TechnoSports Media Group

Tag: NVIDIA

The PlayStation 5 Pro’s Ray Tracing Revolution: What Gamers Need to Know

The gaming landscape is on the brink of a significant transformation with the awaiting arrival of the PlayStation 5 Pro. Recent leaks have unveiled exciting details about the console’s potential to deliver a quantum leap in ray tracing performance, promising visuals that could redefine the gaming experience.

At the heart of these advancements lies the RDNA 4 architecture, which is slated to power the PlayStation 5 Pro. Key features like a doubled ray tracing intersect engine and improved BVH compaction are poised to dramatically enhance the console’s ability to handle the computationally intensive demands of ray tracing.

These improvements are expected to translate to a staggering four times the ray tracing performance compared to the base PlayStation 5 model.

Unlocking the Potential of Ray Tracing

The intersect engine, a critical component in ray tracing, benefits immensely from the proposed enhancements. By repurposing texture units as ray tracing accelerators, the PlayStation 5 Pro can tackle the most demanding aspects of ray tracing with greater efficiency.

This translates to richer, more realistic lighting and reflections that were previously unattainable on console platforms.

Furthermore, the optimization of BVH compaction is another game-changer. BVH structures are essential for accelerating ray tracing calculations, and reducing their memory footprint is crucial for performance.

While NVIDIA has been at the forefront of BVH compaction, the PlayStation 5 Pro’s improvements in this area promise to narrow the gap and deliver significant performance gains.

The Path Tracing Challenge

Despite these impressive advancements, the road to full-fledged path tracing on the PlayStation 5 Pro remains challenging.

While older path-traced games like Quake 2 RTX might be within reach, more demanding titles powered by modern engines like Unreal Engine 5 could still push the console’s limits.

A crucial missing piece in the puzzle is hardware acceleration for ray traversal. This process involves efficiently determining the path of a ray as it interacts with the scene, and it places a significant load on GPUs. While Sony has previously hinted at a dedicated traversal unit through a patent, its inclusion in the PlayStation 5 Pro remains uncertain.

Balancing Act: Performance and Fidelity

The limitations imposed by backward compatibility and the need to deliver a smooth gaming experience on the base PlayStation 5 model further constrain the PlayStation 5 Pro’s ambitions. Developers will undoubtedly face challenges in striking the right balance between visual fidelity and performance.

The Looming Announcement

Rumors suggest that a PlayStation 5 Pro could be released before the year’s end, and an official announcement is expected soon. If these leaks hold true, gamers can anticipate a console that pushes the boundaries of visual realism and performance, even if it falls short of achieving full-fledged path tracing.

The PlayStation 5 Pro represents a significant step forward in console gaming, promising to deliver stunning visuals and immersive experiences. While path tracing may still be a horizon to chase, the advancements in ray tracing technology showcased by the console are undoubtedly a testament to the rapid evolution of gaming hardware.

Also Read: Xbox Game Pass August 2024: A Month of Blockbuster Releases and Indie Favorites

August 1, 2024
NVIDIA Readies GeForce RTX 3050 A Laptop GPU with AD106 ‘Ada’ Chip

NVIDIA is allegedly working on a brand new iteration of the GeForce RTX 3050 Laptop GPU that will introduce the Ada Lovelace architecture to laptops. Despite being an entry-level GPU, the RTX 3050 is set to receive an upgrade. As the PCI-ID database shows, a new version.

Now, it should be noted that, as of now Ampere architecture is still used for the laptop version of the NVIDIA GeForce RTX 3050. This confirms that NVIDIA will be using a new architecture (Ada) rather than the older one, to give another life to its RTX 3050 with Ada Lovelace GPUs, which are already mainstream in desktop and laptop markets.

More Info About NVIDIA GeForce RTX 3050 A

With the RTX 3050 name still in use for an Ada-based laptop GPU lineup that includes the RTX 4050, NVIDIA has adopted this measure to prevent anyone from getting mixed up. In the database, this new variant is named as AD106M [Geforce RTX 3050 A Laptop GPU] where the ‘A’ notification denotes to Ada architecture.

The current RTX 3050 laptop GPU is based on the Ampere GA107 die, but transitioning to an “AD106” mobile chip. The AD106 die is featured in the desktop RTX 4060 Ti, which has 4352 CUDA cores, while the current RTX 3050 has 2048 cores. The AD107 die, which could be used for the RTX 3050 A, supports up to 3072 cores on the RTX 4060 and 2560 cores on the RTX 4050. However, NVIDIA opted for the AD106 die, which has been confirmed by their drivers, according to Videocardz.

The RTX 3050 A has not been released and its full specifications are still unknown. The current RTX 3050 mobile GPU has a GDDR6 VRAM of only 4GB on the memory bus with an effective combined clock speed at which it operates at 12Gbps. The AD106 chip supports the same memory bus so it’s unclear if there would be an improvement in terms of memory speed on RTX 3050 A.

FAQs

What is the new feature of the NVIDIA GeForce RTX 3050 A Laptop GPU?

It will feature the AD106 chip based on the Ada Lovelace architecture.

When will the NVIDIA GeForce RTX 3050 A Laptop GPU be released?

The release date has not yet been announced.

July 26, 2024
NVIDIA GeForce RTX 50 “Blackwell” Gaming GPUs Reportedly Launching at CES 2025

According to Kopite7kimi, NVIDIA’s gaming GeForce RTX 50 “Blackwell” GPUs could now be unveiled during CES 2025. While it is a long way off before these cards actually get to market, the launch has been pushed back yet again, this time to early 2025, these new parts will be the latest high point for gaming dominance from NVIDIA.

More About the NVIDIA GeForce RTX 50 Blackwell

The original speculation was that NVIDIA’s GeForce RTX “Blackwell” GPUs would be arriving in Q4 2024, kicking things off with the RTX 5080 to be followed by the RTX 5090. That was at least the case according to recent reports that it is until it might be shifting towards a CES 2025 launch, or around then. Computex brought conflicting reports from different sources, one of which stated the GPU will launch in Q4 2024 while another pegged it for release on a later date – now pointing to early 2025.

NVIDIA is working on two flagship dies, the GB202 and the GB203, according to some earlier rumors. The GB202 is set to deliver power to the GeForce RTX 5090, offering up to a staggering 192 SM units with a huge 512-bit interface, and the GB203 likely used for the RTX 5080, will have 92 SM units and a 256-bit interface. Either way, both GPUs are expected to use a new GDDR7 memory interface.

According to PSU manufacturer Seasonic, who has the new NVIDIA GeForce RTX 50 GPUs listed on their wattage calculator webpage which shows slightly increased TDPs for its fresh graphics line. The numbers are almost certainly either partial or some form of an estimate. The 2025 launch means that both laptop and desktop GPUs in the Blackwell series will be available within the same year.

Since NVIDIA unveiled its RTX 40 “SUPER” lineup at the CES last year, it is expected that they would showcase their RTX 50 series there as well in line with what we have seen from them historically when bringing new desktop SKUs to life around Q3-Q4 and laptop counterpart near news years.

If these reports are accurate, we are unlikely to see any announcements or teasers regarding the gaming lineup at upcoming events like SIGGRAPH or Gamescom. However, whispers about the next-gen lineup from AIB partners may emerge. The NVIDIA GeForce RTX 50 “Blackwell” Gaming GPUs are expected to significantly advance PC graphics. Simultaneously, AMD is also rumored to launch its Radeon RX 8000 GPUs based on the RDNA 4 architecture around CES 2025.

FAQs

When are the NVIDIA GeForce RTX 50 “Blackwell” GPUs expected to launch?

The NVIDIA GeForce RTX 50 “Blackwell” GPUs are expected to launch at CES 2025 or early 2025.

What are the key features of the GeForce RTX 50 “Blackwell” GPUs?

The GeForce RTX 50 “Blackwell” GPUs will feature the GB202 and GB203 dies, with up to 192 SM units, a 512-bit interface for the RTX 5090, and a 256-bit interface for the RTX 5080. Both will use the latest GDDR7 memory interface.

July 23, 2024
Unreleased NVIDIA GeForce RTX 4070 10 GB GPU Spotted: Features 7168 Cores & 21 Gbps GDDR6X Memory

NVIDIA’s upcoming GeForce RTX 4070 10 GB GPU has been spotted at the Chinese reseller, Goofish, featuring the AD104-275 chip. There was a lot expected of several different configurations for the gaming graphics card before NVIDIA finally introduced its GeForce RTX 4070 GPU. An example of this kind of configuration would include an AD104-275 GPU paired with 10 GB memory. It turns out that this specific configuration was obviously in the works, as a board-partner-only graphics card is making its way to the market.

More About the NVIDIA GeForce RTX 4070 10 GB GPU

@Harukaze5719 discovered this NVIDIA GeForce RTX 4070 graphics card uses an unknown AD104-275 GPU die with 7168 cores, 224 TMUs, and 80 ROPs built. The original part from a year later is exactly what this core configuration suggests; the GeForce RTX 4070 SUPER. This card features clock speeds set at 2355 MHz base and 2520 MHz boost, which are much higher than the RTX 4070 and RTX 4070 SUPER models currently on the market. That works out to 22% more cores than the original RTX 4070.

In the memory front, the NVIDIA GeForce RTX 4070 GPU (unreleased) is configured with a slower 10 GB of VRAM along with a narrower bus interface at just 160-bit. While the memory modules support speeds up to 21 Gbps, providing up to 420.2 GB/s of bandwidth, this is lower than the 504.2 GB/s bandwidth and 12 GB VRAM offered by retail RTX 4070 and RTX 4070 SUPER cards. The graphics card operates on the “95.04.2E.00.01” BIOS and has a Device ID of “10DE 2785 – 10DE 16F8”.

The PCB design is much the same as that of the reference Founders Edition graphics cards with a circular cut-out to the rear. The card includes 5 GDDR6X memory dies and the chance for eight total. Accordingly, power is delivered by means of a single 16-pin connector and a 10-phase. It also features three DisplayPort (DP) ports and a HDMI output. This discovery speaks to how prepared NVIDIA is likely to turn up at launch with extensive pre-screening of potential configs to see what works best for the street.

FAQs

What are the core specifications of the GeForce RTX 4070 10 GB GPU?

It features the AD104-275 chip with 7168 cores, 224 TMUs, and 80 ROPs.

How much memory does the GeForce RTX 4070 10 GB GPU have, and what is its bandwidth?

It includes 10 GB of GDDR6X memory with speeds up to 21 Gbps, providing up to 420.2 GB/s of bandwidth, lower than that of the retail RTX 4070 and RTX 4070 SUPER cards.

July 16, 2024
NVIDIA GeForce RTX 5090 Rumored to Feature Near 3 GHz Base Clock, Marking a Significant Leap Over RTX 4090

NVIDIA appears to be pushing the envelope with its rumored GeForce RTX 5090 GPU, which has a base clock that nears the 3 GHz threshold and represents quite an improvement over current-gen cards like the RTX 4090.

The Rumored NVIDIA GeForce RTX 5090

The latest rumor comes courtesy of a Chiphell Forum member, Panzerlied, who is no stranger to dishing on NVIDIA’s mythical Blackwell “GeForce RTX 50” GPUs. While nothing as been officially confirmed, NVIDIA CEO Jensen Huang hinted at something remarkable during a Computex Q&A, teasing anticipation for the next-gen 50-series.

But the most recent ‘leak’ is hinting that the yet-to-be-announced GeForce RTX 5090 will have a base clock of around 2.9 GHz, almost hitting the hallowed 3 GHz mark. This is a 30% jump over the RTX 4090’s base clock, which tops out at 2235 MHz.

As to the boost clock, it is yet unknown though previous models have pushed this over a 3 GHz threshold due to fetched tradition. The RTX 4090 could regularly boost to speeds of about 2.8-2.9 GHz straight out-of-the-box, with some chips hitting as much as a round 3GHz when conditions were just right.

In particular, the GeForce RTX 5090 has been tipped to feature a 448-bit bus interface (potentially 512-bit for the full chip), 14 GDDR7 modules (16 in the full configuration), and up to 28 GB of VRAM. On the other hand, if official specifications are to be believed then RTX 4090 comes with a similar 384-bit bus, 12 GDDR6X modules, and 24 GB of VRAM.

The GeForce RTX 5090 is slated for release later this year alongside the RTX 5080, though there are rumors of potential delays pushing the launch into early 2025. This timing aligns with competitive pressures as both Intel and AMD gear up to launch their next-gen platforms, intensifying competition in the high-performance desktop PC market.

FAQs

When is the NVIDIA GeForce RTX 5090 expected to launch?

The NVIDIA GeForce RTX 5090 is rumored to launch in late 2024 or possibly early 2025.

What are the key specifications of the NVIDIA GeForce RTX 5090?

The NVIDIA GeForce RTX 5090 is rumored to feature a base clock speed near 3 GHz, a 448-bit memory bus, and 28 GB of GDDR7 VRAM

July 11, 2024
NVIDIA: 25 Times More Valuable Than Intel and the Youngest Among the Top 3 Most Valuable Firms

Wall Street’s artificial intelligence boom could arguably be at full throttle, a development that’s propelled NVIDIA Corporation to a historic peak as the world’s most valuable company. According to NASDAQ’s data, NVIDIA had a market capitalization of $3.33 trillion, nudging the previous leader Microsoft out of the way by about $18 billion at today’s close. For the last month or so, NVIDIA has been trading above and below Apple’s market cap, but it finally did it – passing Microsoft to claim first place on this expensive list.

More About NVIDIA’s Value

The Nvidia took over of first place comes after several quarters of NVIDIA proving its dominance as the company’s AI accelerators and GPUs continue to impress both Wall Street and industry luminaries week after week. Based in the early 1990s, originally focused on consumer gaming graphics cards. Over the years, NVIDIA has shifted gears and began heavily investing in developing AI model training and execution workflow products.

NVIDIA’s CEO Jensen Huang has forged a close relationship with Microsoft-backed OpenAI to supply its latest GPUs for an AI project. NVIDIA differentiates itself from the competition not only by providing advanced products but also by assisting in developing use cases for implementation across various domains (including HPC, automotive, healthcare, and supercomputing).

In the past, the crash caused by the epidemic or a previous crypto havoc might have severely damaged NVIDIA, but not anymore as it now stands at No. 1 in market valuation rankings. Meanwhile, Intel has fallen from grace; when strict and unquestionable in Nordlinge’s era, it now represents something of an airy-fairy microprocessor heaven. NVIDIA’s GPUs, originally designed to process game graphics but refined for AI processing and manufactured under a contract model led by Taiwan Semiconductor Manufacturing Company (TSMC), have overshadowed the traditional CPU-centric models.

Intel, the world’s biggest maker of chips, has an estimated market value of $130 billion, according to NASDAQ data, almost 25 times less than NVIDIA’s estimated $3.33 trillion market value. Of the three companies that are worth the most in Silicon Valley — Apple, Microsoft, and NVIDIA — NVIDIA is best known for being the youngest even though it has been incorporated earlier than its higher-grossing counterparts.

NVIDIA’s market dominance in the semiconductor space is exceptional not only when measured against rival AMD (NASDAQ: AMD) with its own $249 billion market capitalization and its licensing partnership with Taiwan Semiconductor Manufacturing. That second one is the whole market cap of Intel and AMD combined.

In essence, the incredible rise of NVIDIA to being the world’s most valuable company highlights exactly that, its central role in defining the future of computing and AI. This decision focused on innovation using GPUs to deliver is not only great for business but has entirely reshaped the global semiconductor technology scene.

FAQs

What led to NVIDIA becoming the world’s most valuable company?

NVIDIA’s leadership in AI accelerators and GPUs, pivotal in sectors like computing and autonomous driving, drove its market value to $3.33 trillion, surpassing rivals.

How does NVIDIA compare to Intel and AMD in market value?

NVIDIA’s $3.33 trillion market cap is 25 times greater than Intel’s $130 billion and significantly higher than AMD’s $249 billion, showcasing its dominance in the semiconductor industry

June 20, 2024
NVIDIA DLSS 3: Elevating Your Gaming Adventures with Cutting-Edge Technology

This week’s DLSS announcement is all about harnessing the power of RTX technologies like NVIDIA DLSS 3, NVIDIA Reflex, and advanced ray-traced effects to make your gaming experiences more immersive than ever before. From social sandbox MMOs to intense first-person horror games, DLSS 3 enhances performance and visual fidelity across a variety of titles.

Embrace the unknown with Pax Dei, Still Wakes the Deep, and Skye: The Misty Isle, all launching this week with AI-powered DLSS 3 and Reflex

Discover the Magic of Pax Dei

Step into the enchanting world of Pax Dei, a social sandbox MMO available in Early Access on Steam. In Pax Dei, you can carve out your own legend amidst real myths and magic. With the integration of DLSS 3 and NVIDIA Reflex, your adventures are set to run at peak performance. DLSS 3 uses AI to boost frame rates by an average of 3X at 4K resolution, ensuring smooth gameplay as you navigate this mystical realm.

Face Your Fears in Still Wakes The Deep

For those seeking a more structured narrative, Still Wakes The Deep transports you to an oil rig besieged by a violent storm off the coast of Scotland. Cut off from the outside world, you’ll confront a terrifying presence in this first-person horror game. DLSS 3 with Frame Generation and DLAA makes every scream and shadow more lifelike, while Reflex optimizes your chances of survival when the situation gets dire. Experience a 2.6X average frame rate uplift at 4K with max settings. Ready to face your fears? The game is available to play today.

Explore the Isle of Skye in Skye: The Misty Isle

If you’re looking for a less stressful adventure, sail through Aurora Game Studios’ Skye: The Misty Isle. Brave the elements, fend off predators, and uncover the island’s hidden secrets. When Early Access launches on June 19, DLSS 3 with Frame Generation will amplify frame rates by an average of 2.4X, while Reflex minimizes latency, providing a seamless and engaging experience.

More Exciting Updates and Giveaways

The excitement doesn’t stop there! Deep Rock Galactic Season 5 and THE FINALS Season 3 are both incorporating DLSS 3, with THE FINALS also featuring Reflex and Ray Tracing. This summer, the #GeForceSummer of RTX offers chances to win everything from GPUs to fully equipped PCs. To participate, check out the full Terms and Conditions and contest details on the official website.

Stay tuned for more DLSS integrations arriving soon across a variety of games. For the latest updates, trailers, and performance charts, visit GeForce.com. Explore the full list of over 500 RTX-enhanced games and apps and keep returning for the latest in DLSS and RTX advancements.

June 19, 2024
AMD Instinct MI300X Surpasses NVIDIA RTX 4090 as Fastest GPU in Geekbench OpenCL Benchmark

The first benchmarks of AMD’s Instinct MI300X 192 GB GPU have been published on the Geekbench OpenCL benchmark, highlighting the performance capabilities of AMD’s next-generation accelerator. As such, the focus of the AMD Instinct MI300X will be squarely on Data Centers for solving the sort of AI workloads that are in demand today, putting it head-to-head with NVIDIA’s Hopper AI GPU family. Best of all, the MI300X ranks better in the GPU world than almost any of our processors in the Geekbench OpenCL benchmark-led by a significant margin.

More About AMD Instinct MI300X Benchmark

Some quick specs here with the AMD Instinct MI300X featuring the widespanning CDNA 3 arch with its 153 billion transistors in a die-chaplet packaging design. The chip includes a total of 28 dies comprising 8 HBM3 packages that allow for up to 192 GB of memory, the currently largest memory capacity of all Data Center GPUs.

The Instinct MI300X comes with 304 compute units and 19,456 cores on the GPU side. This gives it an enormous 5.3 TB/s of VRAM bandwidth and 896 GB/s of Infinity Fabric interconnect speeds. so with those two combined you’re looking at a gaming performance beast. In terms of power, the MI300X has an rated 750W TDP which is almost double of NVIDIA’s RTX 4090 with 450W rated TDP.

The AMD Instinct MI300X scored 379,660 points in the Geekbench 6.3.0 OpenCL test. These benchmarks were done with an AMD EPYC setup of two 9754 processors for a total of 128 cores / 256 threads and 3TB of system memory. Meanwhile, NVIDIA‘s flagship RTX 4090 scored 319,697 points and AMD’s top gaming GPU, the 7900 XTX scored 207,354 points.

These benchmarks highlight the remarkable speed of modern Data Center and AI GPUs. However, their high power consumption, cooling requirements, and steep prices—such as the MI300X priced around $15,000 USD—make them unsuitable for standard PC platforms.

Additionally, their specialized hardware designs and lack of optimized drivers for gaming and traditional PC applications further limit their viability outside of Data Centers, AI, HPC, and Cloud servers. AMD’s roadmap includes future releases like the MI350 and MI400 series, continuing to advance the capabilities of their Instinct GPUs in specialized computing environments.

FAQs

June 15, 2024
New Games Featuring NVIDIA DLSS, Reflex, and Ray Tracing: Get Ready for an Enhanced Gaming Experience
Every week, new games integrate NVIDIA DLSS, NVIDIA Reflex, and advanced ray-traced effects, making them run better than ever with the power of RTX technology. Today, we’re excited to highlight a selection of story-driven games that are now enhanced by these technologies.

Celebrate PC Gaming with NVIDIA: Enter to Win Big Prizes

To celebrate the vibrant past, present, and future of PC gaming, NVIDIA is launching an exciting giveaway and offering fantastic prizes, including GeForce RTX 40 Series GPUs, sleeper PC builds, RTX laptops, G-Sync gaming monitors, backplates, keycaps, posters, and more. Here’s how you can enter:
1. Follow NVIDIA: Stay connected with on Facebook, Twitter, Instagram, TikTok, and Twitch.
2. Watch for Prompts: Look out for key prompts in our contests or sweepstakes (India is eligible).
3. Use Hashtags: When prompted, use #RTXOn or #GeForceSummer in your posts throughout the summer.
What Can You Play with Your New Gear?

Night Springs – Alan Wake 2 Expansion

Starting today, dive into Night Springs, the new expansion for Remedy Entertainment’s critically acclaimed Alan Wake 2. Help Alan Wake escape the endless loops of the Dark Place by writing new scripts for a TV show he once worked on. Playing on a GeForce RTX 40 Series PC will offer one of the most immersive gaming experiences, thanks to Full Ray Tracing and DLSS 3.5 with Ray Reconstruction.

Upcoming Titles to Get Excited About

Black Myth: Wukong

A new CG trailer has confirmed Black Myth: Wukong’s release date for August 20. Enhanced with full ray tracing and NVIDIA DLSS 3.5 with Ray Reconstruction, this title promises an unparalleled experience on GeForce RTX PCs.

Star Wars™ Outlaws

A new gameplay showcase trailer and artwork reveal the late-August release of Star Wars™ Outlaws. This highly anticipated game will support DLSS 3.5, Ray Tracing, and Reflex. Check out the Star Wars™ Outlaws DLSS 3.5 World Premiere video to see it in action.

Pax Dei

The social sandbox MMO Pax Dei enters Early Access on June 18th with day-one support for NVIDIA DLSS 3 and Reflex, ensuring that GeForce RTX gamers get the best possible experience from day one.

Stay Updated

NVIDIA’s blog features even more DLSS integrations across a wide range of games. Be sure to visit regularly for updates on the latest DLSS and RTX titles. For a comprehensive list of over 500 RTX-enhanced games and apps, head to our detailed RTX Game List.

Stay tuned for more exciting news and updates from NVIDIA, and don’t miss your chance to win incredible prizes this summer!
June 12, 2024
NVIDIA GeForce RTX 50 ‘Blackwell’ GPU Configurations Leaked: GB202, GB203, GB205, GB206, GB207

NVIDIA has been teasing its next-gen GeForce RTX 50 “Blackwell” series, with GB202, GB203, GB205, GB206, and GB207 GPU configurations out today. The next generation GeForce RTX 50 “Blackwell” GPUs are in the works by NVIDIA for gamers and details are being teased by leaker, Kopite7kimi. So to set the table and establish what the base specs for each new Blackwell Gaming GPU are going to be, we’ll assume that they are built on these core specs.

More About NVIDIA GeForce RTX 50 Blackwell Lineup

At the top, NVIDIA GB202, will be the flagship GPU die for GeForce RTX 50 “Blackwell” series. This beast will power ultra-enthusiast GeForce RTX 5090 and beyond. The GB202 GPU boasts 12 SMs, with 8 TPCs per SM, totaling 96 TPCs, and each TPC accommodates 2 SMs, amounting to 192 SMs. It also supports a 512-bit wide memory bus interface that is split up into flexible configurations including 448-bit, 384-bit, 320-bit, and others.

The NVIDIA GB203 GPU, which is targeted toward high-end enthusiast gamers, has 7 GPCs, 6 TPCs per GPC, and 42 TPCs in total, resulting in 84 SMs. The latter has a 256-bit bus interface, and will reportedly underpin the GeForce RTX 5080 graphics card.

NVIDIA GB205 is a mid-tier of high-end performance. The NVIDIA GB205 is the successor to the NVIDIA GA104 and is used to power the GeForce RTX 5070 and RTX 5070 Ti graphics card sisters. This consists of 5 GPCs, 5 TPCs per GPC for a total of 25 TPCs, as well as 50 SMs, with a 192-bit memory bus setup.

The NVIDIA GB206 GPU has 6 TPCs per GPC or a total of 18 TPCs (36 SMs) for 36 SMs on the NVIDIA GB206 GPU for mainstream gamers. This card has a 128-bit bus interface and as a result, it may turn out to be sufficiently pathetic for the next-generation GeForce RTX 5060 (Ti) family. Finally, the NVIDIA GB207 GPU for entry-level models like the GeForce RTX 5060 or RTX 5050, is constructed with 2 GPCs and 5 TPCs each for a total of 10 TPCs and 20 SMs. The chip works with a 128-bit memory interface, which may allow both GDDR7 and GDDR6 memory to be supported.

Finally coming to the launch timeline, NVIDIA is apparently looking at a number of cooling solutions with the flagship GeForce RTX 5090 SKU to feature a dual-slot/dual-fan design. The RTX 50 family is anticipated to debut later this year, with the RTX 5080 arriving first followed by the RTX 5090. However, there are indications that the lineup may be delayed until early 2025, as suggested by discussions with manufacturers during Computex. Additionally, the laptop GeForce RTX 50 lineup, featuring at least six SKUs with up to 16 GB GDDR7 memory, is expected to debut in early 2025 alongside next-gen Arrow Lake-HX and Fire Range-HX designs.

FAQs

June 12, 2024