For years now, smartphone makers have made the camera bump on devices bigger in order to chase camera improvements. Even if that kind of design makes cameras better, at times it creates usability issues. With the Pixel 10a, Google took a new approach of entirely removing the camera bump and making a flat phone that lies completely on surfaces.
While this is a delightful change in the world of big camera bumps, Google hasn’t otherwise made major design changes with its newest budget smartphone. The Pixel 9a looked mostly the same, with a very small camera bump.
I have the plain old black unit, but Google offers the phone in Lavender (a mix of bright blue and purple), Berry (coral), and Fog (a gray-green tone) colors.
Look! No camera bump Image Credits: Ivan Mehta
The screen size of 6.3 inches is the same as last year’s device, but the display is now brighter at 3,000 nits. Google is using the Actua display series of screens that it used with the other Pixel 10 devices to make it more usable in bright conditions. The display is capable of reaching a 120Hz refresh rate, but the unit ships with it set to 60Hz, so you will need to manually change that through the phone’s settings.
Build and specification-wise, the Pixel 10a goes toe-to-toe with the Pixel 10, with a few differences. For instance, the Pixel 10 has Corning Gorilla Glass Victus 2 on the front and the back, while the cheaper 10a has a plastic back and Corning Gorilla Glass 7i protection on the front. The budget device also has a bigger battery of 5,100 mAh, as compared to 4,970 mAh on the base Pixel 10. The Pixel 10 Pro XL has a battery of 5,200 mAh.
Advertisement
There are only small differences between the Pixel 9a, the Pixel 10a, and the Pixel 10, most of them having to do with performance and compute power. The obvious hardware difference is that the budget phones use the Google Tensor G4 chip, as compared to the Tensor G5 in the Pixel 10. The Pixel 10 charges at 30W through USB-C, up from the 23W charging capacity of the Pixel 9a. Wireless charging is supported at 7.5W for the Pixel 9a, 10W for the Pixel 10a, and 15W (magnetic) for the Pixel 10.
Image Credits: Ivan Mehta
The battery capacity and faster charging speed are helpful as the battery lasts easily throughout the day, including for regular apps, a few hours of video watching, and light gaming. Plus, the brighter display makes the device better for all-around experience in different lighting conditions. Yes, the 10a has chunkier bezels than its more costly cousins, but they don’t make too much of a difference in daily use. After all, you’re getting the device for a much lower price than a flagship.
Techcrunch event
San Francisco, CA | October 13-15, 2026
Advertisement
The Pixel 10a uses the Tensor G4 chip, which was also used in the Pixel 9a. That means there are no performance gains this year, which you might notice if you switch between a lot of apps. Because of the older chip and its 8GB RAM combo, the Pixel 10a can’t run the updated Gemini Nano AI model, which means it has fewer on-device AI features than the Pixel 10a series.
The display is bright, but there are thick bezels around it Image Credits: Ivan Mehta
The feature list not available on the Pixel 10a includes notification summaries, the Pixel screenshot app, Magic Cue (a feature that offers contextual suggestions across apps like Gmail, Messages, and Maps), call notes, and on-device call translation.
The phone features a 48-megapixel main camera and a 13-megapixel wide-angle camera, which is the same as last year’s device. The main camera performs fine for most conditions, even in low light. But given the older and smaller sensor on the the wide-angle lens, it tends to lose some details, and it doesn’t have autofocus.
The Pixel 10a has a camera coach AI feature that can guide you in taking a shot of an object by helping frame it better in the viewfinder. There is also Auto Best Take, which merges photos to create the best composite from a bunch of shots — useful when photographing a group. The phone also has support for up to 8x super-res zoom, but the processing and quality aren’t as good as the Pixel 10, which offers up to 100x zoom through this feature.
Notably, some AI features might make it to the Pixel 10a through a Pixel Drop, Google’s periodic software updates that** often bring new capabilities to older models.
Advertisement
Google offers seven years of software updates with this device, which is crucial for receiving both operating system updates, along with feature drops and security updates. While this is not Pixel 10a exclusive, the phone has a quick share feature that now works with Apple’s AirDrop. This means I could simply transfer photos, just like I did for this story, to my MacBook within a few taps. Previously, I had to connect the Pixel 10a to my MacBook with a USB-C cable.
At $499, good battery life, a bright display, and faster charging are the main things going in favor of the Pixel 10a. For this price, the phone offers good value for money in a light and flat design. However, if you already have last year’s Pixel 9a, there is no reason to change. Also worth considering: the Nothing phone 4a Pro, also at $499, poses tough with better specifications, such as a bigger and brighter screen, a more capable Qualcomm processor, a dedicated telephoto lens, and faster charging speeds of 50W.
It’s easy to see why the Honda Accord is one of Honda’s most successful cars, and among the bestselling midsize sedans of last year. It’s consistently at or near the top of recommendation lists from big authorities like Car and Driver and Edmunds, and yes, even we’re fans of it. The gas versions of the car are also fairly affordable, ranging up to roughly $32,000 before the hybrid powertrains take over in the lineup. The Accord is quiet, smooth, and decently fun to drive as well, making it a good all-rounder for most people.
However, Honda isn’t making as many gas-only models these days. In 2026, the gas model only has two trim levels, and they’re the two lowest trims. The hybrid powertrain has largely taken over the lineup, much like what Toyota did with the Camry. That means if you want a gas car, there’s a real chance you might have to branch out from the Accord.
Advertisement
Since fuel prices are what they are, one of the most important metrics for shoppers is no doubt fuel economy. So, if you’re looking for a car to cross-shop with the Accord, and you’re wanting to keep fuel economy in mind, then you’ve come to the right place. The list below includes every sedan we could find that meets or exceeds the Honda Accord’s 32 mpg (29 mpg city and 37 mpg highway) in fuel economy. It is worth nothing that all of these pale in comparison to the hybrid leaders in fuel economy.
Advertisement
Three other cars at 32 mpg
Three other gas-only cars match the Accord’s 32 mpg combined. They include the Mini Cooper, the Acura Integra, and the Hyundai Sonata, which only gets 32 mpg on its base trim (all other trims are in the high 20 mpg range). This is already a list of interesting prospects. The base trim of the Hyundai Sonata costs about as much as the base Honda Accord, and given that it has the same fuel efficiency, that gives you a direct one-to-one comparison since the two are fairly similar in terms of virtually all of their specs.
The Acura Integra is also an interesting option because it shares the same platform as the Honda Civic, which means the two cars are very similar. It costs about the same as the hybrid models of the Accord, ranging up to $40,000 for a fully loaded trim. It loses the fight in terms of size but gains it back by being slightly faster than the Accord. It also comes with a manual transmission, which is becoming increasingly rare.
Of the three, the Mini Cooper seems like the most entertaining option. It comes with an optional 201-hp engine that scoots it to 60 mph a whole second and a half faster than the gas-only Accord, at 5.8 seconds, while still getting the same fuel economy. Mini Coopers are fun to drive and are surprisingly spacious for how small they are.
Advertisement
Kia K4 (33 mpg)
The K4 is the smallest gas-only car in Kia’s lineup, replacing the Kia Forte in 2025. It’s also one of Kia’s most affordable vehicles — the base trim goes for $23,000 and the top-of-the-line trim costs a hair over $30,000. That means you can get a maxed-out Kia K4 for the price of a Honda Accord SE, provided you don’t mind the fact that the K4 is smaller overall.
The Kia K4 gets 30 mpg in the city, 40 mpg on the highway, which adds up to 33 mpg combined, just barely edging out the Accord. You only get that fuel economy with the base engine, however, which is notably slower and a little sluggish in general, according to reviews. The more exciting drivetrain is a 1.6-liter turbo-four that outputs 190 hp, and that’s a few tenths of a second faster than the Accord to 60 mph but reduces fuel economy to 26 mph in the city, 36 mph on the highway, and 29 mpg combined.
Advertisement
This will be a running theme in this article. Most cars that are more fuel efficient tend to be smaller, slower, and more sluggish, and the Kia K4 is the first of several such cars. That’s part of what makes the Accord so compelling is that it’s reasonably quick and reasonably fuel efficient at the same time, and you usually have to give up one to get more of the other.
Advertisement
Nissan Sentra (34 mpg)
The Nissan Sentra has been around for decades, and it remains a good budget value for car shoppers. Nissan fully redesigned the car for the 2026 model year, kicking off the ninth generation of the Sentra. Like most other small cars, it costs significantly less than most other models, starting at $23,000 for the base S trim and ranging up to $29,000 for the SL trim. That means you can get a fully loaded SL for around the same price as a base model Honda Accord.
The Sentra nets 31 mpg in the city, 39 mpg on the highway, and 34 mpg combined. It does this by pairing a 149-hp naturally aspirated four-cylinder engine to a CVT transmission. CVT transmissions have their own set of woes, but fuel efficiency isn’t one of them, and that is on full display here. The downside is that the Sentra’s one single engine option is much slower than the Accord, hitting 60 mph eventually, or if you want actual numbers, about a second and a half slower than the Accord.
On the plus side, reviewers agree that the Sentra is comfortable and composed, so while it may take you a bit longer to do things, at least it won’t be uncomfortable. In any case, we’d still take an Accord over a Sentra, since the 2 mpg isn’t really worth the weaker engine and smaller dimensions in this case, but if you’re pinching every penny, the Sentra is cheaper at the dealership and at the pump.
Advertisement
Volkswagen Jetta (34 mpg)
There’s been a lot of Volkswagen Jetta generations over the years, but the car is still good enough to be in the conversation when looking at sedans, particularly compact sedans (we warned you this would be a running theme). This is one of Volkswagen’s smallest and most affordable cars. It starts at $25,000 and ranges up to $31,270 for the top trim, which puts it in the same price category as a Honda Accord SE. That top trim also includes niceties like ventilated seats, which usually come in autos costing way more.
The Jetta gets 29 mpg in the city and 40 mpg on the highway, averaging out to 34 mpg combined. The 29 mpg figure matches the Accord, which means all of its fuel economy gains are on the freeway. Volkswagen equips all Jettas with a 1.5-liter four-cylinder that makes 158 hp, which is mated to an eight-speed automatic. The combo brings the Jetta to 60 mph in 7.8 seconds, which is only half a second slower than the Accord.
Advertisement
There is a Jetta GLI with an optional manual transmission and a faster engine, but its fuel economy dips into the high 20 mpg range, so it’s not necessarily a good comparison to a modern Accord. However, in the prior Accord generation, there was a 2.0-liter turbo-four that did 22 mpg in the city and 32 mpg on the highway. The Jetta GLI compares more to that version of the Accord than it does any of the newer models.
Advertisement
Hyundai Elantra (35 mpg)
The Hyundai Elantra is the second Hyundai to make the list, and as you may have guessed, is also one of Hyundai’s smallest cars. It’s a pretty good example of the modern sedan landscape, as half of its trims, including the most expensive ones, are for hybrid powertrains. Unlike most automakers, it does have a performance model with the Elantra N that is surprisingly zippy. In any case, gas models of the Elantra start at $23,800 and range up to $28,420 for the Limited trim, which makes it cheaper all-around than the Honda Accord.
The Elantra has two gas-only powertrains. The first is a 147 hp four-cylinder that nets 31 mpg in the city and 40 mpg on the highway, for a combined 35 mpg. It scoots the car to 60 mph in 8.1 seconds, which nearly a second slower than the Accord. The engine is mated to a CVT transmission, which helps explain the fuel economy figures. It’s slow, but there are slower cars out there. If this is too small, the Sonata matches the Accord’s 32 mpg and is in the same size class.
Much like the Volkswagen Jetta, the most entertaining variant of the Elantra is its performance model, the Elantra N. This model ranges up to $40,000 and comes with an engine that makes it go to 60 mph in a scant 4.8 seconds. However, its fuel economy is substantially worse than the Honda Accord, so it’s not applicable in this comparison.
Advertisement
Toyota Corolla (35 mpg)
The Toyota Corolla is Toyota’s third most popular car, following only the Rav4 and the Camry. It earned that spot by being affordable, fuel efficient, and reliable, ranking among the best in its segment. Toyota has added a hybrid powertrain to the Corolla, but still sells several gas-only trims that range from $24,000 to around $28,000, or up to $30,000 if you opt for the hatchback. The current model is in its 12th generation, and it’s probably due for a refresh soon since that generation started in 2020.
The Corolla is one of the most fuel-efficient gas-only cars left on the market. It delivers 32 mpg in the city, 41 mpg on the highway, which adds up to 35 mpg combined. It accomplishes this with a 169-hp 2.0-liter four-cylinder engine mated to a CVT. The excellent fuel economy comes at a cost, though, and that cost is speed. The all-wheel drive Corolla gets to 60 mph in about 9 seconds, which almost 2 seconds slower than the Accord and it’s the slowest car on this list. On the plus side, it’s one of the few sedans with all-wheel drive, although you have to get a hybrid to get it.
Advertisement
The Corolla is a pretty easygoing car. It’s slow, sure, but it’s also fairly comfortable to be in, and the fuel economy is hard to argue with. Toyota sells a couple hundred thousand of these a year for a reason.
Advertisement
Honda Civic (36 mpg)
At the top of the list sits the Honda Civic. This car has been around for 50 years and remains one of Honda’s all-time greatest selling cars. It comes in both hatchback and sedan flavors, and costs much less than an Accord. The base trim starts at a hair under $26,000 and ranges up to $29,000. It gets more expensive after that, but that’s when you get into the hybrid drivetrains. Thus, you can get a pretty good mid-trim package for around the same price as an Accord SE.
The Civic nets 32 mpg in the city and 41 mpg on the highway. That’s good for a 36 mpg combined. We’re not sure why this gets 36 mpg and the Corolla gets 35 mpg, given that they have the same city and highway numbers — the EPA creates these figures, not us. In any case, the gas-only Civic comes equipped with a 150-hp 2.0-liter four-cylinder mated to a CVT, which is very much like the Corolla. It goes to 60 mph in 8.9 seconds, which is one tenth of a second slower than the Corolla. In short, if you want the best fuel economy, you have to get a hybrid or go slow.
Despite its slow engine, the Accord is known for being even more zippy and fun to drive around than the Accord. It features sharp handling but enough on-road refinement to avoid feeling too firm or bouncy. The most fuel-efficient alternative to a Honda is, as it turns out, another Honda.
Beijing’s new framework codifies the technology-tracing approach the NDRC used to unwind Meta’s $2bn Manus acquisition, making cross-border AI deals materially harder.
China has formalised a tougher framework for outbound-investment review, codifying the legal-and-administrative posture the National Development and Reform Commission used to unwind Meta’s $2bn acquisition of AI-agent startup Manus in April.
The updated rules, reported by Reuters on Monday, give Chinese regulators a substantially expanded toolkit for blocking cross-border AI and technology transactions, particularly those that involve technology, talent or intellectual property with Chinese origin even if the relevant company is incorporated outside China.
The Meta-Manus case is the template the new framework formalises. Manus, the Chinese-founded AI-agent startup that relocated its corporate headquarters to Singapore before announcing the Meta acquisition in December 2025, was blocked by the NDRC on national-security grounds in April.
Advertisement
The regulator’s reasoning was structurally aggressive: rather than focusing on the company’s current legal domicile, the NDRC examined where Manus’s technology was developed, where its engineering team accumulated expertise, and how the underlying IP was transferred out of the original Chinese corporate entity.
The new rules codify this technology-tracing approach, asserting Chinese jurisdiction over cross-border deals on the basis of technological origin rather than corporate registration.
The substantive consequence is that the Singapore-or-Cayman-Islands restructuring playbook many Chinese AI startups have used over the past five years no longer reliably protects companies from Chinese regulatory review when they accept foreign acquisition offers.
The previous strategic pattern, founded in China, restructured offshore, sold to a US buyer, has been the standard exit route for Chinese AI talent looking to monetise their work in the global market.
The NDRC’s technology-tracing approach, now formalised, means Beijing retains effective veto power over those exits regardless of which jurisdiction the relevant corporate entity sits in at the time of the deal.
Advertisement
The Manus block was the first publicly confirmed use of China’s foreign-investment security-review mechanism to unwind a cross-border AI transaction.
The new rules now make that approach the default rather than the exception, with the NDRC’s framework explicitly covering technology, IP, and key personnel as triggers for review even when the formal acquisition target is non-Chinese.
The framework sits inside a broader 2026 Beijing push that has included expanded travel restrictions on top AI researchers at private firms, instructions to leading AI startups including Moonshot and StepFun to reject US-origin capital without prior clearance, and the parallel push to anchor Chinese AI firms inside mainland-incorporated corporate structures.
The contrast with the US side is the cleaner editorial layer. Washington has spent the past three years tightening outbound-investment rules and expanding semiconductor export controls in an explicit attempt to slow Chinese AI development.
Advertisement
Beijing’s response, on the evidence of the new framework, is to codify a mirror restriction running in the opposite direction: outbound exits, not inbound capital, are the channel China is now closing.
The US is building a wall to stop AI capability from flowing to China; China is building a wall to stop AI capability from flowing out. Both are conditioning their respective technology workforces on the explicit assumption that the bilateral commercial pipeline is no longer trusted infrastructure.
For Meta specifically, the Manus situation now appears to be permanently unwound. The company has reportedly written off the $2bn position in the most recent quarter and abandoned operational integration plans.
For other US tech companies that had been contemplating Chinese-origin AI acquisitions through offshore-incorporated targets, the new rules effectively close that route. Several similar pending deals are reportedly being restructured or abandoned in response.
Advertisement
The broader Chinese AI commercial map is recomposing accordingly. Moonshot AI, StepFun and others that had used offshore-incorporated entities are considering reincorporation onto the mainland, partly because the offshore-protection thesis is now weaker than it was six months ago and partly because Beijing’s domestic-IPO regime offers a clearer exit pathway for companies willing to anchor inside China.
The Chinese AI talent base, for the same reason, is being more aggressively retained inside the country.
The new outbound-investment rules take effect immediately. Foreign acquirers contemplating Chinese-AI-origin assets now face a substantially higher regulatory bar than even four months ago.
Generally, the frame and other structural parts of an FDM printer use steel or similar, but could you use wood instead for that truly artisan look? As [Mitsu Makes] demonstrates after half a year of work, you absolutely can, and it looks about as amazing as you might imagine.
Naturally, you cannot make everything out of wood – such as the linear rails and lead screws – and there is a fair bit of FDM-printed black PLA in there too, but the wood is both structural and decorative. The stained look does really add something. For the FDM-specific parts, the Voron 0 was taken as the base, including the bed. The motion system isn’t CoreXY but Cartesian for ease of construction and driving the axes, while also providing more torque due to the additional motors.
Since it’s more or less a Voron FDM printer and even has automatic bed leveling, it works basically perfectly after assembly and input shaping. Even if it’s not the most practical way to make your own FDM printer from parts, it definitely makes it look unique and would be the focal point of any printing farm.
I’m torn on the price of the Movestyle, though. I love how affordable it is at $580, putting it within a more mainstream budget than I would have assumed. On the other hand, this is a very unique product, and I think higher-end specs might have been a better choice. This is a VA panel rather than IPS, and that means the color accuracy and saturation are OK, but not the best. Although it’s only rated for up to 250 nits of brightness, it topped out at 310 nits when measured against my colorimeter. But it’s not terribly bright, which could be a problem in a brightly-lit room. The display quality isn’t horrible, and this monitor isn’t made for professional video work.
And yet, in terms of the viewing experience, it doesn’t feel all that high-quality, either. For a similar price, you can get a more capable OLED monitor that’s brighter, faster, more colorful, and capable of HDR. But that doesn’t come with the adjustable, rolling stand. An even higher-end monitor would increase the price by at least a few hundred dollars. The lack of a touchscreen feels like a missed opportunity, too, especially since this could easily be used next to a desk or in a kitchen. There are just some cases where using your fingers is easier than using a remote.
Photograph: Luke Larsen
Interestingly, Samsung does sell a more premium Movingstyle monitor that’s even touchscreen-enabled and has a higher refresh rate of 120 Hz for gaming. But it’s a smaller 27-inch panel, comes with a lower-resolution 1440p display, and costs significantly more at $1,200. Whew. Another handy feature of the pricier model is a built-in battery. That means when the cord is unplugged, it doesn’t just immediately die. Speaking of the length of the cord, that does end up being one of the limitations of this design as a whole.
In a lot of ways, that more expensive model feels like what a Movingstyle monitor should be. For my purposes, the larger 32-inch 4K panel matches my needs better.
Advertisement
LG has its own version of this that moves in that direction, the LG Smart Monitor Swing. It comes with a 4K panel, measures 32 inches, and has a screen that can handle touch inputs. At $1,000, it’s priced in between the two Movingstyle monitors. For Samsung, perhaps the solution would be to sell the adjustable stand separately, which would give you the ability to pair it with whatever monitor you want.
You don’t realize just how much you depend on your appliances until one of them goes down. It can happen when you least expect it, forcing you to decide whether or not to find a local repair service, or just buy a new appliance altogether. Before you take your home appliance in for repair, you should know the 50% rule, which says you don’t want to spend over half of what it would take to replace it.
Things can get tough when a repair estimate approaches that 50% line. For example, if you initially paid $1,500 for your refrigerator and the repair cost is $650, that might feel close enough to justify replacing it. But you also have to factor in the time it takes to research the right model, compare prices, and deal with added costs like delivery and installation. Plus, you could be waiting days or even weeks for a replacement to arrive. If the repair gets the original unit back to top form, it might make sense to just get it done.
Age should also play a role. Most appliances, including washing machines, have a limited lifespan. If a unit is nearing the end of its expected shelf life, it may begin breaking down more often. This means repairs will likely just be prolonging the inevitable. Older appliances tend to perform less efficiently as well, which could cost you more money in the long run.
Advertisement
What to know before getting an appliance repair estimate
Anatoliy Cherkas/Shutterstock
If you need to get a repair estimate for an appliance, it’s important to know that the cost can vary based on a number of factors. The brand and model of the appliance, its age, and the urgency of the repair itself, can all play a part. The complexity of the repair is a factor as well, and the same is true of your location.
But before you call a local repair service, check the warranty on your appliance. If you don’t have that information or can’t remember if you have an extended warranty, you should contact the place of purchase. Home Depot and Lowe’s both offer warranties beyond manufacturer warranties. Just be sure to have your information ready to go and they should be able to help. If the warranty has expired, you can ask about replacement parts.
Advertisement
If you decide to get an estimate, you should try to find a reputable repair company. Some appliance manufacturers also have a network of trusted repair mechanics at their disposal. If not, this can be tough, so reaching out to friends and family for their recommendations can be a good idea. If you’re searching online, you can use Google to find repair services, read reviews, and get contact information. You can also visit the Better Business Bureau’s website, where you can use their Appliance Repair tool to find BBB accredited repair services.
COMPUTEX 2026 Intel’s upcoming Diamond Rapids Xeon will boost core counts to 192, a 50 percent increase over last generation, the x86 giant revealed at Computex in Taipei this week.
But while core counts continue to rise, in doing so Intel has managed to cut thread counts by a quarter.
Advertisement
Yep, Hyperthreading – Intel’s marketing for simultaneous multithreading – is officially dead.
Intel first added support for SMT all the way back in 2002. The technology boosted utilization by enabling two threads to harness idle execution units during a single cycle. While SMT doesn’t double throughput, for certain applications it can deliver double-digit percentage gains.
After slowly abandoning the tech across its consumer product lineup, Intel’s Xeons are latest to get the cut.
Except, wait! It seems Intel may have seen the error of its ways, and is already reversing course on the decision. Intel’s next next Xeon, codenamed Coral Rapids, will bring SMT back.
Advertisement
The jump from 128 to 192 is a big jump for Intel, but still smaller than the AMD is making with its 256-core Venice Epycs. If that weren’t enough, it looks like AMD could beat Intel to market by as much as a year.
Diamond Rapids is now slated for release sometime in 2027.
Echos of Epyc, notes of Monaka
In addition to core count, we also got our first look at how Intel will stitch the chip together. It turns out AMD might have been onto something when it started gluing silicon together back in 2017, because Intel’s next round Xeons look more like an Epyc under the hood than ever.
Here’s a quick rundown of the new details we’ve learned about Diamond Rapids at ComputexCourtesy Intel
We know the chip will be fabbed using Intel’s 18A-P process tech, a refined version of its 2nm-class process tech. Beyond this details get a little fuzzy.
Advertisement
From the renders shared in Intel’s press deck, we can see what appear to be two I/O dies serving four vertically stacked compute assemblies assembled using its Foveros packaging tech.
This isn’t the first time we’ve seen something like this from Intel. Intel’s Clearwater Forest, which is finally launching after years of teasing, also used a similar arrangement, with four 24-core compute tiles sitting atop a base die containing the memory controller and L3 cache.
Moving the L3 cache to the base die frees up a lot of die area on the compute chiplet. In this case, we’re looking at four 48-core compute chiplets.
In this respect, Diamond Rapids looks a lot like another CPU we’ve looked at recently: Fujitsu’s Monaka. That chip uses an almost identical chip layout, albeit with one I/O die rather than two.
Advertisement
While we’re fairly certain Diamond Rapid’s L3 cache will live on the base die, the memory controller could be housed on the four base dies or it could be on the I/O dies, similar to what AMD has done since Rome launched in 2019.
If we had to guess, our bet would be on the I/O die, since it would reduce the number of NUMA nodes to one or two as opposed to four.
Not a mainstream part
Unlike Intel’s last P-core Xeon, codenamed Granite Rapids, don’t expect to see Diamond Rapids deployed widely in enterprise virtualization or storage servers.
According to Intel, Diamond Rapids is “optimized for high-demand IaaS, high-perf/thread,” putting it in the same class as its high-performance-computing (HPC)-centric 6900P-series parts.
Advertisement
The lack of SMT complicates hypervisor licensing models. Where you once got two threads for the price of one, Diamond Rapids customers will now be getting half as many for their dollar. There are of course ways of getting around this. Oracle rented out its Ampere-based instances, which also lack SMT, in core-pairs rather than on a core-per-core basis, but something like this would presumably require buy-in from the likes of VMware or RedHat.
As with past HP- optimized processors, Diamond Rapids will be packing a much beefier memory bus than most folks are going to be looking for. HPC workloads like their memory bandwidth and the next-gen Xeon will have no shortage of it with 16-channels of DDR5.
Intel hasn’t disclosed what memory speeds the chip will support out of the box. With that said, Clearwater is already at 8000 MT/s on standard RDIMMS, and Granite could hit 8800 MT/s on MRDIMMS — in fact, 9600 MT/s DIMMS wouldn’t be an unreasonable assumption. That works out to 1.2 TB/s of bandwidth per socket, which happens to be the same as Nvidia’s LPDDR5X-packed Vera CPUs.
That’s not the only thing we’re still in the dark about. Power consumption and instruction per clock gains from the chip’s new architecture are details we expect Intel to trickle out.
Advertisement
The good news: we won’t have to wait long for the next round of specifications, as Intel will be presenting on Diamond Rapids at Hot Chips in August.
The Windows Defender security team is alerting users with dedicated GPUs about scammers manipulating search engine results to distribute remote monitoring and cryptomining payloads. The hackers are manipulating not only search engine results but also AI chatbot responses. Read Entire Article Source link
Apple’s long-delayed AI overhaul may finally be starting to take shape, and the company appears ready to push Siri far deeper into its ecosystem than before. According to a new report from Mark Gurman, Apple is developing a major Siri upgrade that will synchronize AI conversations across devices through iCloud, turning the assistant into a more persistent and connected AI system inside Apple’s tightly controlled ecosystem.
The upcoming Siri redesign is reportedly being prepared as part of Apple’s broader iOS 27 and iOS 28 strategy, with the company positioning the assistant more directly against AI products like Google Gemini and ChatGPT. Instead of functioning as a simple voice tool, Siri is expected to evolve into a conversational AI assistant capable of maintaining synced chat histories across iPhones, iPads, Macs, and other Apple hardware.
Apple wants Siri to become the centre of its AI ecosystem
According to Bloomberg’s report, Apple is internally testing a completely redesigned Siri interface that resembles modern AI chatbot apps. The new experience reportedly includes a dedicated chat-style interface, persistent conversation history, and cloud synchronization powered through iCloud.
This would allow users to begin an AI conversation on one Apple device and continue it seamlessly on another. Apple is reportedly positioning this as a key differentiator for its AI strategy, leveraging the company’s ecosystem advantage rather than competing purely on raw AI model performance.
Advertisement
SiriUnsplash
The report also suggests Apple is integrating Siri more deeply across its software platforms as part of future versions of iOS, iPadOS, and macOS. Internally, Apple is said to already be preparing iOS 28 features while work continues on iOS 27.
The AI-focused Siri upgrade has reportedly faced multiple delays over the past two years, partly because Apple has struggled to modernize Siri’s underlying architecture quickly enough. Gurman notes that several Apple AI projects, including AI-powered AirPods and smart home products, were also slowed by delays tied to Siri’s redevelopment.
At the same time, Apple is preparing for a broader hardware push built around AI experiences. Bloomberg reports the company is developing smart glasses aimed at competing with Meta’s Ray-Ban smart glasses, with Siri expected to play a major role in those products as well.
Why this matters
Apple has been noticeably slower than rivals like Google, OpenAI, and Microsoft in rolling out consumer-facing AI products. While competitors aggressively integrated generative AI into search, productivity apps, and smartphones, Siri has increasingly felt outdated compared to modern AI assistants.
SiriUnsplash
Apple’s strategy appears different, however. Instead of creating a standalone chatbot platform, the company seems focused on embedding AI deeply into its hardware ecosystem and user workflows. That could make Siri more useful for existing Apple users, especially if conversation syncing works smoothly across devices. But it also further strengthens Apple’s famously closed ecosystem approach, where the best experiences are often limited to users fully invested in Apple hardware.
What happens next
Apple is expected to reveal more about its AI plans during upcoming WWDC announcements, though Bloomberg suggests the most ambitious Siri upgrades may not fully arrive until iOS 28. The company is also reportedly developing future AI-powered hardware, including smart glasses, updated HomePods, and refreshed Apple TV products that could rely heavily on the new Siri platform.
Advertisement
For now, Apple’s challenge is becoming increasingly clear. The company no longer just needs to improve Siri. It needs to convince users that its version of AI is worth waiting for after years of falling behind competitors already moving at full speed.
Acer stepped forward at Computex with a handheld that takes a clear stand. The Nitro Blaze Link exists to pull games from a nearby computer and show them on its screen with built-in controls. It leaves the heavy graphics work on a stronger machine and keeps its own parts minimal on purpose. Designers gave the body ergonomic grips along each side for steady handling during extended play. Standard controls fill the surface in familiar spots, including dual analog sticks, a directional pad, face buttons, shoulder bumpers, and triggers. A seven-inch touchscreen sits front and center with 1920 by 1200 resolution and support for five touch points at once.
The device weighs 464 grams (nearly a pound), is approximately 287 millimeters wide, and has a maximum thickness of 33.5 millimeters. This is all fairly compact, making it easy to grasp in your hand or throw into a bag without any effort. The power comes from an 18-watt-hour battery, which isn’t bad. There is a single USB-C port that can take up to 15 watts of charging, but no data transmission is available through that port, so you won’t be shifting files or bringing in more peripherals. A set of two-watt stereo speakers or a traditional 3.5 millimeter headphone connector provide audio output.
【All-in-One Handheld】 The ultimate portable console in your hands. Enjoy retro games, remote play, cloud gaming, and Android titles on a…
【Cloud Gaming + Remote Play】 Cloud Gaming – Instantly play AAA titles over Wi-Fi with subscriptions to Xbox Cloud Gaming (Beta), GeForce NOW…
【Emulator Retro Games + Android Games】 Emulation – Enjoy retro classics with enhanced compatibility and smoother performance on PS2, GameCube…
It runs on Debian Linux, and Moonlight is configured to receive video streams. To broadcast video and receive controller signals, the host computer must have Sunshine installed. However, games run at full speed on the main system, with the handheld simply decoding video and sending back button presses. That split contributes significantly to the internal hardware’s reduced weight. You’re looking at 1GB of LPDDR4 RAM and 8GB of integrated flash storage, which aren’t very large numbers, especially when compared to what you’d get on a typical gaming device. Interestingly, they appear to handle video decoding perfectly fine. WiFi 6 includes 80 megahertz channel support and a few improvements designed to keep shared home networks working smoothly.
Acer designed this to work seamlessly with its own Predator Helios and Nitro laptops. Owners may play their whole game libraries from another room or a distance because the setup is straightforward and local. Make no mistake: the hardware has limitations. It cannot run games locally and cannot store much additional stuff. Storage and RAM are simply too limited for that; if you try to squeeze in a bunch of extra apps or save files, things become a little crowded in there. Performance is also dependent on having a reliable wireless connection, and the main system must be able to encode the stream correctly. Any issues are mainly caused by the network rather than the device. No cloud services have been confirmed to be supported, and the Linux foundation and focus on Acer-compatible systems limit what you can accomplish with this device when compared to more open handhelds.
This device won’t be available in the United States until the fourth quarter of 2026, and Acer hasn’t revealed any pricing details yet. When compared to more serious handhelds with comparable specifications, the Nitro Blaze Link appears to be a more affordable option. [Source]
Gaming desktops have been getting smarter every year with better cooling, faster chips, and more RGB lights than anyone asked for. MSI has decided that none of that is interesting enough and has introduced something genuinely unexpected at Computex 2026.
The Taiwanese company has unveiled a gaming desktop, and its most interesting aspect is the built-in cylindrical display that exists purely to give your AI companion a physical avatar.
What is the AI Holostage and why does it exist?
The MEG Vision X2 AI+ is MSI’s new flagship gaming desktop, which sits at the top of its MEG product line. The main main highlight is the AI Holostage, a cylindrical display integrated directly into the chassis rather than sitting as a separate accessory.
The Holostage gives digital companions, desktop pets, and custom third-party AI avatars a visible, physical presence on your desk. Out of the box, the system ships pre-configured with MSI’s own AI companion, which is called LuckyClaw.
Advertisement
LuckyClaw is MSI’s agentic AI companion, which responds to natural voice commands and gives users hands-free control over performance profiles, MSI monitor settings, and RGB lighting.
The chassis also features a tool-free upgradeable design. While this might be something of a novelty or a visually appealing factor, it might not serve a practical utility purpose as such.
Pricing and availability have not been confirmed in the official press release. Whether the Holostage becomes the defining feature for gaming PCs or ends up as the most elaborate desktop widget ever built remains to be seen. But as a statement of intent, it is hard to ignore.
Advertisement
Some popular gaming desktop brands, including Asus, Razer, and Lenovo, have tested virtual assistants in their software ecosystems. However, MSI is the first to give the companion a dedicated physical display embedded in the chassis itself.
You must be logged in to post a comment Login