TSMC: N2 To Start With Just GAAFETs, Add Backside Power Delivery Later

https://ift.tt/3ahdUzE

When TSMC initially introduced its N2 (2 nm class) process technology earlier this month, the company outlined how the new node would be built on the back of two new cutting-edge fab techniques: gate-all-around transistors, and backside power rails. But, as we've since learned from last week's EU symposium, TSMC's plans are a bit more nuanced than first announced. Unlike some of their rivals, TSMC will not be implementing both technologies in the initial version of their N2 node. Instead, the first iteration of N2 will only be implementing gate-all-around transistors, with backside power delivery to come with a later version of the node.

So far, TSMC has mentioned two distinctive features of N2: nano sheet gate-all-around (GAA) transistors, and backside power rails. GAA transistors have two unique advantages over FinFETs: they solve many challenges associated with the leakage current since GAAFET's channels are horizontal and are surrounded by gates around all four sides. Meanwhile, backside power rail enabled improved power delivery to transistors, which increases performance and lowers power consumption.

But, as it turns out, TSMC is not planning to start with both nanosheet GAA transistors and backside power rails in the initial generation of its N2 process technology. As disclosed by the company last week at their EU symposium, the first generation of N2 will only feature gate-all-around transistors. Backside power delivery, on the other hand, will come later with more advanced implementations of N2.

At this point the company hasn't said too much as to why they're not rolling out backside power delivery as part of their initial N2 node. But, in discussing the bifurcation, TSMC has noted that backside power delivery will ultimately add additional process steps, which the company is seemingly looking to avoid on their first try with GAAFETs.

The lack of backside power delivery in the original version of the N2 fabrication technology perhaps explains rather moderate performance improvement of N2 when compared to N3E node. While for high-performance computing (CPUs, accelerators, etc.) a 10% to 15% performance improvement at the same power and complexity does not seem to be impressive, a 25% to 30% power drop at the same speed and complexity seems to be very good for mobile applications. 

Advertised PPA Improvements of New Process Technologies Data announced during conference calls, events, press briefings and press releases
	TSMC
	N5 vs N7	N3 vs N5	N3E vs N5	N2 vs N3E
Power	-30%	-25-30%	-34%	-25-30%
Performance	+15%	+10-15%	+18%	+10-15%
Chip Density*	?	?	~1.3X	>1.1X
Volume Manufacturing	Q2 2022	H2 2022	Q2/Q3 2023	H2 2025

*Chip density published by TSMC reflects 'mixed' chip density consisting of 50% logic, 30% SRAM, and 20% analog. 

Considering that TSMC always offers multiple versions of its nodes, that TSMC has several variants planned for N2 is not all that surprising. Nonetheless, it is a bit odd to see that TSMC is taking a rather long road to backside power delivery.

Compared and contrasted to the competition, this will end up being a notable difference from how rival Intel is planning to handle their own GAAFET/backside power transition with the Intel 20A process. Intel intends to introduce its GAA RibbonFET transistors and PowerVia interconnects together in mid-2024 – going so far as to create an internal pseudo node just to focus on RibbonFET development. TSMC, on the other hand, is taking a more cautious approach to risks and innovations, one which potentially has TSMC moving at a slower pace, but is also an approach that has traditionally been a better fit for TSMC's need to deliver more constant and consistent updates to its fab offerings.

And while we're still a few years out, it will be interesting to see what this means for the competitiveness of TSMC's first-generation N2 node. Will a GAAFET process without backside power delivery be at a meaningful disadvantage? Per current schedules, we'll find out the answer to that in the second half of 2025, when TSMC's first N2 node is slated to enter high-volume manufacturing (HVM).

from AnandTech https://ift.tt/veHT8hp
via IFTTT

GPU stock recovery derails the bundle gravy train for motherboard makers

https://ift.tt/nNrCF0K

Motherboard sales are predicted to drop quite sharply as 2022 rumbles on, and one of the reasons for that might surprise you: namely the easing of the graphics card shortage (and prices therein).

This news comes from sources that DigiTimes tapped, who have knowledge of the purported predictions of the two biggest motherboard players in the PC component market. Asus and Gigabyte represent 70% of motherboard sales, or thereabouts, and both companies anticipate that the volume of units shifted will drop by 25% through the rest of this year.

Apparently motherboard shipment numbers for Q2 have already shown a drop in numbers – a larger than expected fall, in fact – and forecasts show this situation will deteriorate further in Q3 (and quite possibly in Q4, too).

A drop of a quarter in terms of mobo sales is a hefty downward turn, and the main contributing factors cited are waning demand in the Chinese DIY PC building market, and also the fact that the end of the road has been reached for bundling GPUs with motherboards in order to make for a tempting (but pricey) bundle proposition.

This is a practice you’re likely familiar with since GPU stock levels started to really suffer after the pandemic began. When gamers couldn’t get hold of a graphics card – at all, let alone at inflated prices – they were willing to spend even more cash to secure one as part of a hardware bundle.

That included GPU plus motherboard bundles, and in some cases with system RAM as well. Indeed, there were a small number of desperate folks who would buy an entire PC just to get the RTX 3000 or RX 6000 graphics card inside them, swap that GPU into their rig, put their old GPU into that freshly bought PC, then sell it off on eBay (or indeed just sell all the components off separately).

A messy and frankly ridiculous way to go about things, but that’s how frankly ridiculous trying to buy a GPU got at the worst times.

---

Analysis: Fresh hardware not coming to the rescue for mobo sales, either

(Image credit: MSI)

The GPU crisis finally easing is great news for consumers, of course, but bad news for motherboard makers. Not that there’ll be much sympathy for the latter, considering that effectively forcing motherboard sales upwards by bundling them with much-sought-after graphics cards was, shall we say, not the kindest of practices.

It’s clear, then, that GPU plus mobo bundles must’ve been popular enough, if they helped to drive motherboard sales to an extent that there’s a measurable impact of this practice now becoming irrelevant.

But wait a minute, you might be thinking – aren’t there new processors from AMD and Intel coming later this year that should help to push up sales volumes of motherboards? This is because AMD’s Zen 4 chips are going to need a new socket and therefore motherboard, and we’ve got Intel’s 13th-gen Raptor Lake CPUs inbound as well.

However, the sources DigiTimes spoke to, believe that these fresh hardware launches aren’t going to do anything to help stop the slide in motherboard sales – although if things change in the broader world, that could have a more positive impact to stem losses. By which they mean the end of the war in Ukraine, or a slowdown in the seemingly relentless pace of inflation.

Of course, Raptor Lake will have a limited impact anyway in terms of motherboard upgrades, because anyone who bought Alder Lake won’t need to upgrade their mobo (they’ll use the same socket). And Zen 4 may arrive a bit too late in the year to pep up motherboard sales all that much.

While the rumor mill indicates a September launch for Zen 4 (and has done several times), that could be wrong, or indeed that might be the timeframe for a reveal – but it may not be until a while later that an appreciable volume of Ryzen 7000 processors are on shelves.

Via Tom’s Hardware

from TechRadar: computing components news https://ift.tt/r8WJDBS
via IFTTT

AMD Updates Ryzen Embedded Series, R2000 Series With up to Four Cores and Eight Threads

https://ift.tt/TACqZHg

One area of AMD's portfolio that perhaps doesn't garner the same levels of attention as its desktop, mobile, and server products is its embedded business. In early 2020, AMD unveiled its Ryzen Embedded R1000 platform for the commercial and industrial sectors and the ever-growing IoT market, with low-powered processors designed for low-profile systems to satisfy the mid-range of the market.

At Embedded World 2022 in Nuremberg, Germany, AMD has announced its next-generation of Ryzen Embedded SoCs, the R2000 series. Offering four different SKUs ranging from 2C/4T up to 4C/8T, which is double the core count of the previous generation, AMD claims that the R2000 series features up to 81% higher CPU and graphics performance.

The AMD Ryzen Embedded R2000 Series compared to the previous generation (R1000), now has double the core count, with a generational swing from Zen to the more efficient and higher performance Zen+ cores. All four SKUs announced feature a configurable TDP, with the top SKU, the R2544, operating at between 35 and 54 W. More in line with the lower power target of these SoCs, the bottom SKU (R2312) has a configurable TDP of between 12 and 35 W.

AMD Ryzen Embedded R2000-Series APUs
AnandTech	Core/ Thread		Base Freq (MHz)	1T Boost Freq (MHz)	Memory Support	L2 Cache	L3 Cache	GPU CU's	TDP Range (W)	Launch  (Expected)
R2544	4	8	3350	3700	DDR4-3200	2 MB	4 MB	8	35-54	October 22
R2514	4	8	2100	3700	DDR4-2667	2 MB	4 MB	8	12-35	October 22
R2314	4	4	2100	3500	DDR4-2667	2 MB	4 MB	6	12-35	In Production
R2312	2	4	2700	3500	DDR4-2400	1 MB	2 MB	3	12-25	In Production

Another element delivering additional performance compared to the previous generation is better iGPU performance via increasing the number of Radeon Vega graphics compute units. The entry R2312 SKU comes with 3 CUs, while the R2544 comes with 8 CUs. The Ryzen Embedded R2000 series also benefits from newer video decode and display processor blocks, bringing support for decoding 4Kp60 video and driving up to three 4K displays.

AMD has also equipped the SoCs with 16 PCIe Gen 3 lanes on the R2314, R2514, and R2544 SKUs, while the R2312 gets eight. The R2000 series has support for two SATA 3.0 ports, up to six USB ports with a mixture of USB 3.2 G2 and USB 2.0, and OS support for Microsoft Windows 11/10 and Linux Ubuntu LTS. 

The application benefits of AMD's Ryzen Embedded R2000 series include the commercial and industrial sectors, as well as robotics, with a planned product availability of up to 10 years, ensuring a long life cycle for each product. Some of AMD's Ryzen Embedded R2000's Ecosystem partners include Advantech for its gaming and gambling machines, as well as DFI, IBASE, and Sapphire, so these new SoCs are already being adopted and planned into existing thin-client and small form factor systems.

AMD states that the Ryzen Embedded R2544 (4C/8T) and R2514 (4C/8T) will be available sometime in October 22, while the R2314 and R2312 SKUs are currently in production.

Source: AMD

from AnandTech https://ift.tt/lOVCuB9
via IFTTT

TSMC to Expand Capacity for Mature and Specialty Nodes by 50%

https://ift.tt/312xY0Q

TSMC this afternoon has disclosed that it will expand its production capacity for mature and specialized nodes by about 50% by 2025. The plan includes building numerous new fabs in Taiwan, Japan, and China. The move will further intensify competition between TSMC and such contract makers of chips as GlobalFoundries, UMC, and SMIC.

When we talk about silicon lithography here at AnandTech, we mostly cover leading-edge nodes used produce advanced CPUs, GPUs, and mobile SoCs, as these are devices that drive progress forward. But there are hundreds of device types that are made on mature or specialized process technologies that are used alongside those sophisticated processors, or power emerging smart devices that have a significant impact on our daily lives and have gained importance in the recent years. The demand for various computing and smart devices in the recent years has exploded by so much that this has provoked a global chip supply crisis, which in turn has impacted automotive, consumer electronics, PC, and numerous adjacent industries.

Modern smartphones, smart home appliances, and PCs already use dozens of chips and sensors, and the number (and complexity) of these chips is only increasing. These parts use more advanced specialty nodes, which is one of the reason why companies like TSMC will have to expand their production capacities of otherwise "old" nodes to meet growing demand in the coming years.

But there is another market that is about to explode: smart cars. Cars already use hundreds of chips, and semiconductor content is growing for vehicles. There are estimates that several years down the road the number of chips per car will be about 1,500 units – and someone will have to make them. Which is why TSMC rivals GlobalFoundries and SMIC have been increasing investments in new capacities in the last couple of years.

TSMC, which has among the largest CapEx budgets in the semiconductor industries (which is challenged only by Samsung) has in recent years been relatively quiet about their mature and specialty node production plans. But at their 2022 TSMC Technology Symposium, the company outlined its plans formally.

The company is investing in four new facilities for mature and specialty nodes:

• Fab 23 Phase 1 in Kumamoto, Japan. This semiconductor fabrication facility will make chips using TSMC's N12, N16, N22, and N28 nodes and will have a production capacity of up to 45,000 300-mm wafer starts per month.
• Fab 14 Phase 8 in Tainan, Taiwan.
• Fab 22 Phase 2 in Kaohsiung, Taiwan.
• Fab 16 Phase 1B in Nanjing, China. TSMC currently makes chips on its N28 in China, though the new phase was once rumored to be capable of making chips using more advanced nodes.

Increasing mature/specialized capacity by 50% over the next three years is a big shift for the company, and one that will improve TSMC's competitive positions on the market. What is perhaps more important is that the company's specialty nodes are largely based on its common nodes, which allows at least some companies to re-use IP they once developed for compute or RF for a new application. 

"[Our] specialty technology is quite unique as it is based on common technology platform  [logic technology platform], so our unique strategy is to allow our customer to share or reuse many of the [common] IP," said Kevin Zhang, senior vice president of business development at TSMC. "For example, you have RF capability, you build that RF on a common logic platform, but later you find 'hey someone need a so-called ULV feature to support an IoT product application.' You want to build that on a common platform so you can allow different product lines to be able to share IP across the board, this is very important for our customers so we do want to provide a integrated platform to address the market needs of customer from product perspective.' 

There are other advantages too. For example, TSMC's N6RF allows chip designers to combine high-performance logic with RF, which enables them to build products such as modems and other, more unique solutions. Many companies are already familiar with TSMC's N6 logic node, so now they have an opportunity to add RF connectivity to something that benefits from high performance. GlobalFoundries has a similar approach, but since the U.S.-based foundry does not have anything comparable to TSMC's N6, TSMC has an indisputable advantage here.

With its common platform approach for mature nodes as well as specialized technologies, and 50% more capacity, TSMC will be able to offer the world more chips for smart and connected devices in the coming years. Furthermore, it will also benefit TSMC by significantly increasing the company's revenues from mature and specialized nodes, as well as increasing pressure on their rivals.

from AnandTech https://ift.tt/ZLSiVhj
via IFTTT

TSMC Unveils N2 Process Node: Nanosheet-based GAAFETs Bring Significant Benefits In 2025

https://ift.tt/4s3O1B5

At its 2022 Technology Symposium, TSMC formally unveiled its N2 (2 nm class) fabrication technology, which is slated to go into production some time in 2025 and will be TSMC's first node to use their nanosheet-based gate-all-around field-effect transistors (GAAFETs). The new node will enable chip designers to significantly reduce the power consumption of their products, but the speed and transistor density improvements seem considerably less tangible.

TSMC's N2 is a brand-new platform that extensively uses EUV lithography and introduces GAAFETs (which TSMC calls nanosheet transistors) as well as backside power delivery. The new gate-all-around transistor structure promises well-published advantages, such as greatly reduced leakage current (now that the gates are around all four sides of the channel) as well as ability to adjust channel width to increase performance or lower power consumption. As for the backside power rail, it is generally designed to enable better power delivery to transistors, offering a solution to the problem of increasing resistances in the back-end-of-line (BEOL). The new power delivery is slated to increase transistor performance and lower power consumption.

From feature set standpoint, TSMC's N2 looks like a very promising technology. As for actual numbers, TSMC promises that N2 will allow chip designers to increase performance by 10% to 15% at the same power and transistor count, or reduce power consumption at the same frequency and complexity by 25% ~ 30%, all the while increasing chip density by over 1.1-fold when compared to N3E node.

Advertised PPA Improvements of New Process Technologies Data announced during conference calls, events, press briefings and press releases
	TSMC
	N5 vs N7	N3 vs N5	N3E vs N5	N2 vs N3E
Power	-30%	-25-30%	-34%	-25-30%
Performance	+15%	+10-15%	+18%	+10-15%
Chip Density*	?	?	~1.3X	>1.1X
Volume Manufacturing	Q2 2022	H2 2022	Q2/Q3 2023	H2 2025

*Chip density published by TSMC reflects 'mixed' chip density consisting of 50% logic, 30% SRAM, and 20% analog. 

Versus N3E, the performance improvements and power reductions enabled by TSMC's N2 node are in line with what the foundry's new nodes typically bring in. But the so-called chip density improvements (which should reflect transistor density gains) are just a little over 10%, which is not particularly inspiring, especially considering that N3E already offers a slightly lower transistor density when compared to vanilla N3. Keeping in mind that SRAM and analog circuits barely scale these days, mediocre improvements in transistor density of actual chips should probably be expected these days. However, a chip density improvement of 10% in about three years is certainly not great news for GPUs and other chips that live or die based on rapidly increasing their transistor counts. 

Bearing in mind that by the time TSMC's N2 enters production the company will also have the density-optimized N3S node, it would appear that the foundry will have two process technologies based on different types of transistors yet offering very similar transistor densities, something that has never happened before.

As usual, TSMC will offer their N2 node with various features and knobs to allow chip designers to optimize for things like mobile and high-performance computing designs (note that TSMC calls HPC everything that is not mobile, automotive or specialty. which includes everything from a low-power laptop CPU to a high-end compute GPU aimed at supercomputers). Also, platform offerings include something that TSMC calls 'chiplet integration', which probably means that TSMC enable its customers to easily integrate N2 chips into multi-chiplet packages made using various nodes. Since transistor density scaling is slowing down and new process technologies are getting more expensive to use, multi-chiplet packages are going to become more common in the coming years as developers will be using them to optimize their designs and costs.

TSMC expects to start risk production of chips using its N2 fabrication process sometimes in the second half of 2024, which means that the technology should be available for high volume manufacturing (HVM) of commercial products in the second half of 2025. But, considering the length of modern semiconductor production cycles, it's likely more pragmatic to expect the first N2 chips to become available either very late in 2025 or 2026, if everything goes as planned.

from AnandTech https://ift.tt/RZOkb6m
via IFTTT

TSMC Readies Five 3nm Process Technologies, Adds FinFlex For Design Flexibility

https://ift.tt/DR012x9

Taiwan Semiconductor Manufacturing Co. on Thursday kicked off its 2022 TSMC Technology Symposium, where the company traditionally shares it process technology roadmaps as well as its future expansion plans. One of the key things that TSMC is announcing today are its leading-edge nodes that belong to its N3 (3 nm class) and N2 (2nm class) families that will be used to make advanced CPUs, GPUs, and SoCs in the coming years.

N3: Five Nodes Over Next Three Years

As fabrication processes get more complex, their pathfinding, research, and development times get stretched out as well, so we no longer see a brand-new node emerging every two years from TSMC and other foundries. With N3, TSMC's new node introduction cadence is going to expand to around 2.5 years, whereas with N2, it will stretch to around three years. 

This means that TSMC will need to offer enhanced versions of N3 in order to meet the needs of its customers who are still looking for a performance per watt improvement as well as transistor density bump every year or so. Another reason why TSMC and its customers need multiple versions of N3 is because the foundry's N2 relies on all-new gate-all-around field-effect transistors (GAA FETs) implemented using nanosheets, which is expected to come with higher costs, new design methodologies, new IP, and many other changes. While developers of bleeding-edge chips will be quick to jump to N2, many of TSMC's more rank & file customers will stick to various N3 technologies for years to come.

At its TSMC Technology Symposium 2022, the foundry talked about four N3-derived fabrication processes (for a total of five 3 nm-class nodes) — N3E, N3P, N3S, and N3X — set to be introduced over the coming years. These N3 variants are slated to deliver improved process windows, higher performance, increased transistor densities, and augmented voltages for ultra-high-performance applications. All these technologies will support FinFlex, a TSMC "secret sauce" feature that greatly enhances their design flexibility and allows chip designers to precisely optimize performance, power consumption, and costs. 

Advertised PPA Improvements of New Process Technologies Data announced during conference calls, events, press briefings and press releases
	TSMC
	N4 vs N5	N4P vs N5	N4P vs N4	N4X vs N5	N4X vs N4P	N3 vs N5	N3E vs N5
Power	lower	-22%	-	?	?	-25-30%	-34%
Performance	higher	+11%	+6%	+15% or more	+4% or more	+10-15%	+18%
Logic Area Reduction* % Logic Density*	0.94x -6% 1.06x	0.94x -6% 1.06x	-	?	?	0.58x -42% 1.7x	0.625x -37.5% 1.6x
Volume Manufacturing	2022	2023	H2 2022	2023	2023	H2 2022	Q2/Q3 2023

*Note that TSMC only started to publish transistor density enhancements for analog, logic, and SRAM separately around 2020. Some of the numbers still reflect 'mixed' density consisting of 50% logic, 30% SRAM, and 20% analog. 

N3 and N3E: On Track for HVM

TSMC's first 3 nm-class node is called N3 and this one is on track to start high volume manufacturing (HVM) in the second half of this year. Actual chips are set to be delivered to customers in early 2023.This technology is mostly aimed at early adopters (read: Apple and the like) who can invest in leading-edge designs and would benefit from the performance, power, area (PPA) advantages offered by leading-edge nodes. But as it's tailored for particular types of applications, N3 has a relatively narrow process window (a range of parameters that produce a defined result), which may not be suitable for all applications in terms of yields.

This is when N3E comes into play. The new technology enhances performance, lowers power, and increases the process window, which results in higher yields. But the trade-off is that the node features a slightly reduced logic density. When compared to N5, N3E will offer a 34% reduction in power consumption (at the same speed and complexity) or an 18% performance improvement (at the same power and complexity), and will increase logic transistor density by 1.6x. 

It is noteworthy that, based on data from TSMC, N3E will offer higher clockspeeds than even N4X (due in 2023). However the latter will also support ultra-high drive currents and voltages of above 1.2V, at which point it will be able to offer unbeatable performance, but with very high power consumption. 

In general, N3E looks to be a more versatile node than N3, which is why it is not surprising that TSMC has more '3nm tape outs' at this point than it had with its 5 nm-class node at a similar point of its development.

Risk production of chips using N3E is set to start in the coming weeks (i.e., in Q2 or Q3 2022) with HVM set for mid-2023 (again, TSMC does not disclose whether we are talking about Q2 or Q3). So expect commercial N3E chips to be available in late 2023 or early 2024.

N3P, N3S, and N3X: Performance, Density, Voltages

N3's improvements do not stop with N3E. TSMC is set to bring out N3P, a performance-enhanced version of its fabrication process, as well as N3S, density-enhancing flavor of this node, some time around 2024. Unfortunately, TSMC is not currently disclosing what improvements these variants will offer compared to baseline N3. In fact, at this point TSMC does not even show N3S in all versions of its roadmap, so it is really not a good business to try guessing its characteristics.

Finally, for those customers who need ultra-high performance no matter power consumption and costs, TSMC will offer N3X, which is essentially an ideological successor of N4X. Again, TSMC is not revealing details about this node other than that it will support high drive currents and voltages. We might speculate that N4X could use backside power delivery, but since we are talking about a FinFET-based node and TSMC only going to implement backside power rail in nanosheet-based N2, we are not sure this is the case. Nonetheless, TSMC probably has a number of aces up its sleeve when it comes to voltage increases and performance enhancements.

FinFlex: N3's Secret Sauce

Speaking of enhancements, we should definitely mention TSMC's secret sauce for N3: FinFlex technology. In short, FinFlex allows chip designers to precisely tailor their building blocks for higher performance, higher density, and lower power.

When using a FinFET-based node, chip designers can choose between different libraries using different transistors. When developers need to minimize die size and save power at the cost of performance, they use double-gate single-fin (2-1) FinFETs (see the illustration). But when they need to maximize performance at the trade-off of die size and higher power, they use triple-gate dual-fin (3-2) transistors. When developers need a balance, they go with dual-gate dual-fin (2-2) FinFETs. 

Currently, chip designers have to stick to one library/transistor type either for the whole chip or the whole block in a SoC design. For example, CPU cores can be implemented using 3-2 FinFETs to make them run faster, or 2-1 FinFETs to reduce their power consumption and footprint. This is a fair tradeoff, but it's not ideal for all cases, especially when we are talking about 3 nm-class nodes that will be more expensive to use than existing technologies.

For N3, TSMC's FinFlex technology will allow chip designers to mix and match different kinds of FinFETs within one block to precisely tailor performance, power consumption, and area. For complex structures like CPU cores, such optimizations give a lot of opportunities to increase core performance while still optimizing die sizes. So, we are eager to see how SoC designers will be able to take advantage of FinFlex in the looming N3 era.

FinFlex is not a substitute for node specialization (performance, density, voltages) as process technologies have greater differences than the ibraries or transistor structures within a single process technology, but FinFlex looks to be a good way to optimize performance, power, and costs for TSMC's N3 node. Ultimately, this technology will bring the flexibility of FinFETs a little closer to that of nanosheet-based GAAFETs, which are slated to offer adjustable channel widths to get higher performance or reduce power consumption.

Summary

Like TSMC's N7 and N5, N3 will be another family of long-lasting nodes for the world's largest contrast maker of semiconductors. Especially with the jump to nanosheet-based GAAFETs coming up at 2nm for TSMC, the 3nm family will be the final family of "classic" leading-edge FinFET nodes from the firm, and one that a lot of customers will stick to for several years (or more). Which, in turn, is why TSMC is prepping multiple versions of N3 tailored for different applications – as well as FinFlex technology to give chip designers some additional flexibility with their designs.

The first N3 chips are set to enter production in the coming months and arrive to the market in early 2023. Meanwhile, TSMC will keep producing semiconductors using its N3 nodes long after it introduces its N2 process technology in 2025.

from AnandTech https://ift.tt/U3f9kZ5
via IFTTT

The ASUS ROG Maximus Z690 Hero Motherboard Review: A Solid Option For Alder Lake

https://ift.tt/gYed1cV

Over the last six months since Intel launched its 12th Gen Core series of processors, we've looked at several Alder Lake desktop CPUs and seen how competitive they are from top to bottom - not just in performance but price too. To harness the power of Alder Lake, however, there are many options in terms of Z690 motherboards, and today we're taking a look at one of ASUS's more premium models, the ROG Maximus Z690 Hero.

They say hard times don't create heroes, but ASUS has done for many years with good results. Equipped with plenty of top-tier features such as Thunderbolt 4, Intel's Wi-Fi 6E CNVi, and support for up to DDR5-6400 memory, it has enough to make it a solid choice for gamers and enthusiasts. It's time to see if the Z690 Hero option stacks up against the competition and if it can sparkle in a very competitive LGA1700 market.

from AnandTech https://ift.tt/uxASyFX
via IFTTT

AMD’s RDNA 3 graphics cards could arrive much later than Nvidia’s RTX 4000

https://ift.tt/1AGVEjb

AMD’s next-gen graphics cards might not be out until November, with Zen 4 CPUs debuting just before the RDNA 3 GPUs.

This rumor comes from one of the Twitter hardware leakers with a better track record, Greymon55, but we should of course help ourselves to a couple of handfuls of salt here – and remember that even if this is AMD’s intended plan right now, that could change.

Launch: late October, no later than mid-November😉June 14, 2022

See more

As you can see, the purported launch date for what should be RX 7000 GPUs is late October, or possibly mid-November, and no later than that (well, it couldn’t really be much later given that RDNA 3 cards are scheduled for this year, and at that point, there’s not much of 2022 left).

It’s notable that Greymon sounds confident enough in this prediction, adding in a further tweet that this is “new info and very reliable”.

The leaker also responds to clarify that Ryzen 7000 CPUs will launch a bit before RDNA 3, possibly in late September or early October. Again, let’s be careful about putting too much weight on that assertion, but this isn’t the first time we’ve heard a September date pinned on the launch of Zen 4 chips.

---

Analysis: Nvidia to beat AMD to the next-gen GPU punch – and by some way?

Previously the rumor mill has theorized that AMD is going to kick off with RDNA 3 graphics card in Q4, but all along, the only promise Team Red has made is that these next-gen GPUs will be out before the end of 2022.

There was a suggestion from one leak that the mid-range RX 7000 models – built on the Navi 33 GPU – might just sneak out before Q4 rolls around, but this new rumor certainly pours cold water on that idea. It seems increasingly clear that we’re looking at Q4 now, and indeed this is a reality check that we’ll be lucky to see the first RDNA 3 product in October, really, as the ‘late’ October prediction makes that month seem a bit more tentative.

The problem for AMD is that Nvidia is rumored to be getting its RTX 4000 series off the ground in September most likely – though we’ve heard a few dates floating about (even August, in fact, for what’s supposedly going to be the first Lovelace GPU, and that’s the RTX 4090).

So broadly, AMD is looking like it might be a couple of months behind Nvidia with RDNA 3 graphics cards, which would give Team Green quite an opportunity to establish itself in the market with next-gen products. Not an ideal situation for AMD, particularly as if the most recent rumor is right, by November Nvidia will have its RTX 4070 deployed, alongside the RTX 4090 and 4080 (which are supposedly due in September and October respectively).

While we don’t know what card AMD will produce from its RX 7000 deck first, it’s interesting to note that rumors previously pointed to Navi 33. When questioned about whether that might still be the case – or if it could be the flagship Navi 31 offering debuting first, as VideoCardz now seems to believe – Greymon wasn’t drawn to comment (whereas the leaker responded to other questions in that tweet). That seems to indicate that while confident about the overall launch schedule, Greymon isn’t so sure about which graphics card AMD might get out of the door initially.

AMD may feel the need to make a big impact with the flagship as the initial launch – running with an all-new multi-chip design as was recently confirmed – or the better (or indeed only viable) strategy may be to get a more affordable mid-range Navi 33 model launched first.

Or alternatively, we could witness the simultaneous arrival of multiple RDNA 3 GPUs, but if Nvidia already has its top three Lovelace graphics cards out there – pushed out in staggered fashion across the previous couple of months – by the time that AMD launch rolls around, that’s going to give Team Green a fair old head start, as we already mentioned.

Unless Nvidia elects to use the extra time to hold the release of RTX 4000 GPUs for an additional month, kicking off with the RTX 4090 in October rather than September, in order to sell off existing RTX 3000 stock, which is another possibility that’s filtered down from the grapevine recently. The rumors around this are something of a writhing nest of possibilities, for sure, but one way or another, the overriding sentiment is that Nvidia appears to have an advantage in terms of having the luxury of time, and perhaps still a choice to make on how to best use that.

from TechRadar: computing components news https://ift.tt/VPwCayZ
via IFTTT

Intel Core i9-13900 leak shows a 24-core CPU that could scare the hell out of AMD

https://ift.tt/uHB9V5i

Intel’s Raptor Lake CPUs are set to arrive later in 2022, and we’ve just caught a glimpse of the potential power of the top dog 13th-gen processor – which could frighten AMD’s engineers, at least with the apparent gains in some respects.

The Core i9-13900 has appeared in multiple results in the SiSoftware Sandra database, as VideoCardz reports, with SiSoftware rounding up those benchmarks as a ‘performance preview’ in an article which also summarizes the purported specs for the flagship CPU (view all this with a skeptical eye, naturally).

As previously rumored, the Core i9-13900 will run with eight performance cores backed with 16 efficiency cores, meaning it’s a 24-core CPU (though it’ll only have 32-threads, as the efficiency cores don’t feature hyper-threading). In other words, the performance cores will remain the same in number as Alder Lake, but Intel is doubling up on the efficiency cores – a big jump indeed.

We’re supposedly looking at a 20% larger L3 cache with Raptor Lake compared to Alder Lake, too, and L2 cache is twice the size for both performance and efficiency cores (2MB per core for the former, and 4MB per cluster of four cores for the latter).

We’ll apparently also see support for faster DDR5 system RAM (5600MT/s, which is a 16% increase compared to Intel’s current 12th-gen chips), if the details provided by SiSoftware are correct, of course (again, let’s remain skeptical).

So, what about those performance findings which are based on several SiSoftware Sandra benchmarks of a sample Core i9-13900 chip?

In ALU/FPU (arithmetic) testing, the Raptor Lake flagship managed a huge improvement of between 33% to 50% over the current Core i9-12900’s benchmark results provided for comparison. Furthermore, this was at pre-release clocks for the Raptor Lake sample, namely 3.7GHz for the performance cores and 2.76GHz for the efficiency ones (compared to 5GHz and 3.8GHz for the Alder Lake flagship).

With vectorized/SIMD tests, the Core i9-13900 made gains to the tune of around 5% to 8% over the Alder Lake flagship.

---

Analysis: How much bite might this Raptor have, then?

While any benchmarking of early sample silicon needs to be regarded with caution, what we see here – for heavyweight performance benchmarks – give us some broad ideas of what Intel might be looking to achieve with Raptor Lake.

That massive leap for the ALU/FPU tests is certainly very impressive, and although the gains elsewhere aren’t nearly as eye-opening, SiSoftware observes they are still ‘encouraging’, and we do have to bear in mind the nerfed clocks of the engineering sample being tested here.

With Intel keeping performance cores at the same number (8), but doubling up on efficiency cores to a huge 16, we can surely expect some compelling multi-core performance boosts, even if the acceleration with single-core performance is more modest.

Given that there’s no movement with performance cores compared to Alder Lake, and 16 rather than ight efficiency cores, we can also guess that Intel is aiming to do better with power-efficiency with Raptor Lake. Although doubtless with that many efficiency cores being introduced, clock speeds will need to be reined in somewhat compared to Alder Lake, as doubtless something has to give with the power requirements, of course.

With better efficiency, Raptor Lake could be great news for laptops, but this matters less for desktop PCs, and it’s here that there remains an element of doubt about exactly how Raptor Lake might perform when we’re hearing some tantalizing things about Zen 4, and how far AMD might be pushing with clocks speeds. Indeed, Team Red looks set to make some seriously big leaps with its next-gen chips able to muster all-core boosts of up to 5.5GHz (admittedly situationally – we discuss that in more depth here).

Still, Alder Lake is winning the battle in terms of gaming versus AMD’s now rapidly aging Zen 3 processors right now, and Raptor Lake is also rumored to boost to impressive levels – plus we mustn’t forget about that bolstered cache, which is going to come to the aid of gamers running Intel’s 13th-gen silicon, too.

Clearly it’s too early to start making predictions with any weight around who might win the battle of the next-gen CPUs, but it’s shaping up to be quite a scrap. And as ever when that’s the case, the winner hopefully ends up being the consumer, as both Intel and AMD drive each other to do better – or to be more relatively affordable, to even things up on the pricing front, perhaps.

from TechRadar: computing components news https://ift.tt/feAhxvp
via IFTTT

Intel 4 Process Node In Detail: 2x Density Scaling, 20% Improved Performance

https://ift.tt/nZFavsp

Taking place this week is the IEEE’s annual VLSI Symposium, one of the industry’s major events for disclosing and discussing new chip manufacturing techniques. One of the most anticipated presentations scheduled this year is from Intel, who is at the show to outline the physical and performance characteristics of their upcoming Intel 4 process, which will be used for products set to be released in 2023. The development of the Intel 4 process represents a critical milestone for Intel, as it’s the first Intel process to incorporate EUV, and it’s the first process to move past their troubled 10nm node – making it Intel’s first chance to get back on track to re-attaining fab supremacy.

Intel’s scheduled to deliver their Intel 4 presentation on Tuesday, in a talk/paper entitled “Intel 4 CMOS Technology Featuring Advanced FinFET Transistors optimized for High Density and High-Performance Computing”. But this morning, ahead of the show, they re publishing the paper and all of its relevant figures, giving us our first look at what kind of geometries Intel is attaining, as well as some more information about the materials being used.

from AnandTech https://ift.tt/wkBlU8e
via IFTTT

Intel Arc A730M leaks give hope with one hand, and take with the other

https://ift.tt/26yscLK

Intel’s second-fastest Arc Alchemist GPU for laptops has witnessed some leaked benchmarks, and they’re a mixed bag of results – though we should be very cautious around interpreting them.

There are two sets of benchmarks that come from a Weibo user – purportedly for both real-world gaming performance, and synthetic tests – as flagged up by HXL on Twitter (and spotted by Tom’s Hardware and VideoCardz).

The denizen of Chinese social media platform Weibo, who goes by the name ‘Golden Pig Upgrade’, has apparently purchased a Machenike gaming laptop (which have very recently become available, but only in China) with an Intel Arc A730M and Core i7-12700H (Alder Lake) processor. As mentioned, the A730M is part of the top-of-the-range A7 family, but is the second best offering next to the flagship A770M – these GPUs have 24 and 32 Xe-cores respectively.

Let’s look at the synthetic testing first, which saw the A730M hit a graphics score of 10,107 in 3DMark TimeSpy, which is faster than an Nvidia RTX 3070 laptop GPU (and not that far off the 3070 Ti, for that matter).

Furthermore, 3DMark Fire Strike scores were also shared for the Arc A730M – by the laptop maker, in this case – and that placed the Arc GPU a little lower, between the RTX 3060 and 3070.

Real-world gaming performance metrics turned out rather differently, though, going by this Weibo leak.

Golden Pig Upgrade tried out the Arc A730M with a small selection of games, seeing an average of 70 fps (frames per second) recorded with Metro Exodus in 1080p resolution (on high details), and 123 fps with F1 2020 on 1080p (high details). Assassin’s Creed Odyssey, however, only managed a way shakier looking 38 fps on 1080p (with very high details), plunging to jerky lows of 10 fps as the minimum frame rate recorded.

Also, speaking of minimum frame rates, while Metro Exodus managed a good average frame rate here, it did drop to some seriously poor lows of 9 fps – even more sluggish than Assassin’s Creed in that respect.

Clearly, the results here are something of a mixed bag, and certainly peg the A730M at a lower level than the 3DMark testing. As Tom’s Hardware points out, the Metro Exodus frame rates are comparable to the RTX 3060, but with Assassin’s Creed Odyssey, performance ranks more like the RTX 3050.

---

Analysis: A confusing overall picture of performance

There’s a lot to unpack here, but all of this should be taken with large heaps of assorted seasonings, for sure. There are a whole host of caveats, not least of which is that we should be wary of purported benchmark results posted on Weibo (although admittedly, the laptop maker Machenike did also supply a 3DMark result, as noted).

The 3DMark scores certainly seem impressive, and indeed perhaps overly so when you remember that this is Intel’s second best A7 GPU, and not even the flagship – yet it’s apparently outdoing the RTX 3070 in one benchmark. However, synthetic tests like these are not nearly as valuable a measurement as actual gaming performance, and Intel may have optimized its graphics driver considerably for these kinds of tests.

On the other hand, the graphics driver clearly hasn't been fully honed for some of the games tested, given those dips to really low minimum frame rates, which are indicative of 'drivers in progress'. Indeed, we don’t know which driver version was used here, but it’s clearly an early working version (one without official support – given that the Alchemist driver with A730M support was only just released today). The Weibo user also observed that Shadow of the Tomb Raider proved impossible to run for benchmarking purposes, simply crashing out and not loading at all.

There are other possible variables in the mix with the game benchmarks, too, like whether the Dynamic Tuning Technology (DTT) driver was enabled – this is Intel’s take on intelligently allocating power between the CPU and GPU, but it has been seen to still be somewhat wonky and hamper performance.

Overall, the potential performance of the A730M remains a difficult picture to decipher from these early benchmark glimpses, but it all points to one thing which we’ve theorized for a while now. Namely that Intel still needs to work on the software for Arc GPUs, and the delays we’re seeing on the desktop graphics card front are likely wrapped up in this need for further work on driver optimization.

The obvious problem with the delay is that as Intel spends more and more time hammering software into shape, the launch of next-gen graphics cards from both AMD and Nvidia gets closer – and then GPU comparisons will become a much more difficult area for Arc Alchemist products.

• AMD vs Nvidia: as it stands right now, who is king of the desktop GPU world?

from TechRadar: computing components news https://ift.tt/0GtrAsb
via IFTTT

Nvidia RTX 4080 GPU rumor could make you think about a power supply upgrade

https://ift.tt/TBEAuve

Nvidia’s RTX 4080 graphics card could crank up power usage to 100W more than the RTX 3080, if the GPU grapevine is on the money.

This latest rumor about next-gen Lovelace GPUs comes off the back of a great deal of speculation on the topic of power consumption (or TGP), and Twitter-based leaker Kopite7kimi now believes that the RTX 4080 will demand 420W from the power supply (PSU).

Possible RTX 4080, PG139-SKU360, has a 420W TGP, still uses AD103.June 6, 2022

See more

Obviously, this is just a rumor, so remain skeptical as ever, but as a general suggestion of where things may be heading with RTX 4000 power usage, it doesn’t sound too comforting on the face of it.

Remember that the RTX 3080 draws 320W (the original 10GB incarnation, that is), so as mentioned this is a theoretical 100W increase.

In fact, the RTX 3090 pulls 350W, so the 4080 is well ahead of that, and not too far off the level of the RTX 3090 Ti which has a TGP of 450W – and Nvidia recommends a PSU of 850W for that GPU, no less. (Remember, that’s a minimum spec too, meaning those who are going to be pushing things with faster top-end CPUs and overclocked components in their PCs would be wise to go for some decent headroom over and above that).

---

Analysis: A bitter PSU pill to swallow? Possibly, but that’s no real surprise

Does that mean we’re looking at an RTX 4080 which requires a power supply of 800W, then, or thereabouts? Well, that seems a likely possibility if that figure of 420W for the TGP is correct.

All this may seem rather outlandish, but it’s not really a surprise. There’s been plenty of talk about major power draw for Lovelace on the grapevine for ages now, as Nvidia apparently looks to put the pedal firmly to the metal in terms of accelerating performance. And regarding the RTX 4080 specifically, we’ve previously been given the expectation of something like 350W to 450W for the TGP. We were hoping 400W would sit about right if that guestimate turns out to be correct, but 420W isn’t far off.

Of course, that doesn’t mean this won’t be a bitter PSU pill for some gamers to swallow, as far from everyone has an 800W+ unit nestling in their PC. And it’s not just about extra cost with a power supply upgrade, as it’s one of the fiddlier procedures to undertake in terms of the cabling; the whole thing’s a bit of a faff.

The TGP situation could’ve been worse by all accounts, though, because Kopite7kimi was still talking about the possibility of the RTX 4080 hitting 450W not much more than a few days ago.

Naturally, it might be the case that this speculation, or the leaker’s sources, are considerably off the mark, but at this point, that seems unlikely (and the rest of the rumor mill doesn’t disagree, broadly). Furthermore, Kopite7kimi asserts that they are 99% confident in this most recent 420W prediction, which is a pretty strong statement to make.

Another interesting point here is that if the RTX 4080 does end up at 420W, for argument’s sake, where does that leave the RTX 4090? Most recently we’ve heard that the flagship will demand 450W, which isn’t all that much more than the purported draw for the 4080 – just 30W.

That has led some to suggest that we might be looking at more like 500W for the 4090, but that’s not necessarily the case. There could be only a 30W difference between the RTX 4080 and 4090 going by Nvidia’s past form, because indeed if you look at the 3080 and 3090, the gap was exactly that – a 320W power consumption for the former, and 350W for the latter. The obvious problem now, though, is that rather than just flagship territory – which is expected to be difficult and demanding GPU terrain – we’re looking at the xx80 tier supposedly calling for what could be an 800W (or maybe even greater) PSU requirement.

Whatever the case, these RTX 4000 graphics cards don’t sound like they’re going to be an easy ride for PSUs. The nearer we get to the Lovelace launch, the more accurate these predictions should theoretically be – and if the rumor mill is right about Nvidia’s planned release timeframe, the RTX 4090 could land as soon as August, which isn’t far off now.

• We also just heard about AMD’s RDNA 3 GPU power demands – and it’s better news

from TechRadar: computing components news https://ift.tt/an5xq8G
via IFTTT

Intel’s Arc Alchemist flagship GPU spotted in a gaming laptop

https://ift.tt/p51EuCN

Intel’s top dog laptop graphics card, the A770M, is in a high-end gaming laptop just announced by Clevo, paired with an Intel Alder Lake mobile CPU.

Clevo makes laptops that other manufacturers customize and then sell under their own brands (for example, Eurocom and XMG use Clevo designs), and the company just revealed the Clevo X270 on YouTube (as spotted by well-known hardware leaker HXL on Twitter).

Clevo X270https://t.co/0VUM5XTZ9G pic.twitter.com/t4Cqboa2jJJune 2, 2022

See more

The X270 comes loaded with an Arc Alchemist A770M, the flagship of Intel’s new laptop GPUs, alongside an Alder Lake H-series processor.

The brief teaser video also underlines that a robust cooling system is in place to ensure thermals remain favorable despite the hardware packed in here – in theory anyway – which includes a liquid metal topping for the CPU to ensure temperatures are 10-degrees cooler than they otherwise would be (or that’s the claim). A host of copper heat pipes – seven of them – are also present to facilitate better cooling of the internals (there’s a pair of fans, too).

Not much else is revealed about the spec, aside from the gaming laptop having a 17.3-inch Mini-LED display, and that the range of ports provided includes a pair of Thunderbolt 4 connectors.

---

Analysis: A speedy portable, no doubt, that’ll hopefully arrive before long

This notebook should be pretty fast when it comes to frame rates given that combo of an Alder Lake H-series chip and Arc Alchemist A770M. That said, we don’t yet know which 12th-gen Intel mobile CPU will be employed, but we’d expect it’ll be one of the peppier Core i7 or likely Core i9 models, given that it’s twinned with Intel’s flagship GPU.

The other thing we don’t yet know is exactly how performant the A770M will be, but it hopefully won’t disappoint – previously we’ve seen comparisons in the ballpark of the RTX 3060 or 3060 Ti (take such leaks with plenty of caution, of course).

Whatever the case, the power on tap should be impressive enough given that Clevo has opted to run with a Mini-LED panel here, marking this notebook out as a more premium gaming model (and as such, it’d hardly be sluggish).

Sadly, we don’t know when this gaming laptop is scheduled to come out, as the teaser clip doesn’t give us any idea. But its very presence surely indicates that the X270, and more to the point, gaming notebook models built on the design, shouldn’t be too far off…

• What about Intel Arc desktop GPUs? Sadly, the news on their launch is disappointing

Via Notebookcheck.net

from TechRadar: computing components news https://ift.tt/btLXgqY
via IFTTT

Intel Meteor Lake CPUs could arrive later than expected, requiring a new motherboard

https://ift.tt/p51EuCN

Intel’s Meteor Lake processors are due to arrive late in Q4 of next year, with laptop chips turning up earlier, according to a fresh rumor which also insists that the 14th-generation desktop CPUs will require a new socket (though that isn’t much of a surprise – we’ll discuss that in more detail later on).

In other words, this means that desktop owners who’ve bought into Alder Lake which introduced its own new socket – or the incoming next-gen Raptor Lake which will continue to use that socket – will need a new motherboard if they want to upgrade to Meteor Lake (which represents a big step forward for Intel in terms of finally getting to 7nm).

This rumor peddling comes from Moore’s Law is Dead (MLID) on YouTube, a regular source of hardware leaks that insists that the purported new LGA 2551 socket won’t be much different to the current LGA 1700 in size (just slightly bigger), but it’ll be a lot more dense with pins (having 50% more of them, in fact, the leaker believes – airing a photo of what’s supposedly a Meteor Lake chip showing the pin configuration).

We already knew Meteor Lake is due for 2023, but it’ll apparently be the fourth quarter before desktop chips are released, with laptop CPUs debuting earlier, most likely in Q3 (though the mobile processors could just sneak in at the very end of Q2, MLID doesn’t seem very convinced about that).

Arrow Lake, which would be Intel’s 15th-gen CPUs, are still set to turn up in 2024 (in the second half of the year), the leaker asserts, and as you might expect will stick with the same socket as Meteor Lake.

With Arrow Lake, the theory is that Intel will really load up with efficiency cores, piling on 32 of them with the flagship (alongside 8 performance cores, which is apparently what Team Blue is going to stick with for the next few generations, electing to drive forward with efficiency cores instead).

---

Analysis: Some slippage, but no real surprise on the socket front

As noted, Intel has already told us that Meteor Lake is due for 2023, but previous rumors indicated that the 14th-gen chips might come earlier in the year than MLID now believes. What seems to have happened is that Intel’s planned schedule has seen some slippage, so desktop Meteor Lake won’t arrive until the final quarter of next year now – and indeed the leaker reckons next-gen Raptor Lake has been held up too.

Apparently, we won’t see those 13th-gen desktop chips until potentially December 2022 (though a paper launch might happen around September or October – in other words, an initial reveal to at least whet our appetite).

At any rate, going back to Meteor Lake, with an apparent delay this silicon could end up squaring off against Zen 5, MLID theorizes, rather than Zen 4 as Intel wanted. Of course, take major handfuls of salt with all of this, and we could equally throw question marks over the possibility of AMD coming out with Zen 5 in a timely fashion next year, frankly.

The other news here is that Meteor Lake will shift socket and require a motherboard upgrade, meaning that the new LGA 1700 socket is only going to be good for two generations of processors, Alder Lake and Raptor Lake.

However, this is no big surprise really, as Meteor Lake is a big change – dropping to 7nm – and it’s fairly standard practice, for Intel at least, only to support a couple of generations of processors with a socket (LGA 1200, which preceded LGA 1700, was only good for Comet Lake and Rocket Lake, if you recall).

• AMD Ryzen 7000 CPUs could muscle in and steal Intel’s Raptor Lake thunder

Via VideoCardz

from TechRadar: computing components news https://ift.tt/wc7vG0X
via IFTTT