Category: Computers

  • Griffin Cove and Nova Lake IPC: Intel to Replace E-Cores with P-Cores

    Griffin Cove and Nova Lake IPC: Intel to Replace E-Cores with P-Cores

    Intel’s CPU plans appear to be ever-changing, based on the latest leaks. So far, it’s been revealed that Intel has scrapped Beast Lake, Beast Lake Next, and has also disbanded Jim Keller’s Royal Core initiative. Moreover, there are reports suggesting that Team Blue could launch Panther Lake desktop chips to compete against AMD’s Zen 6 by 2025.

    Future Innovations

    In a recent update regarding IPC metrics for Nova Lake, Moore’s Law Is Dead indicates that Intel might be developing a groundbreaking CPU architecture set to debut in 2027 or later. The source mentions that Intel is working on Griffin Cove P-cores for a yet-to-be-named architecture expected to arrive in 2027 or beyond. These P-cores might utilize the Intel 14A-P, which is an upgrade of the Intel 14A technology anticipated to commence production in 2026.

    Design Changes Ahead

    Interestingly, MLID suggests that Intel could adopt a "Unified Core" approach for E-cores, mirroring AMD’s strategy. The upcoming CPUs based on Griffin Cove or Griffin-Next might eliminate the distinction between P and E-cores, taking inspiration from Zen 5 and Zen 5c cores. Unlike Intel’s separate P and E-cores, the Zen 5 series features similar architecture with minor changes in the Zen 5c for size reduction. Intel seems to be planning to implement a similar design for its 2027 architecture, moving away from the two distinct core types within a single CPU.

    Performance Expectations

    On the performance front, MLID speculates that the Griffin Cove P-cores could see a 10-20% IPC increase compared to Nova Lake and its Coyote Cove P-cores. However, since Griffin Cove is still quite a ways off, any talk of “IPC enhancement” should be approached with skepticism. MLID is aware of this uncertainty and cautions in their video that “anything after Nova Lake is NOT finalized”.

    In addition, MLID claims a potential IPC boost of 9 to 18% for the Coyote Cove P-cores in Nova Lake, although no numbers are provided for the Arctic Wolf E-cores. The leaker further suggests that Nova Lake might implement the refreshed Intel 18A-P process instead of the previously rumored 14A and TSMC 2 nm.

    As we move forward, it remains to be seen what Intel has planned with Nova Lake. In the meantime, we can look forward to the reviews of the Arrow Lake Core Ultra 200 CPUs, which should roll out in the coming years. The end of 2024 is shaping up to be an exciting period for gamers, especially since AMD has announced that the Ryzen 9000X3D will launch next month.

    Moore’s Law Is Dead on YouTube, Teaser image: Intel, Moore’s Law Is Dead, edited.

    Image 1
    Tags: , ,

  • Intel Core 270H Leak: 14-Core Mobile CPU Hits 5.8 GHz

    A recent leak has shown Intel’s upcoming mobile processors, giving us insight into seven new chips from the Core 200H and Core 200U series. A popular hardware leaker, Jaykihn, shared these details on X, highlighting the Raptor Lake refresh products.

    Core 270H: The Performance Beast

    At the forefront is the Core 270H, engineered for top-tier performance. This processor features 14 cores, with six designed for high performance and the other eight dedicated to efficiency. It kicks off at a base clock of 2.1 GHz but can elevate to an impressive 5.8 GHz when pushed to its limits.

    Diverse Options in the 200H Lineup

    The Core 200H series includes five distinct models catering to various performance levels. The Core 250H maintains the 14-core configuration of the 270H but scales back slightly, reaching a boost of 5.4 GHz. The Core 240H, on the other hand, provides ten cores with a maximum boost of 5.2 GHz, while the Core 220H has 12 cores and peaks at 4.9 GHz. For those looking for budget options in the H-series, the Core 210H offers eight cores and can achieve a boost clock of 4.8 GHz.

    U-Series Chips for Lightweight Devices

    Intel is also introducing two U-series processors designed for thin laptops and entry-level notebooks. The Core 250U and Core 220U each come equipped with ten cores, split into two performance and eight efficiency cores. The 250U operates at a 1.1 GHz base frequency and can ramp up to 5.4 GHz when needed, whereas the 220U starts at 0.8 GHz and reaches a maximum of 5.0 GHz. Both of these U-series processors include 12MB of L3 cache and maintain a thermal design power (TDP) of 15W for better cooling.

    Not Competing with Core Ultra Lineup

    These new 200 series chips won’t interfere with Intel’s Core Ultra lineup. Instead, they utilize the Raptor Lake architecture, while the Core Ultra processors, such as the 200V and 200H/HS, are built on the newer Lunar Lake or Arrow Lake architectures. Anticipate the launch of these new 200-series processors sometime next year, likely around the same period as the Core Ultra 200H/HX Arrow Lake processors.

    Techspot (translated in English) via @jaykihn0 (translated in English).

    Tags: , ,
  • New Leaks Reveal Microsoft Surface Laptop 8 and Surface Pro 12 Details

    New Leaks Reveal Microsoft Surface Laptop 8 and Surface Pro 12 Details

    Microsoft has not frequently updated its Surface lineup in recent years, especially after the failures of the Surface Duo and Surface Neo. Nonetheless, this year it has launched four new Surface products: the Surface Laptop 6 for Business, Surface Laptop 7, Surface Pro 10 for Business, and Surface Pro 11, with the latter available for $1,099.99 on Amazon.

    New Business Variants

    Although there have been ‘for Business’ models before, this time is significant because they have different processor designs compared to the consumer versions. New reports suggest that Microsoft might be going back to refreshing its Surface Pro and Surface Laptop series annually, particularly the ‘for Business’ editions.

    Upcoming Devices

    Recently, images of what could be the Surface Laptop 8 for Business surfaced online. A post on Chiphell indicates that Microsoft is working on this device under the codename ‘831_WID’. Additionally, a new Surface Pro, codenamed ‘830_QUN’, is also in the works and is expected to launch as the Surface Pro 12 for Business.

    Technical Specifications

    Both the Surface Laptop 8 for Business and Surface Pro 12 for Business will likely be equipped with Core Ultra 5 226V, Core Ultra 7 266V, and Core Ultra 7 268V processors. This means that they will come in both 16 GB and 32 GB RAM options. Moreover, both new models are anticipated to showcase the design tweaks that Microsoft introduced earlier this year with their ARM-based consumer versions.

    Interestingly, this aligns with an earlier report that appeared on Reddit recently. Although that post has since been removed, it was claimed to be from a reliable Microsoft partner over the years. So, it wouldn’t be shocking if Microsoft reveals information about a new Copilot+ or a business-focused event around March or April of 2025.

    Image 1
    Tags: , ,
  • Loongson 3B6600M: Chinese CPU Competes with 7nm x86 Processors

    Loongson 3B6600M: Chinese CPU Competes with 7nm x86 Processors

    Loongson Technology, a semiconductor company from China, has announced major advancements in its CPU development and future plans during the 2024 Loongson Industrial Ecosystem Conference. The chairman, Hu Weiwu, discussed the latest updates related to their three key product categories: high-performance CPUs, embedded SoCs, and MCU microcontrollers.

    Recent Developments in CPUs

    The latest desktop CPU offered by Loongson, the Loongson 3A6000, stands strong against 14nm x86 processors. Operating at 2.5 GHz, it matches the performance of Intel’s 3.6 GHz Core i3-10100, demonstrating Loongson’s focus on optimizing design for enhanced performance without relying on the latest technology.

    Future Plans and Innovations

    Looking to the future, the company is developing the 3C6000 server CPU, which will aim to rival 10nm x86 processors. Additionally, their forthcoming desktop CPU, the 3B6600M, is expected to compete with 7nm x86 processors, potentially lining up against Intel’s 12th and 13th-gen Core chips in the mid-to-high-end segments.

    Loongson is also experimenting with turbo frequency technology to achieve speeds of 3 GHz. Furthermore, they are incorporating x86 binary translation instructions, ensuring that Windows applications and systems will run smoothly on their CPUs.

    Graphics and More

    On the graphics front, Loongson is in the process of developing the 9A1000 GPGPU chip, with initial versions anticipated to be released in the first half of 2025. They aim for this chip to be comparable to AMD’s RX 550, capable of handling AI tasks and supporting technologies such as OpenGL 4.0, PCIe 4.0, HDMI 2.1, and DP 2.1.

    Loongson’s focus isn’t solely on CPUs; they have made significant strides in their embedded SoCs and MCU microcontrollers, now in their fourth generation. These chips are meeting or exceeding market expectations in terms of performance and value.

    Their proprietary LoongArch architecture, introduced four years ago, has passed intellectual property assessments and weathered independent lawsuits. It enjoys full support from operating system kernels, BIOS, and six prominent compilers, which include GCC, LLVM, GOLANG, Java, JavaScript, and .NET.

    Image 1
    Tags: , ,
  • AMD Announces Ryzen 7 9800X3D Launch Date in Early November

    AMD Announces Ryzen 7 9800X3D Launch Date in Early November

    After a lot of leaks, guesses, and chatter, the highly awaited AMD Ryzen 7 9800X3D is set to launch. You can grab it from selected sellers starting November 7. The price will be announced later, but rumors hint it might be between $500 and $550. Moreover, price reductions for other Zen 5 models will start on October 20.

    Performance Insights

    Recent leaked benchmarks indicate that the Ryzen 7 9800X3D can reach a maximum boost of 5.6 GHz when using Precision Boost Overdrive (PBO). This processor comes with eight cores and 16 threads, featuring a base clock speed of 4.7 GHz, a thermal design power (TDP) of 120 Watts, and a total of 96 MB of L3 cache.

    Similarities to Previous Models

    When it comes to specs, the Ryzen 7 9800X3D is quite similar to the Ryzen 7 7800X3D. The performance improvements will primarily stem from the new Zen 5 architecture and the higher base and boost clock speeds. The added support for overclocking will also give users a chance to squeeze out even more power from this chip.

    AMD

    Tags: , ,
  • AMD Ryzen 7 9800X3D Clock Speed, V-Cache, TDP Details Leaked

    AMD Ryzen 7 9800X3D Clock Speed, V-Cache, TDP Details Leaked

    A recent Cinebench 2024 test indicated that the AMD Ryzen 7 9800X3D may increase its base clock speed to 4.7 GHz, compared to the 4.2 GHz on the Ryzen 7 7800X3D. However, its performance in gaming might not be as stellar in certain games. A document from a motherboard manufacturer was discovered by Videocardz, which confirmed this clock speed and provided additional insights about AMD’s new CPU.

    Specifications Unveiled

    The AMD Ryzen 7 9800X3D retains the same 96 MB of L3 cache as the previous two Ryzen 7 X3D models. It also features a thermal design power (TDP) of 120 Watts. While other specifications remain unclear, it is anticipated that it will maintain the 8-core configuration, with one chiplet having complete access to the additional L3 cache.

    Anticipated Launch Date

    Although AMD has not officially announced it yet, the Ryzen 7 9800X3D is projected to be released on October 24, with sales expected to start a week or two later. The other two Zen 5 X3D models, the Ryzen 9 9900X3D and Ryzen 9 9950X3D, are anticipated to debut next year, and if the rumors hold true, they might feature enhanced 3D V-cache.

    Image 1
    Tags: , ,
  • TSMC Denies Intel Fab Acquisition Rumors

    TSMC Denies Intel Fab Acquisition Rumors

    TSMC, the biggest contract chipmaker in the world, has clearly shut down any discussions about acquiring Intel’s manufacturing facilities. This comes at a time when Intel’s foundry division is having a hard time attracting clients, raising questions about CEO Pat Gelsinger’s IDM 2.0 strategy and its potential success.

    TSMC’s Firm Stance

    During a recent earnings call, TSMC’s CEO, C.C. Wei, was very straightforward, stating, “Are we interested to acquire one of IDM’s fabs? The answer is no, OK? No, not at all.” This direct response effectively ends the rumors about who might take over Intel’s fabs if they choose to sell their manufacturing operations.

    Reasons Behind the Speculation

    The talk surrounding Intel’s fabs stems from various considerations. To begin with, there’s the financial aspect. Intel’s extensive manufacturing infrastructure would require a substantial amount of money to take control of, and even TSMC, despite its vast resources, would think long and hard before proceeding. Additionally, there’s the technical challenge. Intel’s fabs are specifically designed for their own products and processes, making it a difficult and costly task to adapt them for use by other companies.

    Different Business Models

    The contrasting operational styles of the two companies further contribute to TSMC’s lack of interest. TSMC maintains a streamlined corporate structure, which is quite different from Intel’s approach of tackling large projects and engaging in deep fundamental research. Furthermore, any significant acquisition like this would likely face considerable pushback from antitrust regulators and could potentially escalate geopolitical tensions.

    A Fundamental Clash

    Perhaps the most significant reason for TSMC’s disinterest lies in their differing business philosophies. TSMC focuses solely on being a pure-play foundry, while Intel adopts an IDM model, managing both design and manufacturing internally. These diverging strategies simply do not align, making the possibility of TSMC acquiring Intel’s fabs virtually impossible.

    Tags: , ,
  • Intel Arrow Lake H Series Leaks: 5 Models, 45W TDP Variant

    Intel Arrow Lake H Series Leaks: 5 Models, 45W TDP Variant

    Intel’s Arrow Lake H series has had its details exposed online. This information comes from the reliable leaker Jaykihn, who also revealed details about Arrow Lake-HX and Lunar Lake earlier this year. The lineup consists of five SKUs, featuring one Core Ultra 9 model at the top. We get a glimpse into its CPU and integrated GPU performance.

    Core Configurations

    It was previously speculated that Arrow Lake-H would include three types of CPU cores. The P-cores are identified as Lion Cove, while the E-cores are Skymont, and the LP-E cores are Crestmont, which are derived from Meteor Lake. It’s still unclear why Intel chose to reintroduce LP-E cores, particularly given their previous underwhelming performance.

    Thermal Design Power

    The only model with a higher performance, the Core Ultra 9 285H, has a TDP of 45 Watts. The other models are limited to 28 Watts, which aligns closely with prior H SKUs. The entire range appears capable of boosting up to 65 Watts (PL2), based on the available information.

    Integrated Graphics

    Furthermore, the Core Ultra 7 265H, Core Ultra 7255H, and Core Ultra 9 285H all feature an integrated GPU with 8 Xe cores. In contrast, the Core Ultra 5 235H and Core Ultra 5 225H come equipped with 7 Xe cores. Each model’s maximum boost clock speeds decrease progressively.

    Jaykihn shared this information on X.

    Tags: , ,
  • M4 MacBook Pro May Offer Up to 28GB Unified Memory

    M4 MacBook Pro May Offer Up to 28GB Unified Memory

    The internet is buzzing with rumors about Apple’s next M4-powered MacBook Pro models. While the typical insights from market experts are expected, the recent unboxing videos of the M4 MacBook Pro on YouTube have raised some serious concerns about the supply chain.

    Early Look at Upgrades

    These unboxing clips offered a sneak peek at various enhancements that the refreshed base-model 14-inch MacBook Pro will feature, such as double the base memory and an extra Thunderbolt port for connecting additional displays. Additionally, a user on Reddit shared a screenshot that seems to show the purchase page for the upcoming M4 MacBook Pro, potentially revealing more details about the laptop.

    Memory and Storage Specs

    From the screenshot, it appears that the M4 MacBook Pro will kick off with 16 GB of unified memory for the base model, along with 512 GB of storage. Interestingly, higher-tier versions of this base model come equipped with 1 TB of storage and 28 GB of memory, which seems like a strange configuration.

    In contrast, the M3 Pro-equipped MacBook Pro was launched with 18 GB of memory through a triple-channel setup. The 28 GB memory option raises eyebrows, and it will be intriguing to find out how Apple reached that figure. However, there’s a significant concern regarding this rumor—the unboxing video suggested that the M4 MacBook Pro would be available in a Space Black color, which does not appear in the Reddit screenshot.

    Confusion and Speculation

    Honestly, it has become very difficult to discern what’s true and what’s not at this point. The sight of a retail version of an unreleased MacBook Pro being unboxed and benchmarked nearly a month before its actual launch is something nobody anticipated. Yet, here we are again, caught up in the whirlwind of speculation.

    Of course, there’s a chance that the screenshot is merely a result of someone using the ‘Inspect Element’ feature on Apple’s website. Yet, this does not clarify the evident pre-release aspect of the webpage. Furthermore, if that were the case, it would be reasonable to assume that the individual would have opted for a more standard memory configuration like 32 GB, right?

    Awaiting Confirmation

    Overall, it is tough to verify the authenticity of the screenshot at this time. We’re less than two weeks from the projected launch of the M4 MacBook Pro, which means we won’t have to wait long to discover what Apple is preparing for us.

    Reddit.

    Image 1
    Tags: ,
  • ASUS ROG Thor III PSUs: Premium Power with OLED Display

    ASUS ROG Thor III PSUs: Premium Power with OLED Display

    ASUS has unveiled its newest high-end power supply unit, the ROG Thor III series, which boasts impressive features and advanced technology. This range includes models starting from 1,000W and going up to a powerful 1,600W, making it ideal for enthusiasts who want top-tier setups.

    Key Features

    A standout aspect of these PSUs is the magnetic OLED display. Unlike older versions that are fixed in one location, this display can be positioned wherever it’s most convenient, making it adaptable to different case designs. It also provides real-time information on power consumption.

    ASUS has embraced GaN technology throughout the entire series, potentially improving efficiency by up to 30% compared to standard MOSFETs. This innovation means the units operate cooler and more effectively while adhering to the ATX form factor.

    Enhanced Stability

    The lineup also features GPU-First voltage sensing along with a smart voltage stabilizer. According to ASUS, this setup delivers 45% more stable power to the graphics card, enhancing overall performance.

    Turbo mode activates to handle prolonged peak loads, thanks to a carefully calibrated fan curve. The fans are dual-ball bearing types, designed for a lifespan of up to 80,000 hours, ensuring significant reliability over time.

    Power Options and Components

    The ROG Thor III series offers three power alternatives:

    Every model includes low-ESR capacitors, ROG-branded heatsinks, all-aluminum exteriors, and modular cables adorned with ROG style.

    While ASUS hasn’t disclosed the pricing yet, the previous ROG Thor 1600T was priced at $699, so you can anticipate these new options to be similarly priced.

    On a positive note, ASUS supports these power supplies with a robust 10-year warranty, providing customers with confidence regarding long-term performance and durability.

    Tags: , ,