Tag: NVIDIA

  • Graphene Semiconductors: Closer to Reality | CheckMag

    Graphene Semiconductors: Closer to Reality | CheckMag

    For many years, silicon has ruled the semiconductor world, supporting everything from computers to cellphones. But now, silicon is reaching its limits in both performance and scalability. This is where graphene comes in, a material that’s been dubbed a miracle of material science. With its unmatched conductivity and electron mobility, graphene could enable processors to work at terahertz speeds, which is much faster than the gigahertz limits of today’s silicon chips.

    What is Graphene?

    Graphene consists of a single layer of carbon atoms that are arranged in a hexagonal pattern, giving it amazing strength and electrical characteristics. In contrast to silicon, graphene allows electrons to move with much less resistance, leading to quicker and more efficient processing. However, there is a problem: graphene does not have a band gap. This band gap is essential for semiconductors to switch between on and off states. Without it, graphene cannot act like a traditional transistor—but this has changed now.

    Breakthrough in Research

    Scientists have finally figured it out. By attaching graphene to silicon carbide and “doping” it with atoms that donate electrons, they have successfully developed a working graphene-based semiconductor. This method, called epitaxial graphene fabrication, creates a band gap while still preserving the unique qualities of graphene.

    The outcome? Transistors that are not just ten times faster than silicon ones, but also mostly compatible with current manufacturing methods. This compatibility allows for a seamless shift from silicon to graphene chips, which is a crucial element for making this technology commercially viable.

    Industry Perspectives

    Graphene’s capabilities go beyond just high-speed electron flow. Even those at the top of the industry recognize that silicon’s dominance is fading. Nvidia’s CEO Jensen Huang famously said, “Moore’s Law is dead.” For those who may not know, Moore’s Law suggested that the number of transistors on a chip would double every two years, leading to massive increases in computing power. However, as transistors become smaller, problems like heat production and switching speeds have notably slowed advancements.

    Thanks to its excellent performance and potential for scalability, graphene could be a solution to these challenges. It may even continue or replace the path that Moore’s Law once paved.

    Challenges Ahead

    As with any emerging technology, some issues need to be resolved before graphene can take the lead. Expanding production and incorporating graphene semiconductors into consumer electronics will call for a substantial investment and a dedication to further innovation. Additionally, there’s the question of whether graphene can surpass other upcoming superconducting technologies in the competitive field of quantum computing.

    Nevertheless, the outlook is bright. With its compatibility with current manufacturing processes and ongoing research into quantum applications, graphene semiconductors are more than just a distant possibility—they represent a glimpse into the future of technology.

    While graphene-based semiconductors might not immediately resolve all of silicon’s challenges, they signify an important advance. Whether it’s enhancing the performance of your next laptop or realizing the full potential of quantum computing, this breakthrough could change the landscape of technology.

    The future is rapidly approaching, and it might just be driven by graphene.

    Source:
    Link

  • Nvidia GeForce RTX 5070 Ti Release Date Leaked Information

    Nvidia GeForce RTX 5070 Ti Release Date Leaked Information

    VideoCardz has released new information about the launch of the GeForce RTX 5070 Ti. Just under a month ago, Nvidia unveiled this graphics card, which is priced at $749, along with the GeForce RTX 5080, GeForce RTX 5090, and the more basic GeForce RTX 5070.

    Performance Insights

    The GeForce RTX 5070 Ti is reported to provide performance that is nearly double that of the previous GeForce RTX 4070 Ti, which is currently priced at $879.99 on Amazon. However, it comes with a slight increase in total graphics power (TGP), rated at 300 W. To boost performance, Nvidia has utilized the GB203 GPU manufactured using TSMC’s 4 nm technology, packing in 8,970 CUDA cores and achieving a maximum boost clock of 2.45 GHz. Moreover, the GeForce RTX 5070 Ti is equipped with 16 GB of GDDR7 VRAM, which enables a memory bandwidth of 896 GB/s through a 256-bit bus.

    Launch Schedule

    According to previous leaks, VideoCardz claims that the GeForce RTX 5070 Ti will be available for purchase starting on February 20. Additionally, they mention that this date will also mark the release of reviews for non-MSRP models. In contrast, reviews for cards that stick to Nvidia’s suggested retail price of $749 will be published a day earlier, on February 19.

    Future Releases

    As per VideoCardz, Nvidia has yet to confirm the launch dates for the GeForce RTX 5070 to its board partners. For context, the GeForce RTX 5070 is currently the weakest option in Nvidia’s Blackwell lineup. The company is anticipated to launch the GeForce RTX 5060 and GeForce RTX 5060 Ti later this year, which will feature variations of the GB206 GPU.

    Source:
    Link

  • Nvidia GeForce RTX 5060 Ti Launching in Two Versions

    Nvidia GeForce RTX 5060 Ti Launching in Two Versions

    Even though Nvidia didn’t fully reveal the GeForce RTX 50 series at CES 2025, a previous leak provided a general view of what the lower tier might be like. However, X leaker @Olrak29_ has found some surprising news regarding one of Nvidia’s entry-level models: the GeForce RTX 5060 Ti.

    Specifications Unveiled

    On the EEC (Eurasian Economic Commission) website, there seem to be two versions of the graphics card – one featuring 8 GB of VRAM and the other boasting 16 GB. In contrast, the standard GeForce RTX 5060 will only come in a single model with 8 GB of video memory. Essentially, Nvidia aims to provide a more affordable RTX 5060 Ti option for those who don’t require the additional VRAM.

    Anticipated Launch Details

    At this point, not much is known about the GeForce RTX 5060 Ti. It’s rumored to be equipped with a GB206 GPU on a PG152 board. Nvidia is likely to announce it in March, coinciding with the launch of its non-Ti version, presumably around the time AMD reveals the Radeon RX 9070 XT and RX 9070. However, the RTX 5060 and RTX 5060 Ti won’t be direct rivals, as they are expected to compete against AMD’s Navi 44-based Radeon RX 9060 XT (tentative) and others.

    Performance Insights

    Recently, we had a look at how the GeForce RTX 5080 compares to its previous-gen model. There isn’t much to be excited about in terms of raw performance, and this trend will probably extend to the RTX 5060 Ti and RTX 5060. Gamers will once again need to depend on AI-driven enhancements to achieve any substantial performance improvement over Ada Lovelace.

    Source:
    Link


  • Nvidia Ends Driver Support for GTX 1080 Ti and Older GPUs

    Nvidia Ends Driver Support for GTX 1080 Ti and Older GPUs

    With the launch of CUDA 12.8, Nvidia is stopping any further development on driver features for the widely used GTX 1080 Ti and other GeForce GTX 1000 series cards. Although Nvidia’s latest CUDA drivers still provide official support for the Pascal architecture, which underpins the entire GTX 1000 series, there will be no additional enhancements or new features introduced moving forward.

    Legacy Status for Older Architectures

    The Maxwell and Volta architectures, which are the direct predecessors to the RTX 2000 series, have also been marked as “legacy.” Nvidia has categorized these architectures as “deprecated” in the release notes for CUDA 12.8. While the GTX 1000 series remains compatible with CUDA 12.8, this move clearly shows that Nvidia is shifting its attention to more recent generations of graphics cards.

    CUDA’s Importance in Computing

    CUDA serves as Nvidia’s platform and programming model for parallel computing. It empowers developers to harness the capabilities of graphics cards for general computing tasks, including machine learning, scientific calculations, and video editing. Since no new features will come for these older architectures, users of these GeForce GPUs might miss out on future enhancements or optimizations when using CUDA-based applications.

    Continued Popularity Among Gamers

    Even though they are quite old, graphics cards based on the Pascal architecture are still very much in demand among gamers, as highlighted by Steam’s hardware and software statistics. Several models from the GTX 1000 series are among the top 20, and collectively, these cards represent over 7% of the market. Unfortunately, this leaves owners of these GPUs facing an uncertain future regarding official driver support.

    Source:
    Link

  • RTX 5080 Time Spy Score: Beats RX 7900 XTX, Up 15% vs RTX 4080

    RTX 5080 Time Spy Score: Beats RX 7900 XTX, Up 15% vs RTX 4080

    After recent reviews of the RTX 5090 surfaced, attention has now shifted to the next GPU in the RTX 50 series, the RTX 5080. Set to launch on January 30, it comes with a starting price of $999. Early leaks regarding the RTX 5080’s performance indicate a modest improvement over the RTX 4080, but not a groundbreaking one.

    Performance Insights

    For example, previous reports indicated that the RTX 5080 was merely 10% and 3% faster than the RTX 4080 in Blender and OpenCL benchmarks, respectively.

    Recently, a new result from the 3DMark Time Spy test for the RTX 5080 has been shared online, revealing a more favorable performance compared to earlier leaks involving OpenCL and Blender.

    Leaked Benchmarks

    According to findings posted on Baidu (via Olrak29 on X), the RTX 5080 scored 32,701 in the Time Spy Graphics test. Our records show that the RTX 4080 and RTX 4080 Super average around 27,605 and 28,344 points in the same test. This means the leaked RTX 5080 is ahead by 18.5% and 15%, respectively, compared to the RTX 4080 and RTX 4080 Super.

    However, gamers hoping the RTX 5080 would surpass or equal the RTX 4090 might feel let down, as the RTX 5080’s score is still 10% lower than that of the RTX 4090 based on our evaluations.

    AMD’s Position

    Currently, AMD doesn’t have a direct rival for the RTX 5080 in its RDNA 4 series, but it may not need one if the 3DMark scores are any clue. The RX 7900 XTX, for instance, averages around 31,540 points in the Time Spy Graphics test—just 4% behind the anticipated RTX 5080 score.

    With the RX 7900 XTX available for under $950 on platforms like Amazon and Newegg, many gamers might find it hard to justify the RTX 5080’s price. This situation is further complicated by reports of limited initial supply for the RTX 5080, which could drive prices even higher than the $999 MSRP.

    Final Thoughts

    Nevertheless, the landscape could shift in favor of the RTX 5080 when considering features like DLSS 4 and Multi Frame Generation, which some may view as essential.

    Ultimately, only time will reveal the true performance of the RTX 5080 and its availability at launch. Until then, it’s wise to approach all leaks and speculations with caution.

    Source:
    Link


     

  • Nvidia Reveals Unique Features of Founders Edition RTX 5090

    Nvidia Reveals Unique Features of Founders Edition RTX 5090

    The Nvidia RTX 5090 FE stands out as a top-tier graphic card crafted to meet high power demands and manage significant heat output. Its surprisingly slim two-slot design breaks the mold for GPUs within this power range. As a premier offering, it showcases cutting-edge engineering and innovation, marking the 5090 FE as a significant advancement in Nvidia’s continuous evolution of graphics processing unit technology.

    Prototype Insights

    A recent video presented a prototype of this GPU, which featured an ambitious quad-slot design with a fully flow-through cooling solution and a vertical PCB. While this initial concept displayed the impressive capabilities of the hardware, the final version of the RTX 5090 FE was streamlined into a more compact design. This adjustment avoids the impracticality of the bulky 4-slot configuration, which wouldn’t fit well in numerous modern cases, opting instead for a design that better aligns with contemporary system builds and user needs.

    Thermal Innovations

    The RTX 5090 FE brings forth new ideas in thermal management. It employs liquid metal on the GPU die, held in place by a specially designed rubber gasket to prevent leaks that might harm the PCB. The GPU features a semi-pass-through setup that integrates a modular PCB with a two-thirds blow-through cooling system, ensuring temperatures remain optimal. Additionally, a 3D vapor chamber equipped with artery wicks significantly boosts heat dissipation efficiency.

    Usability Features

    This design also includes user-friendly elements like repositioned power connector ports for better accessibility and an anti-fingerprint coating, although its effectiveness has yet to be fully evaluated. These enhancements aim to improve maintenance and ease of use, but measuring their actual impact in real-world scenarios might be challenging.

    The RTX 5090 FE is a formidable graphics card that challenges the limits of technology and hardware design. With an impressive power consumption of 575W, it offers top-notch performance in a sleek two-slot layout, as verified through our testing. Such extreme power needs raise concerns regarding its practicality for everyday use and the heat it generates in smaller builds.

    Future of GPU Design

    The compact design of the 5090 FE reflects the trend toward minimalist PC cases, prompting the industry to reconsider the increasing size of GPUs. Where could this lead us? Ideally, towards a future with slimmer, more efficient GPU designs, steering us closer to a standard that is less than 3 slots.

  • RTX 5080 and GeForce 50 Series AIB Prices Explained

    RTX 5080 and GeForce 50 Series AIB Prices Explained

    History may soon repeat itself with Nvidia’s upcoming GeForce 50 series. The RTX 5080 and RTX 5090 are set to launch on January 30, while the RTX 5070 and RTX 5070 Ti are expected to arrive on February 20, according to leaks. Given the current state of affairs, it seems likely that we will witness stock shortages, price hikes, and scalping, leading many unhappy PC users to vent their frustrations on forums about Nvidia and its AIB partners. A recent post on the Chiphell forum, shared by @harukaze5719, suggests that this scenario is very much on the horizon.

    Pricing Strategies Under Scrutiny

    The post highlights that Nvidia is to blame for forcing custom board partners to resort to aggressive pricing measures to maintain profits. If you’ve ever wondered why Asus charges $800 above MSRP for its Astral RTX 5090 or even $900 more for the Astral RTX 5080, the reason could be that Nvidia has “left the manufacturers with a very, very low gross profit.” It’s suggested that creating a basic version according to MSRP is like doing charity work. Although it’s tough to feel bad for these companies since profit is profit, it seems Nvidia’s business approach is causing this predicament.

    Anticipated Stock Shortages

    Additionally, there are worries about stock shortages similar to those seen during the GeForce 30 series launch, which was significantly impacted by the cryptomining boom. Thus, even if Nvidia advertises the GeForce RTX 5080 FE at $999, the likelihood of purchasing one at that price appears very low. Gamers and PC builders will likely have to deal with the tricky landscape of third-party sellers exploiting the current market conditions, where demand surpasses supply. Unfortunately, some may end up buying from scalpers. In contrast, the previous generation cards have a much better availability, like the Gigabyte GeForce RTX 4060 OC, which starts at $299.99 on Amazon.

    Source:
    Link


  • Nvidia GeForce RTX 5090 Leak: 46% Performance Boost Explained

    Nvidia GeForce RTX 5090 Leak: 46% Performance Boost Explained

    Nvidia’s soon-to-be-released GeForce RTX 5090 has been the talk of the town lately, and rightly so. A recent leak has shown what its performance may look like, and it seems to be a significant improvement over the RTX 4090. With as much as 46% extra power in synthetic benchmarks, this top-tier GPU has already turned heads.

    Benchmark Insights

    Some 3D Mark scores for the RTX 5090 have appeared online, showcasing some remarkable figures. Here are the results:
    – 3D Mark Speed Way (1440p): 14,383
    – 3D Mark Steel Nomad (4K): 14,133
    – 3D Mark Port Royal (1440p): 36,673
    – 3D Mark Time Spy (1440p): 48,732
    – 3D Mark Time Spy Extreme (4K): 25,485
    – 3D Mark Fire Strike (1080p): 101,767
    – 3D Mark Fire Strike Extreme (1440p): 60,778
    – 3D Mark Fire Strike Ultra (4K): 33,114

    These scores indicate that the RTX 5090 is approximately 30-46% quicker than the RTX 4090, depending on which benchmark you look at. Nvidia’s own gaming tests support this, hinting at a 30% increase in pure gaming power. For example, in the challenging game Cyberpunk 2077, the RTX 5090 requires DLSS to achieve playable frame rates when Path Tracing is activated. This clearly shows that there’s some serious power within.

    A True Powerhouse

    When it comes to these benchmarks, the RTX 5090 is poised to be a powerhouse. While synthetic benchmarks might not give the entire picture—since performance can differ based on your CPU, RAM, and game optimizations—one thing is clear: this graphics card is set to dominate desktop GPUs for the foreseeable future.

    Don’t plan on AMD or Intel taking its crown anytime soon. AMD’s Radeon RX 9070 XT and even its future RDNA-based GPUs probably won’t reach the immense power of Nvidia’s flagship. The same goes for Intel’s Battlemage and Celestial series.

    Why It Matters

    You should definitely care about this. The RTX 5090 will likely be the fastest graphics card available for the time being. Of course, we’ll need to wait for real-world tests to see how it performs across various games and tasks, but if the leaks hold true, this GPU is going to raise the bar. For those seeking the absolute best, it’s tough to picture a better option. For more regular updates, don’t forget to check out our News Section.

    Source:
    Link

  • Nvidia Responds to RTX 5090 Power Connector Melting Issues

    Nvidia Responds to RTX 5090 Power Connector Melting Issues

    With the Nvidia RTX 5090 priced at $1999, there’s a growing concern about a potential melted power connector. This issue previously affected the RTX 4090, leading both Nvidia and third-party manufacturers to update their cabling solutions. The revamped 12V-2×6 16-pin cable has helped reduce incidents, but the new flagship GPU demands even more power. A Korean tech site, Quasar Zone, reported that Nvidia attempted to ease worries about this problem at a recent press event.

    Initial Concerns and Blame

    Soon after the RTX 4090 launched in October 2022, users began to voice complaints about overheating connectors. Nvidia initially pointed the finger at users, suggesting that improper seating of the connector could result in melting. However, this explanation didn’t sit well with many seasoned system builders. Ultimately, Nvidia encouraged the shift to the newer 12V-2×6 cables included with the more recent RTX 4090 cards. This change nearly eradicated instances of damaged GPUs, although older models with the previous connector were still available for sale.

    Power Draw and System Stress

    The earlier versions of the RTX 4090 came equipped with the 12VHPWR connector, capable of handling a 600-watt power draw. In contrast, the updated 12V-2×6 cables are designed to be more robust, featuring shorter and more efficient pins. The RTX 5090 requires 575 watts, an increase from the 450 watts of the previous generation Nvidia GPU, putting additional strain on power systems. Nevertheless, during the GeForce Update Editor’s Day in 2025, Nvidia expressed confidence that the 12V-2×6 connector would avert future issues.

    Manufacturer Innovations

    Not every manufacturer of the RTX 5090 is convinced that the standard 12V-2×6 design is sufficient. MSI has opted for yellow-tipped power adapters, which provide a clearer indication of a secure connection. With the new Nvidia GPU set to launch on January 30th, testers are currently pushing its performance limits. Should any issues arise due to the high power demands, it is hoped that potential buyers will be informed before making a purchase.

    Source:
    Link


  • GeForce RTX 5090 Launch Faces Supply Issues, Says PC Builder

    GeForce RTX 5090 Launch Faces Supply Issues, Says PC Builder

    It has been a recurring theme during the last two GPU releases from Nvidia: Team Green introduces a flagship 90 series card that boasts significant power but is tough to acquire at its suggested retail price (MSRP) due to limited availability. We saw this scenario unfold with both the RTX 3090 and the RTX 4090. Now, it appears that the upcoming RTX 5090, scheduled for launch on January 30 at an MSRP of $1,999, is likely to face a similar challenge.

    Supply Concerns

    Recent reports indicate that the RTX 5090 and RTX 5080 will have limited launch stock, potentially leading to a shortage of models at MSRP. For example, Moore’s Law Is Dead reported just days ago that even AIBs (Add-in-Board partners) are anticipating a supply situation akin to that of the RTX 3090.

    PowerGPU, a company specializing in custom gaming PCs, has corroborated MLID’s claims and cautioned gamers on X that “the launch of the RTX 5090 will be the worst when it comes to availability.” They do not foresee any improvements “for the first 3 months”.

    Retailer Differences

    It’s important to note that the supply chain used by custom PC builders may differ from that of larger retailers like Best Buy. This means that the overall stock levels could be somewhat better for these major retailers.

    If the RTX 5090 does indeed face stock shortages at launch, gamers can expect that the models priced at the official MSRP will be the most difficult to find. This aligns with the fact that the RTX 5090 variants priced at $2,000 will represent the lowest price point. AIBs typically prefer to equip flagship GeForce cards with premium cooling systems and sell them at higher prices than Nvidia’s Founders Edition models.

    Scalper Activity

    Moreover, if the demand for these GPUs exceeds supply, scalpers will likely swoop in to grab as much stock as they can, reselling the cards at inflated prices. We observed a similar pattern during the launches of the RTX 3090 and RTX 4090.

    In conclusion, it remains to be seen whether the launch of the RTX 5090 will be more favorable or less so compared to the RTX 4090 and the RTX 3090. We sincerely hope that the rumors regarding stock difficulties are exaggerated.

    Source:
    Link