The immediate impact on publishers

Let’s start with the publishers. These are the organizations and individuals who invest countless hours and resources into creating the high-quality content we consume daily. With Google’s AI summaries providing key information upfront, users have less incentive to visit the original websites. This means fewer clicks for publishers, leading to a drop in ad revenue, subscriptions, and overall engagement.

As publishers, we craft our articles with care, adding depth, context, and a unique voice. But when our content is reduced to a couple of sentences on a search page, we lose the opportunity to connect with our readers. This reduction in traffic doesn’t just affect the bottom line; it undermines the relationship between publishers and their audience. The trust and loyalty built over years can erode when readers no longer engage directly with the source.

Also Read: Gadget shopping hacks with a twist

Readers get shortchanged

Now, let’s consider how this affects us as readers. When publishers see diminishing returns from their content, they’re forced to adapt. Some may decide to produce more generic, click-friendly pieces tailored to fit AI algorithms, stripping away the nuanced analysis and distinctive voices that make journalism engaging and informative.

Imagine reading article after article that feels bland and impersonal because it’s been optimized for AI summarization rather than human connection. The richness of diverse perspectives, in-depth reporting, and thoughtful opinion pieces could give way to homogenized content that fails to challenge or inspire us.

Moreover, if high-quality publishers start pulling out or paywalling their content due to revenue losses, we might find ourselves with access only to lower-quality sources. This not only diminishes our reading experience but also poses a risk to how well-informed we are about the world.

The inevitable consequence for Google

You might think that Google benefits from keeping users on its search page, but this strategy could backfire in the long run. Google’s search engine is only as valuable as the quality of the content it can provide access to. If publishers reduce the quality or quantity of their content, or restrict Google’s ability to index it, the overall usefulness of Google’s search results declines.

Users may begin to notice that the information they’re getting is less reliable or lacks depth. Trust in Google’s ability to deliver accurate and comprehensive search results could wane. In a digital age where alternatives are just a click away, this could lead users to seek out other platforms or search engines that prioritize quality content and respect the publisher-reader relationship.

Connecting the dots

Connecting the dots, it’s clear that Google’s AI summaries initiate a domino effect. Publishers suffer first as reduced traffic leads to lower revenue and forces them to alter their content strategies, often at the expense of quality and originality. Readers feel the impact next; we receive less engaging, less informative content, the diversity of voices shrinks, and the depth of journalism declines. Ultimately, even Google stands to lose. As the quality of content decreases, so does the value of Google’s search results. User trust erodes, potentially driving people away from the platform.

Is convenience worth the cost?

While the allure of instant information is strong, it’s worth questioning whether this convenience is worth the broader costs. High-quality journalism and content creation are pillars of an informed society. They foster critical thinking, hold power to account, and enrich our cultural landscape. By disrupting the ecosystem that supports these endeavors, we risk not only the livelihoods of those who create content but also the quality of information available to all of us. In the end, an internet filled with generic, soulless content benefits no one—not publishers, not readers, and not even the platforms that host it. 

Where do you stand on this issue? Do you see AI-generated summaries as the way forward, or do you fear they might commoditize journalism and ultimately destroy its true essence? I invite you to share your thoughts on this sensitive topic, which has serious implications for the very foundation of the internet as we know it today. Please write to me at soham@digit.in—I look forward to hearing your perspective.

Q) I’ve been using Mygate for about seven years and have observed its significant evolution beyond basic gatekeeping into offering numerous new features and value-added services. How would you define what Mygate stands for today, and how would you describe its mission and vision in 2024?

Mygate was always about solving the problems that are thrown up by new ways of living. The rise of gated communities, coupled with ever-growing deliveries, was a big change which created a new set of challenges. In solving these, we tackled the gate first and solved the friction by easing the approval process at the gate. So, we were always focused on solving for the changing way of living and how we could make it better. Today we’ve evolved into a comprehensive lifestyle platform that enhances everyday living within gated communities through technology. In 2024, our mission remains to create a secure, connected, and convenient environment for our users. The vision is centred around ‘Living Experience Tech,’ where we integrate advanced technology into daily life to foster a more engaging, safer, and socially connected community. Whether it’s home security, managing deliveries, or connecting residents, we’re building an ecosystem that improves the quality of living for millions of Indian families.

Also Read: Indians want more refurbished smartphones, Cashify explains why

Q) The launch of Mygate Locks marks the company’s foray into the smart consumer devices segment, aligning well with its overarching goal of home security. How does Mygate plan to compete with established players in this segment, especially considering that consumers often trust well-known brands like Godrej for security and are willing to pay a premium for them? Additionally, what synergies and integrated features have you incorporated between Mygate Locks and the Mygate app to make the adoption of smart locks more seamless and attractive?

We recognize that established players have built a legacy of trust over decades. However, our unique advantage lies in understanding the specific needs of gated communities and modern households, thanks to our years of experience serving millions of Indian families. Mygate Locks are not imported products; they are thoughtfully designed and built specifically for Indian households. With this in mind, we offer a fully integrated solution that combines security and convenience, surpassing what standalone products on the market provide.

For example, Mygate Locks come with a Tamper Alert feature that can instantly notify relevant authorities, such as the security guard at the gate, in addition to alerting the user in case of tampering. Moreover, our locks allow for recurring timed access, making it easier for daily help to enter the home at set hours, which adds an extra layer of convenience for residents.

The synergy between Mygate Locks and the Mygate app is a core differentiator. Our locks seamlessly integrate with the app, enabling users to manage both their community gate access and home door from a single platform. This experience includes features like real-time notifications, remote access, and scheduled access for daily help, all designed to make adoption smoother and more attractive. And since our app is already a part of daily life for millions, this integration streamlines the onboarding process for smart locks.

Q) What future product offerings can we expect from Mygate, and how do they align with the company’s core values and mission? Currently, what portion of the addressable apartment market in India does Mygate cover, and what strategies are in place to reach and serve the remaining market?

Looking ahead, we are expanding into the broader smart home ecosystem with integrated devices such as security cameras, video doorbells, and home automation solutions. Each of these products is designed to work seamlessly within the Mygate ecosystem, enhancing security, convenience, and the overall living experience in a unified platform. This approach aligns with our core values of fostering a safe, connected community and delivering a smooth, user-friendly experience from a single app.

In terms of market penetration, we currently serve over 25,000+ communities across India, covering a significant portion of the addressable apartment market. However, there is substantial potential for growth. Our strategy includes expanding into Tier 2 and Tier 3 cities, where gated community living is on the rise, and extending our solutions to individual homes, so more people can benefit from our technology. This integrated approach not only addresses the needs of large communities but also offers a consistent, high-quality experience across different residential setups.

Q) As a technology-driven company, how is Mygate leveraging artificial intelligence in its current offerings, and are there plans to introduce AI-led solutions in the future? As the founder and CTO, where do you envision Mygate five years from now? Will the focus be more on expanding market coverage, or on bringing innovative products and solutions to the market to stay ahead of technological trends?

Currently, Mygate uses AI to unlock efficiency in development and to enhance customer support, with a significant number of queries now handled directly by AI agents. Our goal is to have AI resolve up to 90% of support questions in the near future. At the same time, we’re experimenting with AI models to

enhance community life. We have been working on creating an AI community helper. For example, a problem that is often faced is in managing community WhatsApp groups, with frequent move-ins and move-outs, and new residents often rely on others to introduce them to the right group. AI-driven solutions to manage the chat groups within Mygate can allow communities to simplify this process, making community integration smoother. Another POC we are doing is a community AI knowledge bank which can suggest the best pediatric doctor around and the best eatery nearby, all based on community discussions.

Reflecting on our journey, we’ve consistently expanded market coverage while driving continuous innovation. We believe that growth and innovation go hand-in-hand; without innovating, we cannot grow.

Q) Given that Mygate handles a significant amount of user data, how does the company utilize data and insights to create and enhance its products and services? What measures are in place to ensure user privacy and data security while leveraging this data for innovation?

Customer experience is at the heart of what we do, and data-based insights allow us to refine our products and develop new services that cater to the evolving needs of our users. For instance, we use data to optimize security features, improve delivery management, and tailor our app’s user experience.

When it comes to data privacy and security, we take a rigorous approach. All user data is encrypted, and we follow industry-leading standards such as ISO 27001 for data security and are DPDP-ready. Additionally, all our data is stored within India on secure Mygate servers. We are committed to maintaining the trust our users place in us, and we ensure that any data used for innovation is anonymized and handled responsibly, with strict privacy controls in place.

Lookout for fake deals!

In previous years, I’ve eagerly upgraded my devices during festive sales, enticed by flashy discounts and the allure of new features. However, I’ve noticed that these “Diwali offers” aren’t always the best deals. For instance, when I was looking to upgrade my microwave oven earlier this year in May, the price offered then was a good 15% lower than the “Diwali dhamaka deal” being advertised now. It seems that sometimes prices are inflated during festive seasons to make discounts appear more attractive. This experience taught me not to get lured by every discount and to check price histories whenever possible.

Buyback is good

One practical step is to take advantage of buyback or trade-in programs. While it might seem tempting to hold onto old devices in hopes of getting a better deal later, they often lose value the longer they stay unused. For example, I’ve been sitting on a very old gaming desktop PC, convincing myself that one day I would rebuild it. That day has never arrived, and now the parts are no longer functional. I could have given it to someone who might have used it, instead of letting it gather dust for years. Eventually, devices become obsolete, and disposing of them irresponsibly adds to the e-waste problem. Buyback offers provide a convenient and responsible way to dispose of old gadgets, ensuring they’re recycled or refurbished appropriately. This Diwali, as you consider purchasing from our top 100 gadget recommendations, I encourage you to utilize these programs.

Need vs. Want, where do you stand?

It’s also worth reassessing whether an upgrade is truly necessary. Festive deals and marketing campaigns can create a sense of urgency for the latest models, but perhaps our current devices still meet our needs. Upgrading out of necessity rather than novelty helps reduce unnecessary consumption and the resulting e-waste. By being mindful of our actual requirements, we can make more sustainable choices.

“The bottom line is to shop smart: buy what you need, not just what’s offered at a seemingly great discount”

Don’t ignore offline

It’s also beneficial to compare online and offline prices. I’ve seen offline retailers not only match the best online offers but also include additional goodies. By shopping around, you might find better deals than simply jumping at the first discount you see. The bottom line is to shop smart: buy what you need, not just what’s offered at a seemingly great discount.

Old gadgets deserve a responsible send-off

My drawers are now a collection of gadgets that seemed essential at the time but ended up unused. This accumulation doesn’t just clutter my space; it contributes to the growing problem of electronic waste. E-waste is a significant environmental concern, with millions of tons generated globally each year. Old gadgets contain hazardous materials that can harm the environment if not disposed of properly. By hoarding unused devices, we’re inadvertently adding to this issue.

Another meaningful option is to donate functional gadgets to those who need them. Many schools, non-profit organizations, and individuals could greatly benefit from technology we no longer use. Donating devices can help bridge the digital divide, offering others access to resources that enhance education and connectivity. Researching products thoroughly ensures we select devices that truly meet our needs and offer longevity. Investing in high-quality gadgets rather than multiple lower-quality items is more economical and environmentally friendly in the long run. Setting a budget and sticking to it can prevent overspending driven by impulse rather than necessity.

This Diwali, I’m committing to responsible tech consumption. By not purchasing new gadgets until I’ve responsibly managed my old ones, I’ve become more mindful of my habits and their broader impact. Sharing this journey, I hope to inspire you to reflect on your purchasing decisions during the festive season.

Which gadget did you buy this Diwali? Share your Diwali shopping haul with us! Send me your top deal grab at soham@digit.in and you could be featured as one of Digit’s top deal hunters in the next issue!

However, we did get our hands on some of the big Lunar Lake laptop launches at IFA 2024, and here are our initial impressions. Note that we weren’t allowed to run benchmarks to test Lunar Lake performance at the launch events, so our impressions are based on laptop specifications and features.

Also read: How Intel Lunar Lake is taking the fight to the competition

Asus Zenbook S 14
The Zenbook S 14 is probably our pick of all the Intel Lunar Lake-based premium thin and light laptops revealed at IFA 2024. We like the design innovation that Asus brings to the table with the Zenbook S 14. The “Ceraluminum” treatment (finish) on the Asus Zenbook S 14 is unique in many ways, from the use of materials to the feel and look it adds to the laptop. Asus has attempted much-needed meaningful differentiation in material science, something the category desperately needs, as we start seeing very similar-looking and feeling (to touch) premium thin and light laptops that come across as generic metal slabs with little identity or character.

We tested and reviewed the Asus Zenbook S 15 earlier in July 2024 and liked what we saw; the Zenbook 14 is essentially a 14-inch version of the Zenbook S 15, albeit with Intel Lunar Lake inside. The Asus Zenbook 15 we tested was based on the AMD Ryzen AI 300 series processor. The most interesting bit, though, is that if we take Intel’s word, the Asus Zenbook S 14 should be able to beat the Asus Zenbook S 15 across every benchmark, be it gaming, battery life, AI performance, or even single-core CPU performance. And just for the record, the Asus Zenbook S 15 is the fastest thin and light laptop we’ve tested so far at our test labs, so we can’t wait to see if the Zenbook S 14 can take the crown away.

When it comes to specifications, Asus has listed the Zenbook S 14 (UX5406) with four Intel processor options, starting from the very top-end Intel Core Ultra 9 288V processor with 48 TOPS of performance and the fastest variant of the Intel Arc 140V 8-core iGPU. You can read more about the full specification of the Asus Zenbook S 14 here. The Zenbook S 14 felt light, premium, and very well put together. The 14-inch 3K OLED display is sharp, and we won’t need too much testing to confirm that it is one of the better OLED displays in terms of color accuracy and response time; we’ve thoroughly tested the 15-inch variant with the exact same Asus Lumia OLED tech in the Asus Zenbook S 15. Overall, the Zenbook S 14 is one of our top picks when it comes to the new laptop launches that came our way at IFA 2024. We await the review unit.

Samsung Galaxy Book 5 Pro 360
The first laptop from Samsung in the updated Galaxy Book 5 lineup is the Galaxy Book 5 Pro 360, a convertible (2-in-1) laptop with a 360-degree display that allows the laptop to be used in tablet mode for creator-centric workloads. The large 16-inch, 3K Dynamic AMOLED touch display can go up to 120Hz and claims 500 nits of peak brightness. We are fans of Samsung’s latest Dynamic AMOLED 2X touch displays on laptops for their non-reflective treatment of the displays, which surely fix one of the biggest issues we have with AMOLED displays – excessive reflections that impact viewability in brightly lit environments.

On the design front, we appreciate the Galaxy Book 5 Pro 360 for making the most of the 16-inch form factor while keeping the overall size and weight in check. The display has one of the slimmest bezels we’ve seen on a 16-inch laptop. Another area where the Book 5 Pro 360 stands out is the super-large touchpad, which has been optimized to prevent unintended activation during typing by fine-tuning the sensitivity of the touchpad when it comes to the borders. The Book 5 Pro 360 will be offered with Intel Core Ultra 5 and Ultra 7 Series 2 processors; we aren’t sure which model will make it to India. For more details on the Samsung Galaxy Book 5 Pro 360, read our launch story here.

Acer Swift 14 AI
Acer announced its new Intel Lunar Lake-powered lineup with two Swift series models – the Swift 14 AI and Swift 16 AI. We checked out the Acer Swift 14 AI, and the first thing that struck us is that the overall chassis and design are pretty much the same as the Swift AI powered by the Qualcomm Snapdragon Elite SoC. What is different, though, is that the Intel Core Ultra Series 2-powered laptop promises higher battery life at a staggering 29 hours on a single charge and is lighter at 1.26 kilograms. The Acer Swift 14 AI is the highest when it comes to battery life claims on a thin and light laptop that we’ve got our hands on to date.

Another cool thing about the Acer Swift 14 AI is the NPU activity indicator, which glows whenever the NPU is in use. This is a neat feature, as it kind of alerts you when any applications are making use of the NPU. We see it being very helpful for testers and power users who are interested in knowing exactly when the NPU is called for action during running different types of workloads. You can read more about the Acer Swift 14 AI and the Swift 16 AI in our IFA 2024 Acer launch story.

MSI Claw 8 AI Plus
MSI did launch the Prestige, Summit, and Venture series of Intel Lunar Lake-powered laptops at IFA 2024, but what caught our attention was the MSI Claw 8 AI+. MSI launched the Claw 7 on the Intel Meteor Lake series at CES 2024, and now they have upgraded the Claw to a larger, more beefed-up 8-inch handled Windows 11-powered gaming console with the MSI Claw 8 AI Plus. The Claw 8 is going bigger on every aspect – a larger 8-inch display, a big jump in the battery pack with an 80Whr (compared to the 53Whr in the Claw 7), and it will be powered by the latest Intel Lunar Lake-based Core Ultra 200V series processor and will also run on a faster SSD besides offering an additional Thunderbolt 4 port. Surely, we expect the Claw 8 to offer a much better gaming experience and battery life without any thermal throttling issues, thanks to the much more efficient and graphically advanced Intel Lunar Lake platform.

During our brief encounter with the MSI Claw 8 AI Plus, we found the console to have a similar design and overall in-hand feel as the MSI Claw 7. Sure, the device has got a bit heavier, but the balance is well maintained thanks to the increase in volume. MSI has not launched the device yet; they’ve announced it, and we expect the official launch to be in early 2025.

Dell XPS 13
We got our hands on the latest Lunar Lake-equipped Dell XPS 13 at the demo area, and the first thing to report is that all the changes are on the inside. Dell has upgraded the innards of the XPS 13 with Intel Core Ultra 200V series processors, with configurations going up to Intel Core Ultra 9 288V with 32GB of memory. The fact that Dell has kept the overall design and chassis the same is a very good thing. We love Dell’s seamless XPS 13 design with the invisible touchpad. It does have a unique appeal, and Dell has also done a very good job with the implementation. Dell claims 26 hours of battery life on the Dell XPS 13 with WiFi on and streaming Netflix at 1080p resolution. We’d like to test that claim ourselves to be doubly sure on that number, but we do expect the Intel Lunar Lake-powered XPS 13 to take a substantial jump across performance and battery life compared to the earlier Intel Meteor Lake-powered XPS 13. 

There’s no official information on the India launch yet. We do expect Intel Lunar Lake-powered laptops to make it to the Indian market real soon. Let’s hope a few also make it to the shelves before the buying season kicks off. We’d want to test and update our recommendations and laptop buying guides with Intel Lunar Lake-based laptops once we’ve got the Digit Rating out for them.

Also Read: Intel challenges Qualcomm and AMD with Lunar Lake launch

Most power-efficient Intel CPU ever
Intel’s Lunar Lake platform marks a significant milestone with the most power-efficient Intel CPU ever. The company has made bold and ambitious decisions, starting with 8-core configurations across the lineup, notably without hyperthreading. This deliberate choice aims to enhance performance per watt through a two-cluster approach, featuring four power-efficient E-cores for workloads that don’t require high performance. Building on Meteor Lake’s 2x Crestmont E-cores with a 2MB shared L2 cache, Lunar Lake boasts 4x Skymont E-cores with a 4MB shared L2 cache, introducing several enhancements. These include improved instruction prediction, increased parallelism with 26 dispatch ports, and deeper queueing for better resource utilization. Furthermore, the Skymont E-cores feature wider allocation and retire mechanisms, enabling E-core efficiency to match P-core performance levels. The Enhanced Memory subsystem, with 4x 128-bit FP & SIMD vector processing, claims to deliver double the AI throughput compared to the previous Meteor Lake platform. On the performance cluster, featuring four Lion Cove P-cores, Intel asserts the fastest single-core performance in the industry for thin and light AI PCs, within a power envelope of 8W-37W, with a maximum of 37W. Intel’s Lion Cove P-core architecture offers enhanced processing efficiency through improved scheduling capabilities and 18 execution ports, alongside access to up to 12MB of shared L3 cache for accelerated AI and compute-intensive tasks. Additionally, Intel highlights its AI-driven power management system and enhanced memory subsystem as key optimizations for power per area (PPA), with finer control over power consumption via 16.67MHz clock interval control. According to Intel, these advancements achieve a balance between performance and efficiency in its P-cores.

Fastest integrated graphics for thin and light PCs?
Does Intel’s Xe2 GPU Architecture Claim the Title of Fastest Integrated Graphics for Thin and Light PCs? Indeed, it does. The new Intel Xe2 GPU architecture boasts the fastest performance numbers for integrated GPUs in thin and light PCs. The Intel Arc 140V GPU features 8 cores with a maximum clock speed of 2.05 GHz, while the lower variant, the Arc 130V GPU, has 7 cores and reaches a maximum of 1.85 GHz. Intel claims an impressive 31% overall performance increase compared to the previous generation Xe1, which powered Meteor Lake-based laptops. During the performance briefing session, Intel showcased a gaming performance chart demonstrating thin and light laptops running on Intel Lunar Lake surpassing 60 FPS in numerous popular gaming titles at 1080p medium settings, with some games like Counter Strike 2, DOTA 2, and GTA V exceeding 120 FPS at 1080p medium settings. Intel didn’t hesitate to challenge Qualcomm’s gaming performance, claiming a significant 68% lead over Qualcomm Snapdragon X-Elite powered laptops at the same resolution and settings. Additionally, Intel took aim at its rival AMD, asserting an average lead of 16% over AMD when comparing the AMD Ryzen HX 370-based laptop to an Intel Core Ultra 9 288V-powered laptop. While Intel’s dominance over Qualcomm in gaming performance is unquestionable, the comparison with AMD Ryzen HX 370 is more nuanced, as both Intel and AMD will likely exchange blows depending on the games used for benchmarking. We eagerly await the opportunity to test the gaming capabilities of a Lunar Lake laptop.

Also read: Sneak Peek: Intel’s Robert Hallock on What to Expect from Lunar Lake

Staggering batterylife numbers
We checked out a few laptops at the demo zone and the common narrative across the show floor was “Highest battery life ever”. OEMs were eager to highlight the impressive battery life of their devices, and the claimed numbers were nothing short of staggering. For instance, the Lunar Lake-powered Dell XPS 13 boasted an astonishing up to 26 hours of battery life. Intel’s own rigorous benchmarking using the UL Procyon Office Productivity test claimed an impressive 20.1 hours of tested battery life for an Intel Core Ultra 7 268V-powered laptop, surpassing the 18.4 hours achieved by a Qualcomm Snapdragon X-Elite powered machine. Notably, Intel ensured a level playing field by using identical chassis and battery pack capacities for both devices.

While battery life claims are not uncommon, it’s essential to consider the specific configurations, testing conditions, and workloads being used. However, Intel’s demonstration of superior battery life across various aspects, including graphics, processor, and battery life, suggests a strong readiness to challenge Qualcomm and AMD in this critical area. Intel’s consistent comparisons between Lunar Lake and Snapdragon Elite systems underscore its confidence in the platform’s capabilities. With Lunar Lake, Intel seems poised to shatter the myth that X86 PCs can’t deliver exceptional battery life.

AI performance
Before delving into the NPU improvement numbers, Intel emphasized the importance of a comprehensive AI PC framework. Intel asserts that while NPU performance is crucial, it’s not the sole determining factor for all AI workloads. Many AI tasks rely on the CPU and, more significantly, the GPU for execution. We concur with Intel’s stance on the vital role of a capable and power-efficient GPU, particularly in generative AI models where output time is more critical than power consumption. This explains Intel’s focus on its graphics engine in Lunar Lake. According to Intel, 40% of AI application developers prioritize GPU optimization for AI tasks.

Lunar Lake-powered laptops can achieve up to 120 TOPS of combined AI performance, with the GPU reaching 67 TOPS, followed by the NPU at up to 48 TOPS, and the CPU at up to 5 TOPS. While TOPS is currently the standard measurement metric for AI performance, Intel notes that it may not be the most optimal or realistic measure. As AI performance and measurement benchmarks evolve, TOPS may become less relevant. Intel compared Lunar Lake performance numbers across UL Procyon AI benchmarks and Geekbench ML benchmarks to demonstrate its lead over Qualcomm and AMD. We look forward to testing our review unit to determine the true leader in AI benchmarks. By adopting a holistic approach to AI performance, Intel is poised to revolutionize the industry’s understanding of AI capabilities.

Shots fired!
The story is clear: Intel is pulling out all the stops to regain its crown in the space, particularly in terms of efficiency and AI performance. The goal is to shatter the myth that X86 PCs are incapable of delivering excellent battery life. If the numbers are to be believed, they may be well on their way to achieving this. However, this is not just a battle between Qualcomm or AMD and Intel; it’s a resurgence of the classic ARM vs X86 rivalry. We saw AMD emphasize performance during the launch of the AMD Ryzen AI 300 series in July, and now Intel is seeking to lead and define the narrative around power efficiency and performance per watt across CPU, GPU, and NPU. We will reserve our final judgment until we receive our review units based on the Intel Lunar Lake platform.

What can we expect from Intel’s Lunar Lake? We saw Meteor Lake back in January this year, which was the first Intel platform to introduce the NPU story. Now, we’re ready with Lunar Lake, and the question is, how big will it be? 

I’ll relate the Lunar Lake launch to Meteor Lake as a reference point. On graphics, you can expect a 50% increase in performance, and we’ve reduced SoC power by about 40%, which can vary depending on usage. AI is, of course, the talk of the hour for everyone. Depending on the engine, NPU performance is about 4 times faster, and graphics performance is roughly the same from an AI processing point of view. It’s a monolithic product with two dies – one for the chipset and one for compute. This makes it look like a more conventional Intel CPU. We’ve overhauled cores, graphics, and the AI engine to launch a much more efficient chip. When we launch in September, we’ll have leadership in battery life, graphics, and AI.

On graphics, you can expect a 50% increase in performance, and we’ve reduced SoC power by about 40%

Lunar Lake is doing a few things differently. We’re looking at 8 cores across the lineup, and Intel has dropped hyperthreading, which is big news since we haven’t seen Intel drop hyperthreading in over a decade. What’s the approach here? 

It goes back to when CPUs had fewer cores, and getting another core to increase multithreading performance was physically impossible due to transistor size and wafer size limitations. That’s when SMT stepped in to recognize that execution pipelines in CPUs have bubbles or pauses where new information can be inserted. However, duplicating transistors at the front end of the chip costs power and increases costs. A processing thread served by a physical core is always faster than one served by a logical SMT core. With the architectural changes we’ve made in the Skymont E-cores and Lion Cove P-cores, we’ve improved performance per watt by removing SMT. So, it’s a net benefit overall across performance and power to remove it on a product like Lunar Lake. However, SMT will probably return in other places across the roadmap.

Also read: How Intel Lunar Lake chip aims to power new age of AI PCs at scale

With the architectural changes we’ve made in the Skymont E-cores and Lion Cove P-cores, we’ve improved performance per watt by removing SMT

Another big change with Lunar Lake is the introduction of Memory on Package (MOP). What are the trade-offs? It will help with performance, lower latency, and improved efficiency, but it also makes laptops non-upgradeable when it comes to memory. Will that be a big factor? 

I agree it’s a nuanced issue for enthusiasts like me who love upgradable laptops. However, you could potentially add hours of battery life to the system by not having upgradeable memory. The power costs of talking to off-chip memory are one of the larger power costs outside of running the core. Moving memory on-package was the right call for maximizing energy efficiency and battery life, which was the overarching goal of the program. However, it might not necessarily be the choice we make for upcoming products, as newer technologies will allow us to make the same power savings elsewhere. So, MOP is a choice for Lunar Lake but not necessarily for our upcoming products.

So, MOP is a choice for Lunar Lake but not necessarily for our upcoming products.

Speaking about power consumption, efficiency, and battery life, when it comes to AI performance, TOPS seems to be the new battleground, and the latest metric, as per Qualcomm, seems to be TOPS per watt measurement. What are your thoughts on TOPS as a benchmark for measuring AI performance? 

I think TOPS are good and bad. They allow for an objective and factual discussion about the theoretical peak performance of the AI engine, just like GPUs have teraflops. However, we don’t necessarily talk about GPU teraflops anymore because people realized that teraflops are not the most accurate metric. History is littered with examples of GPUs with lower teraflops beating those with higher teraflops in real-world performance due to better optimizations, drivers, and software. I believe AI is ultimately headed in the same direction. The software stack for AI is very similar to gaming graphics, with AI models instead of game engines, AI-based applications instead of games, and runtimes and APIs similar to DirectX. So, I believe that over the coming months, the benchmark ecosystem will mature to a point where we can discuss performance pros and cons of AI engines and decide from there. What we like to talk about internally at Intel is performance per TOP. I have some recent measurements on this, where we took the flagship AMD Hawk Point, the Ryzen 7 8840u, vs the Meteor Lake Core Ultra 7 155H. The Ryzen 7 has 25% more TOPS than Intel Core Ultra 7, but on average, across 50 different workloads tested, we saw Meteor Lake coming on top with an average 13% better performance. For me, that’s where the conversation around TOPS starts to break down. On paper, Hawk Point should be better, but because Intel’s software optimization is better, we can produce more performance per TOP in the engine. As far as performance per watt is concerned, of course, it matters and it matters across the board, be it CPU or GPU. However, the advantage we have is having the largest network of software developers, and we talk to them formally. We engage with them to know their long-term plans on what AI software they are working on and the engines they want to use. And, only 30% want to use the NPU, the other 35% want to use graphics, and the rest still want to use the CPU. The reality is, the software developers in the industry are not overly focused on the NPU. Which is why Lunar Lake has these XMX instructions to run AI workloads on the GPU. Our largest content creation accounts will only use graphics as that offers the best performance. So, it’s not that NPU is unimportant, but at the moment, it’s about just 1/3rd of what we need to care about when it comes to AI compute. And that’s where we disagree very much with Qualcomm. By over-indexing on the NPU, you might be under-serving the performance of the workloads that will run on the GPU and CPU engines. And it will ultimately lead to poor outcomes for users. And that’s why we’ve put strong AI on all three accelerators on Lunar Lake.

By over-indexing on the NPU, you might be under-serving the performance of the workloads that will run on the GPU and CPU engines. And it will ultimately lead to poor outcomes for users. And that’s why we’ve put strong AI on all three accelerators on Lunar Lake.

What’s your thought when it comes to ARM-based products entering the Windows Ecosystem, and where do you think X86 has an advantage over ARM-based products? 

I think we are about to prove that X86 can certainly have a power advantage over ARM. That is one thing that I am really excited to talk about when we get to the Lunar Lake launch. The real difference maker is the compatibility of X86. Almost every piece of software you have ever touched has X86 at its roots. ARM has this huge hurdle in front of it; it’s got to catch up on 60 years of compatibility. And some of that will probably never move away from the performance cost of emulation. And that’s just the hard truth of trying to overturn 60 years of R&D and coming in at the last minute with something different. Having said that, I do think that Snapdragon is a good offering, the team has done a good job with it. But Lunar Lake is a better solution. Because of the compatibility, the power, and the graphics performance, you get a very well-rounded package that the other guys are still fighting for. And they may never get there, or they will take an extraordinarily long time to get there. You don’t have to wait when you choose the Intel CPU or, for that matter, an AMD CPU. That’s the beauty of X86; it just works well by default.

You don’t have to wait when you choose the Intel CPU or, for that matter, an AMD CPU. That’s the beauty of X86; it just works well by default.

Also read: Intel Lunar Lake AI Performance Preview

We will be there at the launch of Intel Lunar Lake in Berlin, giving you our first impressions. We can’t wait to get our hands on the Lunar Lake laptops from all the major laptop brands. Stay tuned for more coverage on Intel Lunar Lake in the coming week.

CIOs and CTOs are looking at LLMs with a cautious lens. It’s not about the upfront investment or the ROI; it’s about Gen AI falling short of expectations; the hype vs reality gap is much wider. What’s the gap? What’s missing? Gen AI is smart and conversational, but it’s still missing the “intelligence” bit. It still depends heavily on the level and depth of training that the ML algorithms go through and the quality of structured data that is fed upon. Where is human-like intelligence? It might be around the corner.

The rumours surrounding OpenAI’s Strawberry project have sparked interesting discussions about the future of AI. The big rumoured update with Strawberry is the arrival of reasoning skills.  As we continue to push the boundaries of what’s possible with artificial intelligence, it’s worth revisiting the question: “Can machines think?”

What if Strawberry changes everything? What if this new model, rumoured to possess advanced reasoning capabilities, bridges the gap between human and artificial intelligence? The implications would be profound, far-reaching, and potentially transformative.

As I eagerly await the unveiling of Strawberry, I’m looking back at the tremendous progress we’ve made in AI. From the early days of rule-based systems to the current crop of large language models, we’ve witnessed significant strides in machine learning and natural language processing. Yet, despite these advances, we’ve still fallen short of true human-like intelligence.

Strawberry could be the game-changer we’ve been waiting for. By enabling machines to think critically, make decisions, and solve complex problems, we may finally be able to answer Turing’s question in the affirmative. But what then? Would we be comfortable with machines that can think, feel, and behave like us? The ethical implications are daunting, to say the least.

Also Read: Battleground AI: The Great Chip War Heats Up

Perhaps, though, I’m getting ahead of myself. Maybe Strawberry will turn out to be a small step towards AI reasoning rather than the revolutionary leap I’m anticipating. Maybe my excitement – and apprehension – is premature. But even if that’s the case, the fact that we’re having this conversation and pushing the boundaries of what’s possible with AI is fascinating and a bit unnerving at the same time.

The pursuit of AGI is a double-edged sword. We want machines to approach human intelligence to harness its full potential, but we’re cautious about the unknown repercussions. As we inch closer to AGI, we’re torn between excitement and apprehension. Will another step forward unleash a Pandora’s box of uncontrollable consequences or open up new possibilities? Critics warn of autonomous entities surpassing human intelligence, displacing workers, and eroding our sense of purpose. We’re hesitant to face the unknown, yet we can’t resist the allure of AGI’s promise. This conundrum leaves us questioning whether we’re ready for the future we’re creating. Either way, ready or not, the race for AI supremacy amongst the tech giants is leading us closer to AGI. In my opinion, the sooner we prepare ourselves, the easier it will be to internalize it and accept the change.

So, I ask you: are you nervous? Are you excited? Are you sceptical? What’s your reaction to large language models like ChatGPT moving towards AGI, logical reasoning, and human-like behaviour? Do you see a future where machines and humans coexist, collaborate, and perhaps even merge? Or do you see a more ominous outcome, where we lose control of our creations and face unforeseen consequences?

At the centre of this battle is the Neural Processing Unit (NPU), a specialized chip designed to handle AI workloads efficiently. AMD is claiming the highest-performing NPU and integrated GPU with its Ryzen AI 300 series, while Qualcomm is focusing on efficiency and battery life with its ARM-based Snapdragon X Elite SoC. Intel, meanwhile, is lining up with its Lunar Lake launch, expected to deliver the highest performance along with great efficiency.

However, amidst the excitement, a cloud of uncertainty hangs over the AI PC market. Microsoft’s Windows Recall feature, touted as a game-changer for AI PCs, has been delayed, and privacy concerns surrounding the feature have raised eyebrows. The Recall feature, which logs and records everything you do on your laptop, allowing you to search across time to find the content you need, has sparked concerns about data security and privacy.

Consumers are taking a wait-and-see approach, hesitant to invest in a technology that is still evolving and surrounded by uncertainty.

This delay and the surrounding controversy have slowed down the urgency of upgrading to an AI-ready laptop with an NPU. Consumers are taking a wait-and-see approach, hesitant to invest in a technology that is still evolving and surrounded by uncertainty.

AMD’s Balanced Approach

AMD’s Ryzen AI 300 series is a powerful contender in the AI PC market. Its powerful integrated GPU and a 50 TOP NPU offer a balanced approach that delivers high performance at great efficiency. AMD’s focus on delivering performance across all three frontiers makes it a top contender when it comes to performance-centric workloads. However, it’s not as power efficient as the ARM-based Qualcomm SoC and hence it will have to strike that balance with laptop designs and configurations with partners to ensure it can compete when it comes to batterylife.

AMD’s focus on delivering performance across all three frontiers makes it a top contender when it comes to performance-centric workloads

AMD’s strength lies not only in its hardware but also in its software ecosystem. With its AMD XDNA2 NPU, AMD is providing developers with a robust platform to build AI applications. The company’s commitment to open standards and collaboration with industry partners ensures that its technology is widely adopted and supported.

Qualcomm’s Efficiency Crown in AI compute

Qualcomm’s Snapdragon X Elite SoC is a game-changer when it comes to efficiency. Its ARM-based design and optimized NPU deliver unparalleled battery life, making it an attractive option for consumers who value portability and convenience. Qualcomm’s focus on TOPS per watt is a clever move, as it shifts the narrative from absolute performance to efficient performance.

Qualcomm’s Snapdragon X Elite SoC is a game-changer when it comes to efficiency

But Qualcomm’s strength lies not only in its hardware but also in its deep understanding of mobile devices. With its extensive experience in designing mobile processors, Qualcomm is well-positioned to deliver AI PCs that are optimized for mobile use cases. Its partnership with Microsoft to deliver the CoPilot+ PC is a testament to its commitment to innovation and collaboration.

Also Read: WazirX hack: Confusing aftermath of the biggest cyberattack on Indian crypto exchange

Intel’s upcoming Lunar Lake looks promising

Intel’s Lunar Lake launch is highly anticipated, and for good reason. With its focus on pure performance and great efficiency, Intel is poised to deliver AI PCs that meet the demands of even the most intensive workloads. Intel’s strength lies in its deep partnerships with PC manufacturers and software development houses, ensuring that its technology is widely adopted and supported.

Intel is poised to deliver AI PCs that meet the demands of even the most intensive workloads. Intel’s strength lies in its deep partnerships with PC manufacturers and software development houses

However, Intel’s challenge lies in its ability to adapt to the changing landscape. With AMD and Qualcomm nipping at its heels, Intel must innovate and differentiate its offerings to stay ahead. Its focus on optimized performance is a step in the right direction, but it must also address concerns around efficiency and battery life.

The Future of AI PCs

As the AI PC market continues to evolve, one thing is clear: the future belongs to those who can deliver efficient, high-performance AI capabilities. With AMD, Qualcomm, and Intel each bringing their unique strengths to the table, the consumer stands to benefit.

Beyond the chip giants’ rivalry, the AI PC market is missing a crucial piece of the puzzle: a killer AI application that can drive demand for laptops with powerful and efficient Neural Processing Units (NPUs). All the current-day NPU-optimized applications that I have tested and seen demos of are creator-centric applications or niche use cases that miss the mass appeal. The Windows 11 Recall feature was poised to be the catalyst, but its potential has been hindered by pressing privacy and data security concerns. To move forward, industry collaboration is essential to establish robust standards and best practices, addressing these concerns and unlocking the full potential of AI PCs.

As the AI Chip Wars unfold, I’m left wondering: what’s your take on this revolution? Are you eager to upgrade to an AI-equipped PC, or are you content with the status quo? Do you see on-device AI processing as a game-changer, or are you comfortable relying on cloud-based compute? And what about data privacy – is it a concern that keeps you up at night, or do you trust the industry to get it right? I’d love to hear your thoughts on this brave new world of AI PCs. Drop me a line at soham@digit.in, and let’s continue the conversation. Your voice matters in this debate, and I’m eager to hear your perspective on the AI PC war that’s knocking on our doors.

Digit: We’ve tested Qualcomm Snapdragon X Elite-powered laptops and compared them to Intel Meteorlake and AMD Zen 5-based laptops; the Snapdragon Elite surely takes the crown when it comes to battery life and efficiency, but when it comes to outright GPU performance, there is a significant gap. Is it that the games and GPU-intensive apps are not able to take advantage of the Adreno GPU, or is it the Prism emulation on Windows that’s holding it back? 

Kedar: It depends on what you are testing with games and how we look at platform definition. We are very focused on optimizing apps and games. I don’t want to mislead the community that this is a gaming platform. At the same time, we have more than 1200 games optimized on the mobile side and we have elite gaming, we bring the legacy of partnerships with net engine and game developers. But at the same time, once you position this as a gaming PC, the expectations from gamers are very different. We are being careful that we don’t overposition the SD Elite PC as being a gaming PC, but we are optimizing gaming performance as well. You should look at the overall gaming experience that you get with the device. 

We are focussed on TOPS per watt. We are looking more at that benchmark than just the TOPS number alone

Kedar Kondap, Sr. Vice President & General Manager, Compute and Gaming at Qualcomm

Digit: So is the Snapdragon X Elite promising more efficiency rather than outright performance? Will we see a beefier version of the X Elite that is targeted more at performance rather than efficiency? 

Kedar: This is the beauty and the versatility of the silicon we provide. Unlike our competition which has multiple series like U, H, etc with varying TDPs. We offer the versatility of giving our customers the choice of what they want to do. We give them a piece of silicon and they can design the laptop with a TDP of their choice. You could design a 12-15W TDP device or you could choose to design the laptop with 55 or 80W TDP and harness the performance of the X Elite. Some devices have launched at 35-40, and you can go higher on the TDP. It’s not commercially launched yet but we did showcase some high TDP devices at the Snapdragon Summit last year. Overall our goal is to scale fast, as developers and apps are going at a rapid pace. We want to make sure that as many consumers can get access to it. There’s a big announcement planned at IFA 2024 in Berlin and you will know more about how we plan to scale and make Snapdragon X Elite powered PCs more accessible to a wider audience. 

Digit: While internationally Snapdragon X Elite based laptops made it to the shelves some time back, in India we are now starting to see launches by all key brands. If you were to market X Elite laptops to laptop buyers, what is that big promise and advantage that SnapdragonX  Elite powered laptops bring to the table? 

Kedar: It is performance across three key vectors, Performance, Power, and AI. From the performance standpoint, you’ve already run the benchmarks and you’ve seen how snappy the performance is. Also, as you start using real-world applications you will experience the crisp and responsive performance of Snapdragon X Elite-powered laptops. When it comes to Power or battery life, you have tested it out for yourself, it is amazing. I use the device full-time and once you start using it, you will understand the difference as a Windows user. And the third is AI. We showed a bunch of applications and how the Apps can take advantage of our NPU. This is just the beginning. And it will bring the excitement back to Windows, we don’t think there’s been performance leadership in Windows and the Snapdragon X Elite will be bringing that leadership back again. 

Digit: A lot of performance leadership today is centered around the NPU. It’s a sweeping statement, but is NPU the new CPU from a focus point of view? 

Kedar: It evolved from the CPU decade to the GPU decade to I believe the NPU decade today. We invested a lot in the NPU development, we’ve had an NPU on the smartphone platform for several years. It’s just that the NPU is getting wider recognition now as models are evolving. So you have more than 100 models that we’ve optimized on the X Elite and once you start to showcase the benefits, you realize the advantages of the efficiency of the NPU. We deliberately picked intensive creator apps to showcase the benefits of NPU. It showcases the versatility of the NPU which leaves a lot of room for the CPU and GPU to run other tasks. You will see more and more applications getting ported to the NPU. 

Digit: The whole story has been about on-device AI, privacy, and the reason why you need a powerful NPU. There’s also a mix of on-device and private cloud computing, we’ve seen Apple announce its Apple Intelligence which combines on-device AI and secure cloud-based computing to augment AI experiences. How do you see AI computing moving? For someone being in the industry for a long time and witnessing the changing trends, where are we heading? 

Unlike our competition which has multiple series like U, H, etc with varying TDPs. We offer the versatility of giving our customers the choice of what they want to do.

Kedar Kondap, Sr. Vice President & General Manager, Compute and Gaming at Qualcomm

Kedar: It will surely be a combination. I think of it in three steps. You have on-device, we’ve shown how the X Elite with 13 billion parameter models. While your own AI Infranet or near-edge or private cloud, whatever we reference it as. You are going to have certain models that you can run in that environment. For example, instead of 13 billion, you can run up to 100 billion in a private cloud environment. But after that, you will go to the cloud. So, there’s always going to be three steps. And over time these numbers will change. On-device performance will scale higher, number will change. Accuracy will improve and you will have more NPU TOPs available on-device, for example, the 13 billion will become 30 billion or 50 billion, so on-device NPUs will become more powerful. So there will always be a hybrid model where certain use cases will run only on a device, certain in a closed infra environment, and certain will go only to the cloud. 

Also Read: ASUS Vivobook S15 OLED Ft. Snapdragon X Elite Review: Better Than Intel Core Ultra 7?

Digit: You just mentioned TOPS, which seems to be the key and only metric when it comes to measuring NPU performance and now we have two numbers being marketed, NPU TOPS and total system TOPS, that take into account the CPU, GPU, and the NPU. And that number is a lot higher than the standalone number for the NPU. Is TOPS the only way to measure and quantify AI performance? 

Kedar: It’s a very good question, and frankly, the answer is, that we don’t know yet to be honest. I don’t think there is a good metric yet that is defined. And it also doesn’t do justice to the performance per watt aspect when it comes to AI computing. Just because you have a higher number, doesn’t mean you have the most efficient NPU. You could be at worse power. So I don’t know if there is a metric yet and that is why we didn’t reference TOPS as much, but we focussed on TOPS per watt. We are looking more at that benchmark. There’s not an easy way to benchmark and compare them yet. And that’s why at Qualcomm we are more focused on driving experiences rather than just numbers. Even now I am very focused on driving experiences and not just pure specs. The reason why everybody is using TOPS is because there is no other benchmark or metric yet to measure performance. To be honest with you, I don’t know what the right metric is and we have started focussing on looking at TOPS per watt number. It’s very hard to tell you which is the right metric today. We are focused on making the consumer understand the kind of superior experience you get from a Snapdragon X Elite powered PC. 

It has evolved from the CPU decade to the GPU decade to I believe what is the NPU decade today.

Kedar Kondap, Sr. Vice President & General Manager, Compute and Gaming at Qualcomm

Digit: Talking about the experience, what’s the reason for a consumer to pick an AI PC today? We’ve seen a delay in the Windows Recall feature which was touted to be this feature that will make everyone get a CoPilot+ PC. So, what’s the reason to upgrade? 

Kedar: I think the question today is if you are buying a PC, as the apps are evolving fast, you want to buy a future-proof PC. Today you don’t want to buy a PC that doesn’t have an NPU. You are hearing all the apps and app developers like Adobe have talked about moving to the NPU. You want to buy a PC where it is ready for the shift to NPU from a software and application point of view. And the developments are super fast, we are talking days and weeks and not years. It’s all moving at such a rapid pace and ISVs are building and optimizing the applications to take advantage of the NPU. We trying to make it simpler for developers with all the tools, we’ve announced a developer kit, and we have the AI hub. So, we are enabling developers with access to hardware and software to accelerate the process. 

AMD’s Zen 5 core, powering the upcoming Ryzen 9000-series CPUs and Ryzen AI 300-series processors, boasts a 16% average increase in instructions per clock (IPC) on the back of architectural enhancements and a node shrink to TSMC’s 4nm process technology. The Zen 5 architecture boasts significant improvements in performance and efficiency, thanks to a redesigned front end, wider execution pipelines, and enhanced cache hierarchy.

How is AMD Zen 5 getting those performance gains?

The new front-end changes include a revamped fetch, decode, and dispatch mechanism that yields more instructions to the back end. Additionally, branch prediction accuracy has been improved, with more predictions per cycle. The dispatch and execution engine has been enhanced to support up to 8 instructions per cycle, with a unified scheduler and expanded execution window.

The cache hierarchy has also been improved, with a 50% increase in data cache size to 48 kilobytes, without increasing latency. Furthermore, the data bandwidth has been doubled, with support for 512-bit loads and stores. The AVX math unit has been doubled to a physical 512-bit wide pipeline, resulting in reduced latency and increased throughput.

What about the latency?

According to AMD, the Zen 5 architecture boasts several enhancements that collectively reduce latency and boost performance. Improved branch prediction accuracy and increased predictions per cycle minimize latency while dual-porting and decreased latency in caches enable more efficient access. Wider execution pipelines allow for simultaneous execution of more instructions, reducing latency.

Additionally, doubled data bandwidth enables faster transfer of data between caches, and an optimized scheduling algorithm minimizes stalls and improves resource utilization. Finally, tuned data prefetch algorithms recognize stride patterns, prefetching data more efficiently and reducing latency. These advancements combined deliver significant performance gains and reduced latency, making Zen 5 a formidable processor architecture.

Mark Papermaster, EVP and CTO at AMD, on Zen 5

“Zen 5 will not disappoint you with the kind of performance improvements that we’ve brought, and we won’t let up going forward,” said Mark Papermaster, EVP and CTO at AMD. “It’s going to be a pedestal that we’re going to build upon the next several generations of Zen,” he said. Mark Papermaster further emphasised, “We’re very happy with the improvements that we made. We could not be more proud with our new math acceleration unit.” Mark Papermaster reiterated their commitment to innovation, stating, “We will not slow down on our innovations at AMD.”

We’ll continue to deliver leadership and bring delightful experiences to our customers.” He added, “We think broadly with our IP designs in terms of how we deploy it, in terms of how we optimize it to be both high performant and highly efficient, and nothing has changed at all in that philosophy and that approach with Zen 5.” Zen 5 is poised to “absolutely excel at AI data preparation and general-purpose CPU workloads.” Mark Papermaster affirmed, “We’re maniacally focused on execution and delivering new capabilities on a regular cadence”, Mark affirms.

Will the iGPU performance be the deciding factor?

As AMD’s Zen 5 architecture gears up to take on Intel’s Lunar Lake and Qualcomm’s ARM-based Snapdragon X-Elite chips, the burning question is: how will these impressive specs translate to real-world performance? The answer lies in the harmonious balance between the Zen 5 core, XDNA 2 NPU, and revamped RDNA 3.5 iGPU. And as we’ve seen with the past few generations, increasingly the GPU is doing all the heavy lifting when it comes to AI workloads. And the focus and reliance on GPU performance is only going to get heavier as we move further into the generative AI era.

Benchmark leaks suggest that RDNA 3.5 iGPU is a major step up in integrated graphics performance, possibly rivaling some entry-level discrete GPUs. We’re eager to review the ASUS ZenBook S16, which we previewed at AMD Tech Day 2024. Our next article will dive deeper into RDNA 3.5’s enhancements.

Also read: AMD launches 8700F and 8400F Zen 4 CPUs without RDNA 3 integrated graphics