A few years ago, my neighborhood went through an extended blackout, and I had to take a few ice cold showers during the cold weather, and that reminded me that I shouldn’t take hot water for granted.

Yet, most people don’t spend much time thinking about their hot water heaters. And when they break down, most of us just replace them with the same type and move on. But heating up that water is expensive—typically close to 20% of your home’s total energy footprint goes into hot water, and it’s usually the second-largest energy-related expense. That’s why you should consider a heat pump water heater—if you can make it work.

How does a heat pump water heater work?

Where traditional water heaters use gas or electricity to directly heat water, a heat pump water heater works like a reverse refrigerator, pulling warm air over hot condenser coils that are either wrapped around a water tank or that go through the tank directly. The heat is transferred to the water, and the result is the same: A constant supply of hot water you can access any time you need it.

There are a lot of benefits that come with a heat pump water heater:

• More efficient. It’s estimated that heat pump water heaters can reduce your energy costs by as much as 75%. Considering that heating water is your second-largest energy expense, that can mean significant savings over time.

• Environmentally friendly. Heat pump water heaters are also pretty clean in terms of pollution. It’s estimated they can reduce the amount of carbon your house contributes to the atmosphere by as much as 75%.

• Tax Incentives and rebates. You can get a tax rebate of up to 30% of the installation cost to a maximum of $2,000 if you install a heat pump water heater, and many states also offer incentives. That means you can get an immediate financial benefit before the efficiency savings kick in.

Downsides to a heat pump water heater

You might be wondering why everyone doesn’t install a heat pump water heater, considering their advantages. There are some downsides:

• Cost. Heat pump water heaters are typically more expensive than traditional water heaters, with total costs as high as $5,000—on par with a tankless water heater’s cost. This can be offset by tax rebates and the energy savings over time, but represents an immediate obstacle for many people.

• Climate. Heat pump water heaters need ambient heat to work—they’re generally designed to work best in temperatures between 40 and 120 degrees Fahrenheit. If you want to install them outdoors (which many people do, because of their size—see below) you’ll need to be located in an area where it doesn’t get too cold, or you might find yourself without sufficient hot water, or with lukewarm water at best.

  It should be noted that if the heat pump water heater is installed in a space where ambient heat is produced—alongside a boiler, for example—it can actually use the waste heat from the other appliances to work. This might take extra planning, however. You can also find hybrid models that include a traditional heating element as a backup that kicks in when the temperatures get too low.

• Space. Heat pump water heaters tend to be pretty big, and they need a lot of air circulation to work properly, which is why they’re often installed outdoors. They can be installed indoors, but it’s generally suggested that they be in a room at least 450 cubic feet in size in order to work at peak efficiency.

• Condensation. Because of how they operate, heat pump water heaters cool down the space where they’re located, pulling heat into their water tank. As a result, there can be a lot of condensation which will need to be drained. That might require a drain be installed if there isn’t one already in place; otherwise, you might need to install a pump to remove the extra water (which can be used elsewhere—to water a garden, for example).

Google has announced a trio of new AI-powered shopping features. The tech company is rolling out more conversational search capabilities for shopping in AI Mode, so that buyers can make more specific and descriptive queries to find exactly what they’re looking for. The AI Mode results will also be formatted to match the type of search, delivering shoppable images for prompts with a more visual angle or side-by-side comparisons of different traits for shoppers considering several options of what to buy. 

There are also two new agentic shopping features. First is an agentic checkout for finding and buying an item at a particular price. Shoppers can use Google’s price-tracking feature to set the sale price they want to pay for a particular product, and can opt to allow Google’s AI to purchase the item when it hits the desired discount. Users will have to confirm the purchase and shipping info before the agentic checkout purchase is completed via Google Pay. For now, agentic checkout is limited to US shoppers at select online retailers including Wayfair, Chewy, Quince and certain Shopify stores; Google said it plans to support more sellers in the future. (Presumably Amazon will not be one of them.)

Finally, once you know what you want to purchase, Gemini can call local businesses to ask about product availability. The agentic AI can gather information about prices, stock and current promotions, ask relevant follow-up questions, then will email the results to the user. The first wave of categories that will support AI phone calling are toys, health and beauty, and electronics, and it will be available to US Search users. 

Google has been rapidly expanding its shopping tools to include AI integrations, with new resources such as virtual clothing try-ons and shoe shopping.

This article originally appeared on Engadget at https://www.engadget.com/ai/google-adds-agentic-ai-checkout-to-shopping-140000930.html?src=rss

Hideo Kojima has unveiled Death Stranding Isolations (working title), an original anime series based on his franchise, coming to Disney+ in 2027. The gaming legend made the announcement at a Disney+ event in Hong Kong alongside series director Sano Takayuki. It will be the first Kojima Productions’ partnership with a global streaming platform, with Kojima serving as an executive producer.  

The series will tell a new story set within the Death Stranding universe and not draw from any of the games’ narratives. All we know so far is that it “will follow a young man and woman as they set out on an adventure” and that traditional hand-drawn 2D animation will be used. Concept art displayed at the event was created by Ghost in the Shell: SAC_2045 character designer, Ilya Kuvshinov. 

Death Stranding (the game) originally launched in 2019 starring Norman Reedus, Mads Mikkelsen and Léa Seydou. The franchise finally got a sequel earlier this year with Death Stranding 2: On the Beach.

In 2023, Kojima revealed that production company A24 was collaborating with Kojima Productions on a Death Stranding live action film project. At the time, he said that “I aim to present a new dimension of cinematic adaptation that goes beyond merely turning a game into a live-action movie.” That film will be directed by Michael Sarnoski (A Quiet Place: Day One).

Live action series and movies based on games seem to be everywhere of late, but the number of animated productions is also growing. Many of those are on Netflix, including Splinter Cell: Deathwatch, Arcane, Castlevania and Tomb Raider: The Legend of Lara Croft. Death Stranding Isolations (working title), appears to be the first gaming IP-based animation series for Disney+. 

This article originally appeared on Engadget at https://www.engadget.com/entertainment/streaming/a-new-death-stranding-series-is-coming-to-disney-in-2027-133018605.html?src=rss

Spotify is testing a new feature that could make it easier to pick up an audiobook that you’ve ignored a little too long. The streaming service is rolling out recaps for audiobooks that can summarize the story up to where a user stopped listening. Recaps will be available once a user has gotten 15 to 20 minutes into the audiobook and will be updated to reflect the point the listener has reached without spoiling anything yet to come. It could help listeners to finish audiobooks they haven’t revisited in awhile without needed to restart from the beginning, and it might be handy just to refresh your memory on the plot. 

This is an AI-powered feature, but Spotify noted that it will not be training any large language models on authors’ content or developing voice generation from narrators’ performances. For now, recaps are only available on Spotify’s iOS app for a select number of English-language titles. The company said it plans to add recaps for more content in the future.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/spotify-is-adding-recaps-for-audiobooks-133000324.html?src=rss

Amazon has early Black Friday savings on its most potent streaming stick. The Fire TV Stick 4K Max is 41 percent off, bringing it close to its record low. You can get it now for $35.

The Fire TV Stick 4K Max is one of Engadget’s favorite streaming devices. The stick supports a solid mix of advanced technologies for its price: Dolby Vision, Dolby Atmos, 4K and HDR10+. As long as the rest of your entertainment setup can handle it, you’ll get a high-quality picture and sound. It also supports Wi-Fi 6E, enabling better, faster connectivity with compatible routers.

The 4K Max has the fastest processor of any Amazon Fire TV Stick, so expect zippy navigation. It also supports Amazon’s Ambient Experience. This mode displays art (like on Samsung’s The Frame) while the device is in standby. It’s also a solid choice for gaming: It supports Xbox cloud streaming and works well as a retro game emulator, too.

The UI is where Amazon appears to be subsidizing the device’s low cost. Expect to see loads of Prime Video content promos, along with other ads. But for $35 (compared to its MSRP of $60), you may find it easier to justify that tradeoff.

Also on sale is Amazon’s Fire TV Stick HD, our pick for the best budget streaming stick. This model doesn’t support 4K; instead, it limits you to 1080p at 60 fps. At $18 for Black Friday, it’s certainly cheap. But if you have a 4K TV (or plan to soon), you may want to consider the slightly more expensive model.

This article originally appeared on Engadget at https://www.engadget.com/deals/amazon-black-friday-deals-include-the-fire-tv-stick-4k-max-for-only-35-132100009.html?src=rss

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

The PowerBoy 2-in-1 Retro Gaming Console and Fast-Charging Power Bank mixes the old with the new, and it’s currently on sale for $79.99 at StackSocial. At first glance, it appears to be a simple handheld game console from the early 2000s, complete with a 3-inch IPS display and a familiar D-pad layout. But hidden inside that nostalgic shell is a 5000mAh magnetic power bank capable of 22.5W fast charging. You can play 1,500 preloaded games, covering everything from side-scrollers to puzzle titles, or use it to recharge your phone during a long commute.

Performance-wise, the PowerBoy runs on a 900MHz ARM9 CPU with 64MB of RAM, which is modest but more than enough for retro titles. Games load instantly, controls are responsive, and the 320×320 full-color screen gives old sprites a surprisingly crisp look. If you enjoy tinkering, there’s room for more, as its TF card slot lets you expand storage and add your own ROMs, although you’ll need to supply your own card. Beyond gaming, it doubles as a mini media player that can handle MP4 videos, MP3s, photos, and even eBooks. The 5000mAh battery isn’t just for the console itself; it can charge your phone, earbuds, or smartwatch multiple times through its built-in multi-device cables, which are compatible with USB, Apple, Android, and Huawei devices. You can even charge three gadgets simultaneously, making it a surprisingly practical companion for travel or workdays.

The design is compact and lightweight, measuring approximately 6.3 inches in width and weighing under a pound, making it easy to carry. It’s not designed for modern gaming or heavy use, but that’s not its intended purpose. It’s for filling quiet moments, revisiting old favorites, or keeping your phone alive when the battery warning hits. The 12-hour playtime gives it plenty of staying power, and since it doesn’t rely on wifi, it’s a solid offline entertainment option. If you want something simple, nostalgic, and functional all in one, the PowerBoy makes a solid case for pocketing a little retro fun with real-world usefulness.

Secret Santa gift exchanges are supposed to be fun, but it’s easy to overthink it. You want a gift that feels thoughtful without being awkward, useful without being boring, and most importantly, affordable. The sweet spot is under $50, which is plenty to find something that fits your recipient’s personality. Whether you’re buying for a co-worker you only chat with at the coffee machine, a friend who already has everything or a cousin who never gives you ideas, there are clever options that will make them smile. These picks prove you don’t need to spend big to win at Secret Santa.

Best Secret Santa gift ideas

Check out the rest of our gift ideas here.

This article originally appeared on Engadget at https://www.engadget.com/the-best-secret-santa-gift-ideas-for-2025-affordable-gifts-from-lego-apple-yeti-and-others-130014081.html?src=rss

SMART goals are a helpful way to turn your ideas and to-dos into actionable plans for the near and far future, but they’re not always the best approach. If you work on a larger team or need to track highly ambitious, long-term goals, for example, SMART might not be the best method to turn to. Other goal-tracking methods, like FAST and PACT, may be a better fit. You’ll need to start by familiarizing yourself with these alternate methods, but once you do, you’ll see that they are more tailored to specific needs and might just be your best bet.

When should you use SMART goals?

SMART goals are specific, measurable, achievable, relevant, and time-bound—that’s what the acronym spells out. When you make them, you make sure your goal sticks to all five of those requirements, usually writing it out in a sentence like this: “By the end of the month, I will have aced three practice quizzes to prepare for my chemistry test.” It’s specific because it says what you’re going to do and why, measurable because you’re setting an amount of quizzes and expected grade on each, achievable because it can reasonably be done, relevant because it has to do with a pressing need, and time-bound because it’s related to a forthcoming deadline.

Because they’re so customizable, SMART goals are used by students, workers, and anyone who needs to break down their to-do list into manageable action items, but they can be limiting for the same reasons they work so well on an individual level. For example, they can be too specific to be broadly applicable, so if you have to do the same sort of task at work every month, you’ll redo the SMART goal every time instead of creating a process to follow. They’re also not ideal for collaboration because while you can share a SMART goal with others, it doesn’t leave much room to be concise and authoritative about who needs to do what to reach the goal. 

When FAST goals make sense

The first helpful SMART goal alternative is the FAST goal, which is especially useful if you’re in a position to lead or delegate, but is generally good for teams overall. FAST stands for the following: 

• Frequent discussions

• Ambitious scope

• Specific measurements

• Transparent

When using SMART goals, you analyze your desired end result through the lens of how achievable it is and when you’ll have it done. With a FAST goal, you’re looking less at the elements that define the final product and more at how you will collaborate to get there. FAST goals enable teams to adapt and evolve as the project goes on because they require frequent discussions and transparency about what the ambitious goal is and how its success will be defined. Still, you need to have a plan in place to follow the FAST goal, since it is a little vague. What are “frequent” discussions, for instance? You need to hammer that out with the team and set a goal dictating what frequent means and how you’ll conduct those meetings.

Say your team at work has to build a report for the end of the quarter. You already know when it’s due, so you don’t need to incorporate timeliness, like you would with a SMART goal. Instead, consider the task through the FAST lens, setting up regular times to meet and discuss it and creating clear communication channels so everyone can stay on the same page. The success of the project might be measured by manager feedback, client response, the enhancement of processes, or increases in sales; you need to define early on what “success” will look like, but staying communicative and transparent will help—and will enable you to make the end goal ambitious. 

Here, I’d recommend incorporating a single source of truth, or SSOT. That’s essentially a folder everyone has access to that includes every resource anyone might need. For the hypothetical project of the end-of-quarter report, your SSOT might include monthly reports, a template for the bigger report, contact information for clients whose data will appear on the report, and so on. The first document in the SSOT should be one outlining the FAST goal by setting requirements for the frequent discussions, detailing the scope, defining the specific measurements of success, and, by its nature, being transparent.

When to use PACT goals

PACT goals, like FAST goals, focus more on the process of achieving a desired result than the measurements of the result itself. Here’s what PACT stands for: 

• Purposeful

• Actionable

• Continuous

• Trackable

Like a SMART goal, PACT works well if you’re tackling something on your own, but it is more process-driven. Let’s say your goal is to get in shape. With a SMART goal, you might have to define it like, “By the first day of summer, I will have worked out for five hours per week, lost 20 pounds, and increased my muscle mass by 5%.” It’s specific, measurable, achievable, relevant, and time-bound, but once you write that down, you’re on your own fighting your way to the finish line. A PACT goal might look more like, “To get in better shape and improve my health, I will work out five times a week and monitor my weight loss and muscle gain using my smart scale.” Instead of being time-bound, this approach takes the process itself into consideration—but also incorporates purposefulness, reminding you why you’re doing what you’re doing. You still take actions and track metrics, but the goal is rooted in purpose and relies on continuity, not a defined end result. 

Keeping the purpose front and center is important, as it helps you stay motivated and reminds you of what you’re really working toward. If you’re building a report at work, you can get caught up in the absolutism of knowing this is your assigned task and you have to do it, which can lead to losing sight of why you’re actually toiling. If you bear in mind that you’re building the report so your company can bring in more clients or enhance internal operations, you remember that this could lead to more business, accolades, or even a raise for you. Tracking your progress is also key to staying motivated and moving toward an end goal, even if it isn’t as rigidly defined as a SMART goal might be. 

Again, you’ll need a document that outlines all this. Think of it like a mission statement and stick it in the SSOT. Here, it might be like, “Our team will compile the report by each dedicating one hour of time to it a day for the next month, inputting our work hours and achievements into the attached spreadsheet, and meeting every Friday to discuss progress.”

In anticipation of what’s expected to be a more hectic than usual season of holiday travel, Uber is adding a collection of new gifting and ridesharing features to its app. The most convenient of the new features is “Send a Ride,” a way to purchase Uber rides for someone else.

Send a Ride lets you cover the cost of a ride directly from the Uber app, making it possible to pay for something like an airport pickup in advance. Uber says you’ll be able to set a number of rides and a spending limit in the app (like two rides for up to $50 each) and then the app will generate a link that you can share with whomever you want to receive the rides. When they click the link, credit will automatically be added to their account and applied to their next trip.

For anyone whose winter travel includes skiing and snowboarding, Uber Ski is a new seasonal ride option Uber is adding that lets you book a ride to and from “nearly 40 of the top mountains across North America and Europe.” When you select Uber Ski in the app, you can reserve an UberXL for up to two guests with gear or an UberXXL for up to four guests with gear. Uber also says it will sell the Epic Pass, a ticket that gets you access to Vail Resorts, directly through the Uber app. 

If you’re figuring out your own ride home from the airport, Uber is also expanding its Uber Share for Airports option that lets you split a ride home with people heading in the same general direction. Uber Share is now available at over 50 airports worldwide, including new additions John F. Kennedy International Airport, LaGuardia AirPort and Orlando International Airport.

Finally, starting in December, Uber will let you add a free video message from stars like Megan Thee Stallion, the Jonas Brothers or Tracee Ellis Ross to any Uber Eats order you send as a gift. The videos don’t sound like they’ll be personalized like a video from Cameo, but whoever you send the gift to will be able to choose their own delivery time so they’re available when their gift arrives.

This article originally appeared on Engadget at https://www.engadget.com/transportation/ubers-send-a-ride-feature-makes-it-easier-to-pay-someone-elses-fare-130000110.html?src=rss

In August, Google announced that it will implement a new safety feature that would require developers to verify their identities if they want Android users to be able to sideload their apps. Now, the company has started inviting developers that distribute exclusively outside of the Play Store to the early access of the identity verification feature in Android Developer Console. Google has also revealed in the same announcement that despite its new rule, it will give experienced users the option to sideload even unverified apps on their Android devices.

The company said it received feedback from developers and power users who want to retain the ability to download unverified apps. That is why it’s now building “a new advanced flow that allows experienced users to accept the risks of installing software that isn’t verified.” Google didn’t delve into how it designed the feature and how it will determine if someone is a “power user,” but it’s already gathering feedback about it and will share more details in the coming months. It did say that it designed the flow to ensure users aren’t duped into bypassing safety checks by scammers, including showing them clear warnings about the risks involved.

As Google explained in its announcement, one common attack in Asia involves scammers calling victims and making them download malware disguised as legitimate applications. They pretend to be employees from a bank, warning victims that their account has been compromised and instructing them to sideload an app to secure their funds. The scammers would also pressure their targets to ignore security warnings while they’re sideloading the application. The malware in the bad actors’ app will then steal the victim’s login and intercept two-factor codes needed to access their bank account.

“While we have advanced safeguards and protections to detect and take down bad apps, without verification, bad actors can spin up new harmful apps instantly,” Google said. “It becomes an endless game of whack-a-mole. Verification changes the math by forcing them to use a real identity to distribute malware, making attacks significantly harder and more costly to scale.” It’s still early days for Google’s developer verification requirement, however, and it won’t be rolling out widely until late 2026.

This article originally appeared on Engadget at https://www.engadget.com/apps/google-will-allow-experienced-users-to-sideload-unverified-android-apps-130000130.html?src=rss

Iceland has formally classified the potential collapse of a major Atlantic Ocean current system a national security threat, warning that a disruption could trigger a modern-day ice age in Northern Europe and destabilize global weather systems. The move elevates the risk across government and enables it to strategize for worst-case scenarios. Reuters reports: The Atlantic Meridional Overturning Circulation, or AMOC, current brings warm water from the tropics northward toward the Arctic, and the flow of warm water helps keep Europe’s winters mild. But as warming temperatures speed the thaw of Arctic ice and cause meltwater from Greenland’s ice sheet to pour into the ocean, scientists warn the cold freshwater could disrupt the current’s flow.

A potential collapse of AMOC could trigger a modern-day ice age, with winter temperatures across Northern Europe plummeting to new cold extremes, bringing far more snow and ice. The AMOC has collapsed in the past – notably before the last Ice Age that ended about 12,000 years ago. “It is a direct threat to our national resilience and security,” Iceland Climate Minister Johann Pall Johannsson said by email. “(This) is the first time a specific climate-related phenomenon has been formally brought before the National Security Council as a potential existential threat.”

Elevation of the issue means Iceland’s ministries will be on alert and coordinating a response, Johannsson said. The government is assessing what further research and policies are needed, with work underway on a disaster preparedness policy. Risks being evaluated span a range of areas, from energy and food security to infrastructure and international transportation. “Sea ice could affect marine transport; extreme weather could severely affect our capabilities to maintain any agriculture and fisheries, which are central to our economy and food systems,” Johannsson said. “We cannot afford to wait for definitive, long-term research before acting.”

Read more of this story at Slashdot.

Here in Bend, where we appreciate both cutting-edge tech and practical value, the AI hardware market is producing some fascinating matchups. The entire industry has been defined by NVIDIA’s “shock and awe” approach: massive GPUs and eye-watering performance claims with price tags to match. NVIDIA’s DGX Spark, a desk-side “mini-supercomputer,” is a prime example. Priced at a hefty $4,000, it’s marketed as the essential tool for serious AI developers. But a new challenger has emerged, not from a server rack, but from the compact world of mini-PCs, and it’s showcasing how AMD is delivering staggering AI performance for a fraction of the cost.

Recent benchmarks have pitted the $4,000 DGX Spark against a new $2,199 mini-PC from GMKtec, the EVO-X2, powered by AMD’s monster APU, the Strix Halo (also known as the Ryzen AI Max+ 395). The results are nothing short of a market disruption. While the DGX Spark holds its own in raw, high-throughput processing, the AMD Strix Halo system demonstrated better real-time performance, with faster token generation and lower startup latency in several key large language model (LLM) tests. This isn’t just a minor win; it’s a profound shift in the value equation for AI development.

The Power of the APU: Why Strix Halo is a Game-Changer

The NVIDIA DGX Spark is a specialized tool, a $4,000 dev kit designed to perfectly mirror NVIDIA’s enterprise-grade data center environment, right down to its proprietary CUDA software stack and advanced networking. Its value is in the entire ecosystem, not just the chip, allowing a developer to test code that will later be scaled up on massive, multi-million-dollar NVIDIA servers.

But AMD’s Strix Halo, the heart of the new mini-PCs, plays a different game. It’s an APU (Accelerated Processing Unit) that integrates a powerful CPU (Zen 5), a high-performance integrated GPU (RDNA 3.5) and a dedicated NPU (Neural Processing Unit) with 50 TOPS of AI performance all on one chip. This is all connected via a unified memory architecture, that, in the GMKtec EVO-X2, can be configured with up to 128GB of shared LPDDR5x RAM.

For a developer running models locally, this unified memory is a massive advantage. It allows them to load and run enormous 70B+ parameter models directly in memory, a feat that would normally require a much more expensive discrete GPU with a limited VRAM pool. The Strix Halo isn’t just “competing” with the DGX Spark in real-time inference; in many common developer scenarios, it’s proving to be faster and more responsive, and costs nearly half as much.

The Target Audience: Who Needs What?

This comparison highlights a critical divergence in the market. These two products, despite similar AI capabilities, are for two very different users.

• The NVIDIA DGX Spark User: This is the Enterprise Developer. Their company is already committed to a multi-million dollar NVIDIA DGX infrastructure in its data center. Their primary goal is not raw performance-per-dollar; it’s ecosystem continuity. They need a desktop-sized box that runs the exact same CUDA-based software stack as the production servers, allowing them to prototype and validate code locally before pushing it to the cloud. For them, $4,000 is the price of admission to that seamless, proprietary workflow.
• The AMD Strix Halo User: This is the Mainstream Developer, Power User or Startup. This user wants the absolute best performance for the money. They are likely using open-source tools like AMD’s ROCm, which are rapidly gaining traction. They need a machine that can both run a massive 70B parameter LLM for development and function as a high-performance workstation for video editing or even AAA gaming. The Strix Halo-based mini-PC isn’t just an AI dev kit; it’s their entire computer, offering a level of versatility the DGX Spark (which runs on an ARM processor and can’t game) simply cannot touch.

The Rise of the Compact AI Workstation

The battle between the Strix Halo and the DGX Spark is the perfect showcase for one of the most exciting trends in computing: the rise of the compact AI workstation. For the past year, AI development has been largely centralized in the cloud, costing companies a fortune. Now, hardware has advanced to the point where powerful, billion-parameter models can be run locally.This shift is democratizing AI. Small businesses, startups and individual developers can now experiment and build complex AI applications without renting expensive cloud instances. This is fueling a boom in the mini-PC market, with manufacturers like GMKtec, Minisforum, and Framework all racing to put these powerful APUs into tiny, desk-friendly boxes. These devices are the “AI PCs” that professionals actually want, offering massive memory and performance in a quiet, efficient, affordable package.

NVIDIA: Ill-Prepared for a Value-Driven Market

This new market dynamic leaves NVIDIA looking surprisingly ill-prepared. NVIDIA’s entire business model is built on selling high-margin, proprietary hardware and a locked-in software (CUDA) ecosystem. At $4,000, the DGX Spark feels like an extension of that high-margin strategy, aimed at protecting NVIDIA’s data center dominance.

NVIDIA doesn’t seem to have a good answer for the developer who doesn’t want to pay $4,000 for a dev kit, especially when a $2,200 AMD machine offers better real-time performance and plays Cyberpunk 2077 at 1080p (Strix Halo gaming benchmarks). AMD is winning the value proposition, and as the AI software world continues to embrace open-source alternatives, NVIDIA’s “CUDA-only” moat looks less like a fortress and more like a liability.

Wrapping Up

The competition between AMD’s Strix Halo and NVIDIA’s DGX Spark represents more than just a battle of specs—it’s a philosophical divide in the future of AI computing. In Bend and beyond, where performance, efficiency and practicality matter in equal measure, this face-off captures a turning point. NVIDIA’s long-standing dominance has been built on a closed, high-margin model that rewards those already inside its ecosystem. But AMD’s emergence with the Strix Halo—and the GMKtec EVO-X2 that showcases it—signals a shift toward openness, affordability and local empowerment. When a sub-$2,200 system can deliver performance on par with, or even exceeding, a $4,000 proprietary rig, it’s clear the market is evolving toward accessibility and versatility.

AMD’s approach is democratizing the AI landscape. The Strix Halo isn’t just another high-performance chip; it’s a platform that brings real AI development into the hands of students, startups and everyday creators without needing to lease cloud GPUs or buy into a locked software stack. As AI hardware becomes more compact, efficient and affordable, developers are realizing they can do more with less and own the entire process from idea to implementation. For those looking to give the gift of power and possibility this season, whether to a young coder, a small business innovator or a creative dreamer, AMD’s open-architecture advantage makes the Strix Halo platform not just a product but a statement that the future of AI belongs to everyone willing to build it.