, demand for products that are relatively niche, including graphics DRAM and consumer DRAM, remained strong. Hence, DRAM suppliers experienced better-than-expected QoQ increases in their DRAM shipment for 2Q21. At the same time, DRAM quotes grew by a greater magnitude compared to the first quarter as well. With both shipment and quotes undergoing growths in tandem, DRAM suppliers registered remarkable growths in their revenues in 2Q21. Total DRAM revenue for 2Q21 reached US$24.1 billion, a 26% QoQ increase.

At this year's Hot chips 33 conference, Samsung has presented its works on the upcoming DDR5 memory standard. The company has managed to achieve a lot of new developments, as the newer standard pairs with new technologies to deliver higher speeds and better capacity. The Korean company designed its DDR5 modules as 8-high (8H) stacked TSV (through silicon via) dies. In the previous DDR4 implementations, Samsung used 4-high (4H) stacked TSV dies, which are actually thicker than the latest 8-high implementations. To achieve the new thin design, Samsung has used thin wafer handling techniques, which resulted in a 40% reduction in gab between stacked dies. The new 8H DDR5 modules are only 1.0 mm thick, compared to the 1.2 mm of the older 4H modules.

When it comes to performance, Samsung expects the new DDR5 modules to deliver big. Running at 7.2 Gbps speeds, the Samsung-made RDIMM/LRDIMM modules can reach up to 512 GB in capacity. This is, of course, limited to the server/enterprise market. Regular consumers/PC users can expect to have UDIMMs with up to 64 GB of capacity. The aforementioned 7.2 Gbps speed is achieved at the specified 1.1 Volts of power, meaning that Samsung's implementation is very efficient. According to some estimations made by the company, the DDR5 crossover for the mainstream market is not expected before 2023/2024, meaning that there is still a lot of time for memory makers to refine their DDR5 products.

In his Q2-FY2022 Results call, NVIDIA CEO Jensen Huang commented that he expects the ongoing chip supply situation to remain bad for the most part of 2022. "Meanwhile, we have and are securing pretty significant long-term supply commitments as we expand into all these different markets initiatives that we've set ourselves up for. And I so I think—I would expect that we will see a supply contained environment for the vast majority of next year is my guess at the moment," he said.

His comments are telling, as NVIDIA relies heavily on cutting edge silicon fabrication nodes for its logic products, such as GPUs, HPC processors, and vehicle SoCs. 2022 will see NVIDIA introduce its "Lovelace" graphics architecture, powering the GeForce RTX 40-series GPUs; as well as a variant powering next-generation HPC processors. The company is looking to design its chips for TSMC's 5 nm silicon fabrication process, unless Samsung can fix its 5 nm-class foundry woes in time, and win back the company.

Samsung Electronics Co., Ltd., a world leader in advanced semiconductor technology, today announced its new wearable processor, the Exynos W920. The new processor integrates an LTE modem and is the first in the industry to be built with an advanced 5-nanometer (nm) extreme ultra-violet (EUV) process node, offering powerful yet efficient performance demanded by next-generation wearable devices.

"Wearables like smartwatches are no longer just a cool gadget to have. They're now a growing part of our lifestyles to keep you fit, safe and alert," said Harry Cho, vice president of System LSI marketing at Samsung Electronics. "With the Exynos W920, future wearables will be able to run applications with visually appealing user interfaces and more responsive user experiences while keeping you connected on the go with fast LTE."

Samsung Electronics today reported financial results for the second quarter ended June 30, 2021. Total consolidated revenue was KRW 63.67 trillion, a 20% increase from the previous year and a record for the second quarter. Operating profit increased 34% from the previous quarter to KRW 12.57 trillion as market conditions improved in the memory market, operations normalized at the Austin foundry fab, and as effective global supply chain management (SCM) helped maintain solid profitability for the finished product businesses.

The Semiconductor business saw a significant improvement in earnings as memory shipments exceeded previous guidance and price increases were higher than expected, while the Company strengthened its cost competitiveness. For the Display Panel Business, a one-off gain and an increase in overall prices boosted profits.

Samsung Electronics, today announced the global launch of its next-generation curved gaming monitor, the Odyssey Neo G9, enhanced with Quantum Mini LED technology. The Odyssey Neo G9 joins the Odyssey lineup, taking gaming to the next level with Quantum Matrix technology, supported by Quantum Mini LED display and Quantum HDR 2000 for an immersive gaming experience.

With the previous launch of the Odyssey G9 in 2020, Samsung pushed the boundaries of premium gaming monitors with smooth and brilliant picture quality. Now with the industry's first Quantum Mini LED curved display delivering the most refined detail regardless of the game played, the Odyssey Neo G9 is catapulting the gaming monitor category forward. Featuring Quantum Matrix Technology combined with super-fast response and refresh rates, Odyssey Neo G9 delivers an unrivaled dynamic picture and premium performance for all gamers.

Intel, in a move comparable to its competitors' Performance Rating system from the 1990s, has invented a new naming scheme for its in-house foundry nodes to claim technological parity with contemporaries such as TSMC and Samsung, that are well into the sub-10 nm class. Back in the i586 era, when Intel's competitors such as AMD and Cyrix, couldn't keep up with its clock-speeds yet found their chips to be somewhat competitive, they invented the PR (processor rating) system, with a logical number attempting to denote parity with an Intel processor's clock-speed. For example, a PR400 processor rating meant that the chip rivaled a Pentium II 400 MHz (which it mostly didn't). The last that the PR system made sense was with the final generation of single-core performance chips, Pentium 4 and Athlon XP, beyond which, the introduction of multi-core obfuscated the PR system. A Phenom X4 9600 processor didn't mean performance on par with a rival Intel chip running at an impossible 9.60 GHz.

Intel's new foundry naming system sees its 10 nm Enhanced SuperFin node re-badge as "Intel 7." The company currently builds 11th Gen Core "Tiger Lake" processors on the 10 nm SuperFin node, and is expected to build its upcoming 12th Gen Core "Alder Lake" chips on its refinement, the 10 nm Enhanced SuperFin, which will now be referred to as "Intel 7." The company is careful to avoid using the nanometer unit next to the number, instead signaling the consumer that the node somehow offers transistor density and power characteristics comparable to a 7 nm node. Intel 7 offers a 10-15 percent performance/Watt gain over 10 nm SuperFin, and is already in volume production, with a debut within 2021 with "Alder Lake."

NVIDIA's upcoming "Ada Lovelace" architecture, both for compute and graphics, is reportedly being designed for the 5 nanometer silicon fabrication node by TSMC. This marks NVIDIA's return to the Taiwanese foundry after its brief excursion to Samsung, with the 8 nm "Ampere" graphics architecture. "Ampere" compute dies continue to be built on TSMC 7 nm nodes. NVIDIA is looking to double the compute performance on its next-generation GPUs, with throughput approaching 70 TFLOP/s, from a numeric near-doubling in CUDA cores, generation-over-generation. These will also be run at clock speeds above 2 GHz. One can expect "Ada Lovelace" only by 2022, as TSMC N5 matures.

Samsung intends to take its partnership with AMD for graphics further, by designing its next-generation Exynos 2200 "Pamir" SoC with a faster GPU based on the AMD RDNA2 graphics architecture. Bound for the second half of 2021, Exynos 2200 will be built on Samsung's swanky new 4 nanometer 4LPP (4 nm Low Power Plus) silicon fabrication node, and integrate an RDNA2-based GPU codenamed "Voyager." Samsung hopes to compete with Qualcomm's Snapdragon 895 SoC and its Adreno 730 GPU. Interestingly, the new Snapdragon is also expected to be built on the same Samsung 4 nm node. It will be interesting to see what device the Exynos 2200 debuts with, given that both the Galaxy S22 and Galaxy Note 21 won't arrive before 2022.

The recent wave of COVID-19 outbreaks in India has weakened sales of retail storage products such as memory cards and USB drives, according to TrendForce's latest investigations. However, demand remains fairly strong in the main application segments due to the arrival of the traditional peak season and the growth in the procurement related to data centers. Hence, the sufficiency ratio of the entire market has declined further. NAND Flash suppliers have kept their inventories at a healthy level thanks to clients' stock-up activities during the past several quarters. Moreover, the ongoing shortage of NAND Flash controller ICs continues to affect the production of finished storage products. Taking account of these demand-side and supply-side factors, TrendForce forecasts that contract prices of NAND Flash products will rise marginally for 3Q21, with QoQ increases in the range of 5-10%.

As third quarters have typically been peak seasons for the production of various end-products, the sufficiency ratio of DRAM is expected to undergo a further decrease in 3Q21, according to TrendForce's latest investigations. However, DRAM buyers are now carrying a relatively high DRAM inventory due to their amplified purchases of electronic components in 1H21. The QoQ increase in DRAM contract prices are hence expected to slightly narrow from 18-23% in 2Q21 to 3-8% in 3Q21. Looking ahead to 4Q21, TrendForce believes that DRAM supply will continue to rise, thereby leading to either a further narrowing of price hikes or pressure constraining the potential price hike of DRAM products.

From the perspective of demand, the stay-at-home economy has resulted in persistently high demand for notebook computers. Although discrepancies still exist among notebook brands' inventory levels of various components, these brands are still making an aggressive attempt at maximizing their production of notebooks. However, as most of these brands are still carrying about 8-10 weeks' worth of PC DRAM inventory (which is relatively high), PC DRAM purchasing strategies from the buyers' side will therefore remain relatively conservative. From the perspective of supply, due to the rising demand for server DRAM, the production capacity allocated to PC DRAM is still in a severe supply crunch. Hence, DRAM suppliers are firm in their attitudes to raise PC DRAM quotes, and TrendForce expects the price negotiations between PC DRAM buyers and suppliers in 3Q21 to become both lengthier and more difficult as a result, with contract prices likely finalized at the end of July. Even so, what is now certain is that both sides have reached some level of understanding regarding the ongoing price hike of PC DRAM products. TrendForce forecasts a 3-8% increase in PC DRAM contract prices for 3Q21.

While the stay-at-home economy generated high demand for notebook computers from distance learning and WFH applications last year, global notebook shipment for 2020 underwent a nearly 26% YoY increase, which represented a significant departure from the cyclical 3% YoY increase/decrease that had historically taken place each year, according to TrendForce's latest investigations. The uptrend in notebook demand is expected to persist in 2021, during which notebook shipment will likely reach 236 million units, a 15% YoY increase. In particular, thanks to the surging demand for education notebooks, Chromebooks will become the primary growth driver in the notebook market. Regarding the shipment performance of various brands, Samsung and Apple will register the highest growths, with the former having Chromebooks account for nearly 50% of its total notebook shipment this year and the latter continuing to release MacBooks equipped with the M1 chip.

Samsung has recently provided a few details of their PM1743 PCIe Gen 5 E3.S 1T EDSFF SSD set to release in Q2 2022. The PM1743 is an upcoming enterprise SSD from Samsung with PCIe 5.0 x4 connectivity which can enable a theoretical maximum speed of 15.7 GB/s. The SSD features V6 TLC NAND flash and comes with 1 Drive Writes Per Day (DWPD) of write endurance. The drive features an enterprise E3.S 1T single-width form factor (111.5 mm × 31.5 mm) popular in server deployments and will likely come with a TDP of 20-25 W. Samsung has provided a basic mechanical drawing of the SSD but we expect to find out more information closer to release.

Semiconductor manufacturing is no easy task. Every company in that business knows that, and the hardships of silicon manufacturing have been felt by even the greatest players like Samsung and Intel. Today, according to the latest report from Business Korea, Samsung is again in trouble with its 5 nm node. It has been reported previously that Samsung is struggling with yields of its 5 nm node, however, we didn't know just how much until now. According to the sources over at Business Korea, Samsung's 5 nm semiconductor node is experiencing less than 50% yields. That means, for example, that out of 100 chips manufactured on a single silicon wafer, only half are functional. And that is not good at all.

Usually, for a node to go into high-volume manufacturing (HVM), the yielding rate needs to be around 95%. In case it is not at that level, manufacturing of that node is not very efficient and not very profitable. The V1 Line in Hwaseong, where this Samsung 5 nm is made, uses EUV tools to manufacture the new node. While the yields are currently below 50%, it is expected to improve as Samsung engineers tweak and tune the node and the tools that are running the facility. We can expect to hear more about the yields of this node in the coming months.

Samsung's ambitious 3 nm silicon fabrication node that leverages the Gate All Around FET transistors, has reportedly been delayed to 2024. The company brands this specific node as 3GAE. 2024 is the earliest date when Samsung will be able to mass-produce chips on 3GAE, which means the company, along with Intel, will begin to fall significantly behind TSMC on foundry technology. The Taiwanese semiconductor fabrication giant will target 2 nm-class nodes around 2024, which leverages EUV multi-patterning, extensive use of cobalt in contacts and interconnects, germanium doped channels, and other in-house innovations. With Intel's foundry technology development slowing to a crawl in the sub-10 nm domain, Samsung is the only viable alternative to TSMC for cutting-edge logic chip manufacturing.

We recently reported that Samsung would be announcing their next-generation flagship Exynos processor with AMD RDNA2 graphics next month. We heard that the RDNA2 GPU was expected to be ~30% faster than the Mali-G78 GPU present in Galaxy S21 Ultra however according to a new 3DMark Wild Life benchmark it would appear the new processor scores 56% higher. This result would give the upcoming Exynos processor the fastest graphics available in any Android phone even matching/beating out the Apple A14 Bionic found in the iPhone 12. This early performance benchmark paints a very positive picture for the upcoming processor however we still don't know how the score will be affected under sustained load or if this will performance will even be replicated in the final product.

Marvell today introduced its new OCTEON 10 DPU designed to accelerate and process a broad spectrum of security, networking, and storage workloads required by demanding 5G, cloud, carrier and enterprise datacenter applications. With the increasing shift of workloads to the cloud, complex security requirements and the growing number of edge devices the demand for data centric compute has accelerated. By combining compute with best-in-class hardware accelerators, Marvell's OCTEON 10 DPU offers a significant TCO advantage and features numerous industry firsts. Delivering three times the performance and 50 percent lower power compared to previous generations of OCTEON, the newly announced solution is the first to be designed on a 5 nm process to incorporate Arm Neoverse N2 cores, as well as the first inline artificial intelligence/machine learning (AI/ML) hardware acceleration, the first integrated 1 terabit switch and the first to incorporate vector packet processing (VPP) hardware accelerators.

"To meet and exceed the growing data processing requirements for network, storage, and security workloads, Marvell focused on significant DPU innovations across compute, hardware accelerators, and high speed I/O," said John Sakamoto, vice president of Marvell's Infrastructure Processors Business Unit. "The OCTEON 10 brings compute leadership, supports networking and security workloads exceeding 400G, and incorporates leading edge I/O including DDR5 and PCIe 5.0."

NVIDIA's $40 billion takeover of Arm Holdings plc from SoftBank, got a shot in the arm, as three major licensees of the IP came out in support of the bid. These include Broadcom, MediaTek, and Marvell Technology Group. This development is key for NVIDIA to fight the perception built up by a rival faction, that the democratized nature of the Arm IP would get lost if a chipmaker like NVIDIA owns it. This rival faction is primarily led by Qualcomm.

It's interesting to note the individual backers of the NVIDIA takeover. There is nothing but love between Broadcom and Qualcomm, especially after the former's failed bid to acquire the latter. MediaTek is a major smartphone and IoT SoC maker, dominating the low-cost and mainstream smartphone segments. Marvell is big in datacenter and storage IP. Each of the three are results of huge IP consolidation over the past decade.

The partnership between Samsung and AMD began in 2019 when the two companies announced that they would work together to integrate Radeon graphics IP in Samsung Exynos processors. We can see the results of this partnership with Dr. Lisa Su confirming at Computex that RDNA2 graphics will be integrated into the next flagship Samsung Exynos SoC. The RDNA2 GPU found in the upcoming mobile chip will include support for raytracing and variable-rate shading with a strong possibility that it will power the next Galaxy S series flagship. Samsung was initially expected to announce this new chipset in June however the event was postponed until July where the complete details and performance numbers will be unveiled.

Update Jun 22nd: The upcoming GPU is expected to be 30% faster than the current Mali-G78 GPU present in Galaxy S21 Ultra which should give it a comfortable lead of ~10% against the next generation Mali GPU. The GPU does appear to suffer from quite severe thermal throttling with a 20% performance drop after the second run and 30% on the third run. Samsung seems pleased with the collaboration and has engaged in talks with AMD to extend the contract for future GPU architectures.

Intel is in talks with the German government to build a European chip factory hoping to counter the global chip shortages and help achieve the EU local chip manufacturing target. Germany is interested in attracting semiconductor companies to increase domestic chip production to improve security for their automotive industry which increasingly relies on foreign chips. Intel wants to open up manufacturing capacity at their foundries to external companies allowing them to compete with TSMC and Samsung in the high-end market. Intel is seeking large subsidies from the German government to the tune of several billion to help make the new factory a reality.

Samsung Electronics, today announced the expanded 2021 Odyssey monitor lineup will be available across the global markets from June 21, 2021, providing gamers of all skill sets with superb picture quality and futuristic design. Following the curved gaming monitor launch in 2020, Samsung now offers a variety of Odyssey monitors in flat-screen design, ranging from 24 to 28 inches. The new lineup delivers hyper-real picture quality, a higher response level, tailored ergonomics and intuitive usability. Together with these latest features, gaming enthusiasts can enjoy real-world colors, pinpoint accuracy and sharp response speeds for their PC and console gaming entertainment devices.

As gaming industry continues to thrive worldwide, Samsung Odyssey has quickly become the number one choice among gamers seeking incredible picture quality and high performance, all in one package. The expanded lineup now ensures gamers can choose a monitor that can accommodate their exact preferences and play needs.

Samsung Electronics Co., Ltd., the world leader in advanced memory technology, today announced that it has begun mass producing its latest smartphone memory solution, the LPDDR5 UFS-based multichip package (uMCP). Samsung's uMCP integrates the fastest LPDDR5 DRAM with the latest UFS 3.1 NAND flash, delivering flagship-level performance to a much broader range of smartphone users.

"Samsung's new LPDDR5 uMCP is built upon our rich legacy of memory advancements and packaging know-how, enabling consumers to enjoy uninterrupted streaming, gaming and mixed reality experiences even in lower-tier devices," said Young-soo Sohn, vice president of the Memory Product Planning Team at Samsung Electronics. "As 5G-compatible devices become more mainstream, we anticipate that our latest multichip package innovation will accelerate the market transition to 5G and beyond, and help to bring the metaverse into our everyday lives a lot faster."

NVIDIA is reported to be cutting down on production of its highly popular RTX 2060 graphics card, in a bid to increase production of the RTX 30-series graphics cards that still elude most consumers looking to get one on their gaming rig. The decision may be motivated by increased margins on RTX 30-series products, as well as by the continuing component shortage in the industry, with even GDDR6 becoming a limiting factor to production capability.

While one might consider this a strange move at face value (Turing is manufactured on TSMC's 12 nm node, whilst Ampere is manufactured on Samsung's 8 nm), the fact of the matter is that there are a multitude of components required for GPUs besides the graphics processing silicon proper; and NVIDIA essentially sells ready-to-produce kits to AICs (Add-in-Card Partners) which already include all the required components, circuitry, and GPU slice to put together. And since supply on most components and even simple logic is currently strained, every component in an RTX 2060-allocated kit could be eating into final production capacity for the RTX 30-series graphics cards - hence the decision to curb the attempt to satiate pent-up demand with a last-generation graphics card and instead focusing on current-gen hardware.

Samsung is preparing to deploy their latest innovations in NAND density with the next-generation V-NAND (7th gen). Samsung says it is preparing products that leverage both V-NAND's higher density (at 176 layers per chip versus up to 136 layers on 6th gen) with the throughput of both PCIe 4.0 and PCIe 5.0. This would of course mean higher density drives available, as well as a reduction in the overall $/GB equation. Due to Samsung's vertical integration (meaning that they are one of the few companies that can design and produce all SSD components in-house), the company is also developing next-gen NAND controllers that can leverage throughputs of 2,000 MT/s transfer rates and thus "optimized for multitasking huge workloads".

Samsung expects to be able to scale V-NAND well past the 1,000 layer mark - a far-cry from the claims made by SK Hynix, who have only talked about a theoretical 600-layer NAND configuration. While the 176-layer, 7-gen V-NAND is only now entering mass production and the final stages of product development, Samsung has already taped out the initial batches of their 8th-gen V-NAND, which feature "more than 200 layers". It's likely that Samsung's 1,000-layer claim actually looks towards the future in a timeframe of decade(s?) and isn't actually something to look forward to in the approximate future.