A55’s Exynos 1480 is a Huge Step Up from its Predecessor, the Exynos 1380

One of the major changes made to the A55 is the SoC, which is the Exynos 1480. This chipset is a huge step up from its predecessor, the Exynos 1380. In comparison to Qualcomm, it delivers performance equivalent to that of the Snapdragon 778G, a chipset that is now three years old.

This time, the Exynos 1480 has been captured in early benchmarks, and the performance metrics have shown significant improvements. Manufactured on Samsung’s 4nm (4LPP) process node, the SoC features an octa-core configuration, comprising 4x Cortex-A78 cores clocked at 2.75GHz and another 4x Cortex-A55 cores operating at 2.05GHz.

Benchmarks & Comparison

YouTuber MadCoby recently put the Samsung Galaxy A55 through its paces, and the new SoC delivered promising results. On Geekbench 6, the phone achieved a single-core score of 1,161 and a multi-core score of 3,398. These figures represent an impressive 18.5% and 22.7% improvement, respectively, compared to previous benchmarks.

On AnTuTu, the A55 scored a decent 713,993 points, showcasing a 20.4% increase in benchmark scores. For the CPU’s thermal efficiency, it was tested through a 15-minute CPU throttling test, where it maintained a relatively stable temperature without any sudden drops.

As for the GPU, the A55 uses Samsung’s Xclipse 530 GPU, based on the RDNA 2 architecture. This GPU unit notably outperforms the Mali-G68 found in the A54. In Geekbench 6 GPU Compute Score, the Xclipse 530 showed a significant 31.4% improvement over the Mali-G68.

This is all we know for now, but rest assured that we will keep you updated as new information becomes available.

The post Exynos 1480 Delivers Big Performance Gains in Early Benchmarks: 22% Higher CPU and 31% Better GPU Scores appeared first on Appuals.

Apple’s Roadmap for Sixteen Upcoming Devices Leaked in New Report

Within the list of sixteen devices, eight were reported to feature existing Apple silicon, including the A17 Pro (7th generation iPad mini), M2 (6th generation iPad Air), and A14 Bionic (11th generation iPad). The remaining eight devices were allegedly reported to sport upcoming hardware, which isn’t really too hard to guess. The M4 will succeed the M3, and the A17 Pro would serve as the predecessor to the A18 Pro.

The iPhone 16 lineup, which is most likely to be announced in the second half of this year, will come with the new A18 chips across the board. For reference, each iteration of iPhones for the past few years has come with different chipset configurations within the same line. The Pros get the new SoCs, while the vanilla models get last year’s Pros’ chips.

Likewise, the iPhone 15 Pro and Pro Max come with A17 Pro, whereas the iPhone 15 and 15 Plus use the year-old A16 Bionic. For the iPhone 16 series, the Pros might use the “A18 Pro” moniker, but the base variants will likely stick to something like the A18 or A18 Bionic.

The M4 Might Be Apple’s Way of Justifying iPad Pros’ Price Hike

The general expectation for the upcoming OLED iPads was that they would share the M3 chip with the new MacBook Air M3, both of which are expected to be released this month. However, this leak suggests the possibility that the new iPad might use the M4 chip. It’s important to note that these iPads are expected to come with a price tag nearly double that of their last-generation counterparts.

Given the significant price discrepancy, Apple might resort to marketing strategies to justify the potential price hike. For reference, the 11-inch model is expected to start at $1500, while the larger 13-inch variant is expected to be priced around $1800.

This is all we know for now, but rest assured that we will keep you updated as new information becomes available.

The post M4 Chip Could Come to the Upcoming OLED iPad Pros appeared first on Appuals.

Qualcomm Expects to Utilize a Multi-Foundry Approach for the Snapdragon 8 Gen 5

The report indicates that Qualcomm may adopt a multi-foundry strategy next year to manufacture the next generation of Snapdragon chips. Previously, Samsung fabricated Qualcomm’s flagship Snapdragon SoCs, including the Snapdragon 8 Gen 1. However, due to concerns about overheating and inefficiency, Qualcomm switched to TSMC, which has been their exclusive foundry partner since then.

This pattern could change next year. Not only is Samsung expected to excel as a mobile manufacturer again, but its foundry division might also make a comeback. While the standard Snapdragon 8 Gen 5 is reportedly slated for fabrication on TSMC’s 3nm N3E process node, the “For Galaxy” variant specifically designed for Samsung’s flagship phones could utilize a 2nm variant produced by Samsung Foundry on their SF2P node.

Will Snapdragon 8 Gen 5 See a Generational Gap Within the Same Lineup?

If the above leaks are considered to be accurate, a nodal jump from 3nm to 2nm within the same year in the same generation of SoC seems highly unlikely. If done, it might be hard for Qualcomm to justify this level of discrepancy to its other clients, especially at a sensitive time like this when competition is picking up from the likes of MediaTek and Google (with the expected improvement in Tensor G5 to match other flagship SoCs).

Since this, like other news, is simply an early leak, we’d advise our readers to take this news with a pinch of salt until stronger patterns or leaks surface online later. This is all we know for now, but rest assured that we will keep you updated as new information becomes available.

via: Revegnus

The post Snapdragon 8 Gen 5 ‘For Galaxy’ Variant Could Be Fabricated On Samsung’s 2nm Process Node appeared first on Appuals.

Samsung Had to Reconsider the Headset Hardware After the Release of Apple’s Vision Pro

Samsung has been planning on launching its own version of an XR headset for quite some time now, but the release of Apple’s Vision Pro inadvertently postponed its launch, since the original hardware specifications were considered inferior to the Vision Pro.

The design and hardware have undergone significant changes, and now appear to be well-suited for the target market. Reports indicate that the headset will be priced somewhere between $500 (Meta Quest 3) and $3,500 (Apple Vision Pro), potentially settling around $2,000, which would directly compete with Meta’s Quest Pro series.

Due to the hardware re-evaluation, Samsung not only increased the raw processing power but also addressed the visual aspects of the headset. Initially, it offered a lower-resolution 2,000 PPI single-eye display compared to Apple’s 3,386 PPI display, and it also lacked behind in the chipset department, using the older, less powerful Exynos 2200 compared to Apple’s dual-chip M2 and R1 combination.

The revised specifications include Qualcomm’s recently released Snapdragon XR2+ Gen 2 chipset, paired with displays offering a 90Hz refresh rate at 4.3K resolution and 120Hz at 3.7K resolution. This chipset is the direct successor to the XR2 Gen 2 found in the Meta Quest 3. Since we don’t have complete performance numbers at this point, we can’t directly compare the XR2+ Gen 2 with the M2.

Samsung’s XR headset has now entered mass production and targets a release around the last quarter of this year. The company forecasts selling around 300,000 units in 2024. The release of the Vision Pro has undoubtedly raised a lot of eyebrows, with multiple companies now vying to push for a place within this rapidly emerging green field market.

This is all we know for now, but rest assured that we will keep you updated as new information becomes available.

The post Samsung’s $2,000 Competitor to Vision Pro Enters Mass Production, Targets Late 2024 Launch appeared first on Appuals.

Intel Lunar Lake to Use TSMC’s N3B Process

A small press session was held after the IFS Direct event, where CEO of Intel, Pat Gelsinger confirmed Intel’s partnership with TSMC to produce Lunar Lake. For the unaware, Lunar Lake is the successor to Meteor Lake and targets the ever-growing 15W laptop segment.

Gelsinger also confirmed the expansion of orders to TSMC, confirming that TSMC will hold orders for Intel’s Arrow and Lunar Lake CPU, GPU, and NPU chips this year, and will produce them using the N3B process, officially ushering in the Intel notebook platform that the outside world has been waiting for for many years. CPU orders.

Pat Gelsinger (Translated)

If we take Pat’s (translated) wording literally, it would imply that the entire Lunar Lake package is built on TSMC’s N3B node. However, Intel’s own slides indicate that Lunar Lake will be a mix of external nodes plus the bleeding-edge 18A node.

The modular design (Foveros) enables them to tackle this obstacle by contracting TSMC with almost all of Lunar Lake. Investors and consumers want reliable products out as fast as possible. If they wait a quarter or two should they choose to design everything using IFS, they’ll have to play catch up once more.

It will be quite interesting to see 18A in action since Lunar Lake is said to arrive by late 2024. As 18A is a 1.8nm equivalent process, odds would be in Intel’s favor as they’d have the most advanced node, if Lunar Lake releases on time.

Lunar Lake: Everything We Know So Far

Lunar Lake succeeds Meteor Lake by offering a substantial architectural uplifts. On the CPU side of things, we have up to 8 Lion Cove and Skymont cores, with no support for hyperthreading. The iGPU has been upgraded from Xe-LPG to Xe2-LPG (8 Xe Cores), also known as Battlemage.

Memory-wise, support for LPDDR5x-8533 memory is rumored. The ‘MX‘ in Lunar Lake-MX indicates packaged memory, that is, the RAM will be directly soldered on to the CPU package, with capacities starting from 16GB.

• Up to 4 Lion-Cove P Cores and 4 Skymont E Cores with no SMT
• 8 Xe2 (Battlemage) iGPU Cores
• Support for On Package LPDDR5x-8533 Memory
• CPU+GFX Tile Built Using TSMC’s N3B Node
• SoC Tile Built Using Intel 18A
• Next-Gen NPU 4.0
• Support for WiFi 7 and Bluetooth 5.4
• 40% Efficiency Uplift Touted
• 2.5 TFLOPS (12W) | 3.8 TFLOPS (Peak Performance)

Lunar Lake’s design is significantly different from Meteor Lake, in a sense that it only features 2 Tiles; CPU+GFX, SoC. This CPU+GFX Tile will verly likely manufactured be using TSMC N3B, as indicated by documentation. Put two and two together, the SoC tile is then left to be fabricated using Intel 18A. Read more about Lunar Lake here.

Source: ChinaTimes

The post Pat Gelsinger Confirms Lunar Lake Will Use TSMC’s N3B Process Node appeared first on Appuals.

Meet the NVIDIA App: A Merger of Control Panel and GeForce Experience With a Modern Interface

NVIDIA App is designed for both desktop and mobile, featuring all the bells and whistles a modern and attractive UI. This app integrates the NVIDIA Control Panel, GeForce Experience, and RTX Apps into a unified package.

While this is a nice and much-anticipated update, we’re surprised it took NVIDIA this long. It is unanimously agreed upon that NVIDIA’s Control Panel looks like a relic from the Windows Vista era.

Besides the Control Panel, you’d need to install a separate app called GeForce Experience which is (was) NVIDIA’s answer to AMD’s Adrenalin. Firstly, you have the stark UI difference between both. Secondly, GeForce Experience required users to create an account, which did not bode well with some members of the community.

Now, NVIDIA users can manage all options, games, and settings within a single software. From the initial images, the interface strikes a resemblance to GeForce Experience with 5 unique tabs; Home, Drivers, Graphics, Redeem and Settings.

NVIDIA App no longer requires users to create an NVIDIA account and provides options top install other NVIDIA apps such as Omniverse.

In-Game Overlay Updates: Enhanced and Simplified UI & RTX HDR

Not only that, but NVIDIA’s in-game overlay has also seen a visual overhaul. While many of the options remain the same, the interface has been made simpler and easier to navigate. You have your typical options like Recording, Instant Replay, Photo Mode, Highlights, Game Filters etcetera.

In the same section, we have in-game statistics which allow users to monitor the FPS, GPU Clocks/Voltages, Latency and whatnot.

RTX HDR is also a key element of the in-game overlay software stack, which actually acts as a filter. It is recommended that you have an HDR-supported monitor to fully utilize this feature. From what we can see, the uplift in visual fidelity difference is quite huge.

Availability

The NVIDIA App is currently in Beta and is available for download here. NVIDIA’s latest 551.61 driver is optimized for this tool, so you can give that a look as well. Being in Beta, the software is still experimental and is missing quite a lot of features. Are you willing to give the NVIDIA App a try? Tell us in the comments.

Source: NVIDIA, Videocardz

The post The New ‘NVIDIA App’ Merges GeForce Experience, NVIDIA Control Panel and RTX HDR Into a Single Package appeared first on Appuals.

Clearwater Forest CPUs are now taped in, set to utilize the CPU Tiles manufactured using Intel 18A (1.8nm). Intel’s fearless ambitions of snatching TSMC’s crown are starting to transpire as Microsoft also announced new chips that use the Intel 18A process. By 2030, IFS, now rebranded to Intel Foundry aims to become the 2nd largest foundry in the world.

Intel 14A Announced: 5 Nodes in 4 Years Now Becoming a Reality

We finally have a successor to Intel 18A. As a 1.4nm class node, Intel 14A will be the industry’s first process node to use high-NA EUV technology. For some context, ASML recently supplied such lithography machines to Intel. TSMC, on the contrary, will not use these measures until 2030.

We neither have performance estimates nor density metrics for Intel 14A. What we do know from Intel 18A and 20A is that it should build upon GAA, RibbonFET, and PowerVia.

Amongst the glim, we also see some other interesting mentions such as Intel 14A-E and whatnot. Intel will expand its portfolio by offering enhanced versions of its process nodes. The P suffix indicates a performance improvement, with T indicating TSVs enabling 3D Stacking and E for Feature Extension.

Clearwater Forest Taped-In: Intel 18A CPU Tile, Intel 3 Base Tile and 3D Stacking

Clearwater Forest, a next-gen Xeon E-Core based processor lineup is now taped-in with design staring in Q2 2024. These CPUs use the Intel 18A fabrication process for the CPU Tiles, stacked atop an Intel 3-built Base Die. 3D stacking is said to be a crucial element, possibly allowing HBM4 memory, enabled by Foveros Direct.

UCIe is also a new and exciting technology launching with Clearwater Forest, which is an open standard for die-to-die interconnect. These CPUs are the first high-volume offerings to make use of this interface.

IFS Sees a Rebrand: Intel Foundry Develops New Partnerships With the Likes of Microsoft

IFS or Intel Foundry Services is being renamed to Intel Foundry, simplifying the name. However, Intel Foundry now positions itself as a ‘systems foundry‘ which as per Intel packs system-level design, technology development, supply chain capabilities, packaging, connectivity fabric solutions and even cooling technology under one umbrella.

ASAT or Advanced System and Test also sees an upgrade with the inclusion of FCBGA 2D+. ASAT integrates Intel’s technologies on custom-built chips. Similarly, we also have a new Intel 12 node in the pipeline that stems from Intel’s partnership with UMC. Following the same trend, Intel will also produce 65nm-based chips for Tower Semiconductors, since not all hardware needs cutting-edge technology.

Adding to the already growing number of affiliated partners, Microsoft has also announced a new chip design using Intel’s 18A technology. Microsoft Chairman and CEO Satya Nadella cites the need of ‘a reliable supply of the most advanced, high-performance, and high-quality semiconductors‘ as to why they picked Intel 18A.

In total, across wafer and advanced packaging, Intel Foundry’s expected lifetime deal value is greater than $15 Billion.

Source: Intel, Tom’s Hardware

The post Intel Announces the 14A (1.4nm) Process Node, Next-Gen 18A Based ‘Clearwater Forest’ CPUs Now Taped-In, and IFS Sees a Rebrand to ‘Intel Foundry’ appeared first on Appuals.

Moreover, the user has also provided us with two distinct images of the connectors so we can easily differentiate one from the other.

Your GPU Could Still Use the ‘Old’ 12VHPWR Power Connector

Identifying both connectors is actually not that difficult. The new revision is demarcated with an ‘H++‘ tag, whereas the old 12VHPWR resorts to ‘H+‘. The former (newer model) can output a max of 675W of power, with the latter going as high as 600W.

The second part of this report suggests that a few RTX 40 Super series models still use the old, 12VHPWR (H+) power connector. While the AIBs remained unmentioned, this can lead to potential issues since the old connector was notorious for being extremely sensitive. We saw boatloads of dead RTX 4090s, with a majority of them being linked to the faulty design of the 12VHPWR connector.

One may blindly buy a new RTX 40 Super GPU, thinking this issue must’ve been alleviated by now. This also puts NVIDIA in the spotlight since custom GPUs from AIBs should need NVIDIA’s approval before launch.

The 12V-2×6 connector supersedes 12VHPWR by ironing out many of its predecessor’s faults. The pins have now been cut down and made more conductive. Hardware Busters, after analyzing the new design concluded that the 12V-2×6 connector maintains a stable temperature, even when improperly inserted, something which was previously termed as ‘User Error‘.

The silver lining is that these GPUs have much lower TGPs than the RTX 4090, a key affectee of the melting catastrophe. If the temperature is kept in check, the chances of your GPU melting should be negligible.

However, the point remains. If your (an RTX 40 Super user) anxiety levels are off charts, try taking the cooler off to check the whether the connector is marked as H+ or H++ (This may void your warranty).

Source: Wxnod

The post Some RTX 40 SUPER GPUs Still Use the Controversial 12VHPWR Connector, Despite A Newer Revision Already in the Market appeared first on Appuals.

RX 7900 GRE Now Available for the UK Market

Despite being exclusive to the Chinese market, the RX 7900 GRE is found in many pre-built systems around Europe. By October of last year, this requirement was lifted with various models available for DIY use. Recent listings show that a few AIB models for the RX 7900 GRE will soon hit shelves around the UK, with a similar DIY launch planned for the US.

The first available model is the XFX Radeon RX 7900 GRE, based on AMD’s reference specifications. The card is priced at £659.99 (~$832), including the VAT. Since supply was extremely limited, these GPUs are now out of stock and as such the listing has been removed.

Now let’s get to the elephant in the room, and that’s the pricing. At £659.99, the RX 7900 GRE does not make much sense since you can snag the faster RX 7900 XT from XFX at £40 more. On that note, the EU also had significantly better deals with Reference spec models avaialble at €542.66 (~$585), bundled with a copy of AVATAR: Frontiers of Pandora.

That’s not all as Videocardz reports that US customers should see a few models from Gigabyte by as soon as later this month. It is worth mentioning that we’re talking about the DIY market here, because the RX 7900 GRE was already available through system integrators at launch.

Source: TechPowerUp, Videocardz, AWD-IT

The post AMD Radeon RX 7900 GRE Shows Up in the UK, Also Planned to Launch in the US DIY Market Later this Month appeared first on Appuals.

Lunar Lake 8-Core Sample in Task Manager

The CPU appears to be an early Engineering Sample, something that is given by its name. For more context, all Lunar Lake processors feature 8 cores as per a previous leak. Hence, it is kind of difficult to pinpoint the model and Core Ultra series.

Speaking of the cores, Lunar Lake packs 8 cores, divided across 4 performance and 4 efficient cores. This specific sample is running at 2.78GHz. Over to the left, we can also see the Intel AI Boost NPU, which Intel touts as a 3x AI bump over Meteor Lake.

A key takeaway from this screenshot is the fact that Lunar Lake abandons hyperthreading. The 4 performance cores only feature 4 threads, making it clear as daylight that Lion Cove does not support hyperthreading. The implications of this decision will only surface once these processors hit shelves.

Furthermore, if we go into the cache side of things, Lion Cove is expected to feature 3MB of L2 Cache. This leaves behind just 2MB for the E-core cluster, which is less than even Raptor Lake (4MB per each E-Core cluster).

Bionic_Squash says that Lunar Lake packs 2.5MB L2 cache per P-cores, leaving behind 4MB for the Skymont-based E-cores. The L3 slice size remains constant at 3MB, however, Skymont E-cores on Lunar Lake will feature no L3 cache.

Lunar Lake-MX will be powered by blazing fast LPDDR5x-8533 memory with support for WiFi 7 and Bluetooth 5.4.

Lunar Lake will feature Lion Cove P-Cores and Skymont E-Cores. The GPU side of things sees a huge upgrade with Battlemage, Xe2-based graphics. As per official leaked documentation, Lunar Lake is designed using the TSMC N3B node (CPU Tile). The same leak claims that Meteor Lake at peak potential can outclass the Apple M2′s iGPU at 3.8 TFLOPS. Read our article for a more detailed overview.

More accurate figures and specs should pop up as we approach the launch date.

Source: zhihu

The post Intel Lunar Lake 8-Core ‘A1’ Sample in Task Manager Confirms No Hyperthreading Support appeared first on Appuals.