Technology

If 2025 is the year you’ve finally promised yourself you will actually get fit, lose weight, run that marathon, or lift those weights, then you should make sure you have as many tools at your disposal as possible.

Starting a new fitness challenge, program, or routine can be really daunting, especially for first-timers. If you own a smartphone, however, there are a ton of great apps out there that can really help you get on top of your fitness goals for 2025. Many of them are free, while some of them require a subscription. Either way, starting out with one of the best fitness apps can be one of the best ways to help your fitness goals.

They can be downloaded to a smartphone you probably already own, and there's usually no immediate upfront cost like there is for a new fitness tracker, smartwatch, or gym membership. So whether you want to run your first marathon, take control of your calories, or just get out and walk, here are some of the best apps to help you achieve your 2025 fitness goals.

1. Fitbod: Workout & Gym Planner

(Image credit: Future)

Fitbot is our number-one pick of the best fitness apps on both iPhone and Android. It’s a pocket personal trainer with a massive range of exercises that caters to both beginners and experienced users. It also has an impressive UI and can be used even without any equipment.

The full version will cost you $12.99/£10/AU$20 a month or $79.99/£63/AU$122 a year, but there’s a very limited free trial if you’d like to test it out.

• Download Fitbod for iPhone
• Download Fitbod for Android

2. Runna

(Image credit: Runna)

Runna is one of our very favorite running apps thanks to its personalized running programs, very solid design, and the broad range of features you get for free.

I’ve personally used this app for a really solid Couch to 5K plan, and the workouts feature really helpful guidance that talks to you as you run. If you want to do any sort of running in 2025, be it your first-ever 5K, an ultra or anything in between, this definitely should be top of your list.

• Download Runna for iPhone
• Download Runna for Android

3. Push

(Image credit: Future)

I’ve used Push for well over a year as my go-to daily workout companion. It’s an AI-powered strength training app with a stunning interface, and workout plans designed to give you maximum results in the gym when it comes to strength training. You won’t find anything except weightlifting for the purposes of building muscle, but if you want to get massive in 2025, this is the one I’d recommend.

• Download Push for iPhone
• Download Push for Android

4. MyFitnessPal

(Image credit: MyFitnessPal)

There’s an old adage that you can’t out-train a bad diet, and it’s sadly true. All that hard work in the gym isn’t going to matter if you aren’t doing the heavy lifting in the kitchen.

Whether you want to shed those pounds, build muscle, or ensure you’re fuelled for long distances, getting your food right is crucial MyFitnessPal is a mainstay calorie tracking app that lets you input your targets and goals and divide up your macronutrients (protein, carbs and fat) to help you hit your target.

The app’s peak feature is undoubtedly the barcode scanner, which gives you crowdsourced nutritional information about the food you buy so you don’t have to put it in manually.

• Download MyFitnessPal for iPhone
• Download MyFitnessPal for Android

5. Strava

(Image credit: Lauren Scott)

If you want to get into the outdoors and feel like you’re part of a fitness community, Strava’s 135 million members are waiting for you.

Ever-popular amongst cyclists and runners, Strava lets you record your exercise and share it with the world, so if you’re ultra-competitive and want to take your fitness to the next level, or you want to join like-minded exercises on your first steps, it’s hard to go wrong with Strava. It also offers analysis and feedback, importing data from your wearable, and kit tracking.

• Download Strava for iPhone
• Download Strava for Android

6. Noom

(Image credit: Noom)

Noom is a sleeper weight loss and diet app that uses personalized programs and psychology to not only track your food and log calories, but also change your behavior and thinking around food.

It’s a dedicated weight-loss app with a serious price tag (at least $209 a year) but could be a winner for those who really want to get inside the psychology of food.

• Download Noom for iPhone
• Download Noom for Android

7. Tasty

(Image credit: Tasty)

Tasty is a much more conventional meal planning and recipe app that can help you come up with great, tasty food to keep your diet going.

There are plenty of indulgent and decadent recipes for cheat days, but there are also easy dinners, five-ingredient meals, and of course healthy options for low-carb, low-calorie, high-protein, or low-fat meals.

• Download Tasty for iPhone
• Download Tasty for Android

8. Headspace

(Image credit: Headspace)

After a life-changing brush with health anxiety a couple of years ago, Headspace is one of the most important apps on my iPhone these days. There’s no point doing all that legwork in the gym and on the dinner table, if you’re not taking care of your mental wellbeing too.

Headspace features a massive range of mindfulness exercises from brief meditations and breathing exercises to 30-day courses. There are even guided meditations for exercises like yoga and cardio, as well as specific courses for mindful eating.

• Download Headspace for iPhone
• Download Headspace for Android

You might also like...

• I only pay for one wellness app every year, and I can’t live without it
• Samsung Health just got one of MyFitnessPal's best features
• Samsung's plan for your health is big, bold, and AI-powered to help you help yourself

Nvidia experienced a memorable 2024 in some ways, and certainly CEO Jensen Huang won’t forget the year in which his company finally outdid Apple as the most valuable firm in the world.

We also received a clutch of powerful new GeForce GPUs from Team Green, along with a huge move on the software front too, and, inevitably, AI continued to be a massive driving force for Nvidia.

Nvidia’s value exceeded $1 trillion in 2023, but it blew through the roof this year, with its stock heading steeply upwards throughout 2024 – driven, of course, in large part by its AI GPUs.

Indeed, Nvidia became the world’s most valuable company in June 2024, beating out the likes of Apple and Microsoft, exceeding a market cap of $3 trillion. Team Green slipped back a bit as the year went on, but then retook pole position from Apple in November 2024, approaching a $3.5 trillion valuation. Nvidia also barged Intel off the Dow Jones Industrial Average index.

Throughout the year, Jensen was gleefully stoking the AI hype furnace, unsurprisingly. A lot of leather jackets can be bought for the kind of cash Nvidia is now worth, and the swiftness with which this financial rise has occurred has been nothing less than breathtaking.

Is Nvidia going to be the first company to hit the $4 trillion mark? It’s a fair bet if the momentum behind AI keeps on chugging the way it has been – and maybe Nvidia is even set to become a household name eventually (something it most definitely isn’t yet, despite all this success).

(Image credit: Future / John Loeffler) The year of the Super graphics cards

Nvidia continued to dominate the desktop GPU market in 2024, to no one’s surprise, to the point of a near-monopoly in fact.

Helping to maintain its momentum here was a trio of new graphics cards for the Lovelace range which Nvidia revealed as the year kicked off. These fresh additions to the best Nvidia GPUs out there were the GeForce RTX 4070 Super, RTX 4070 Ti Super, and RTX 4080 Super.(And yes, we still can’t get over the fact that Nvidia made a ‘Ti Super’ as a way to jam an extra model into a cluttered mid-to-upper-range space).

These Super GPUs replaced the existing models in all but the case of the RTX 4070, which was kept in production to run alongside the RTX 4070 Super, but at a cheaper price point (obviously).

Overall, the new graphics cards were welcome introductions and powerhouse offerings, but what wasn’t so welcome were the price tags attached to them. In true Nvidia style, these mid-to-higher leaning GPUs were all expensive. Notably, the RTX 4070 Ti Super fell a little short in our review, being rather overshadowed by the RTX 4080 Super, which we declared was the enthusiast GPU we’ve all been waiting for.

Our review of the RTX 4070 Super also represented a big thumbs-up for the GPU, and it proved to be our favorite overall of the new graphics cards (even if its 12GB of VRAM limits the card’s 4K chops – the 4080 Super become the new champ here, of course).

On the issue of priciness, it’s true that some of these new Super variants got price cuts to a limited extent as the year rolled on, but overall, AMD’s mid-range remained the much better value picks compared to Team Green.

Otherwise, we didn’t see much else from Nvidia in the way of GPU releases, save for a fresh spin on the RTX 4070 with slightly slower video RAM. It was effectively the same as the original RTX 4070, though, and a move made by Team Green to ensure supply remained strong, we were told.

(Image credit: Future/Jackie Thomas)

There was nothing to be seen elsewhere, and once again, tumbleweeds at the lower-end of the market. The old RTX 3060 remained a strong seller as a result – the 12GB flavor, with that VRAM loadout proving a definite carrot for some gamers – despite chatter from the grapevine around the middle of the year that Nvidia might be discontinuing this model.

Indeed, the RTX 3060 is still the most popular GPU out there going by the Steam hardware survey, with the RTX 4060 making good headway too – that graphics card remains our top pick for 1080p gaming (and it’s solid for 1440p as well).

Despite plenty of rumors suggesting a 2024 launch earlier in the year, we didn’t see the RTX 5000 series turn up this year, with the expectation now being that Nvidia will launch new RTX 5090 and 5080 (and possibly RTX 5070) GPUs at the very start of 2025 at CES.

There was some super new software, too

Broadly speaking, Nvidia GPU owners have had a somewhat rocky relationship with GeForce Experience. A good chunk of gamers with an Nvidia graphics card preferred to just install the graphics driver, and not bother with the companion software, GeForce Experience, at all – and Team Green took the hint. In 2024, Nvidia canned GeForce Experience and made an all-new official app.

The Nvidia App (oh, the hours upon hours that must’ve been spent in brainstorming marketing meetings coming up with that name) arrived in beta in February 2024. It was billed as an all-in-one replacement for GeForce Experience and the Nvidia Control Panel, plus the RTX Experience. All these separate pieces were instead housed under one convenient umbrella (where driver installs are handled, too).

The Nvidia App emerged as a full release in November, after being worked on extensively during the year. Nvidia even actively solicited feedback from gamers on which legacy features should be preserved, and more besides.

That feedback was listened to in terms of implementing elements such as in-game frame rate and latency info (plus much more) in the revamped overlay for the Nvidia App, and it offered some smart new features such as easy GPU overclocking. The new app was generally well-received, appears to run smoothly and responsively in the main, and fortunately, there are no onerous account or login requirements here either – hopefully that’ll remain the case going forward.

There was one notable niggle that cropped up in mid-December, though, when reports emerged that an option in the Nvidia App could cause considerable slowdown of gaming frame rates. Nvidia is currently investigating that problem (at the time of writing), and there’s a fix (of sorts) for the apparent bug.

Also on the software side of the equation, DLSS continued to dominate the frame rate boosting scene, and remains a powerful weapon in Nvidia’s armory of GPU tricks.

(Image credit: Nvidia) AI FTW still

Naturally AI remained an area where Nvidia realized some big successes this year, and as we already mentioned, it helped to drive the company’s market value skywards (or moon-wards, perhaps we should say).

Team Green was predictably keen to push forward with new Rubin AI chips, the successor to Blackwell for AI workloads, being announced just a few months after the latter.

It’s possible that Rubin, which was the focus of Computex 2024, could power the GeForce graphics cards that follow the next generation – mirroring the way Blackwell has been deployed for AI GPUs and RTX 5000 desktop boards – so this could have been our first (sort of) sighting of RTX 6000. (If RTX 6000 ever happens, and here at TechRadar, we’ve made arguments as to why there are valid reasons to doubt this – namely the juggernaut profits in the AI world, of which there are a bunch of major concerns around, it has to be said).

(Image credit: Nvidia) GeForce Now continues to evolve – with a catch

Last year was an important one for Nvidia’s streaming service for PC gamers, with the new ‘Ultimate’ subscription coming in (offering up to 240 frames per second, and a less laggy experience – for those with an internet connection up to the task, that is).

Early in 2024, Nvidia introduced a Day Pass for GeForce Now – allowing you to try out the full service for a day, to see how it works for you – and then towards the tail end of the year, a big upgrade was announced for the middle tier ‘Priority’ membership. This was renamed ‘Performance’ and Nvidia boosted its visual quality from 1080p to 1440p resolution (with ultrawide monitor support on top), all with no additional cost.

At least there was no financial cost, but there was a catch in the form of a monthly time limit imposed on these subscribers (and Ultimate tier members too). On the face of it, capping play time was a move to help Nvidia shorten queues and keep the streaming quality running smoothly, but it went down badly with some subscribers, who felt the 100-hour limit was too stingy.

Note that the new time limit doesn’t come into force until 2025 starts, and won’t apply to existing subscribers until 2026. Still, this led to plenty of threats of quitting GeForce Now on online forums – we’ll just have to see if that turns out to be mere noise, or whether a bunch of subscribers are indeed about to exit stage left.

(Image credit: Nvidia) Concluding thoughts

In 2024, Nvidia brought some high-quality graphics cards into its Lovelace range, although the more affordable end of the GPU market was totally neglected (again). The company continues to dominate the desktop graphics card arena, and that won’t likely change in the near future – especially not at the higher-end where Nvidia won’t even be challenged going forward, if the rumors are right.

The Nvidia App was a high point for Team Green – and it’s great to see gamer feedback helped to shape the software – with DLSS also helping to build on, and reinforce, its consumer GPU success.

AI was the massive money-spinner, though, and the driving force behind Nvidia exploding to become the world’s most valuable company.

As we touched on above, the worry for PC gamers might be that the blistering success of AI GPUs – if it continues, and it doesn’t seem a good idea to bet against that – could mean Nvidia’s gaming graphics cards are eventually side-lined, or perhaps dispensed with entirely. With AMD also looking to cash in on AI, and Intel’s GPU prospects seeming shakier these days, all this does prompt some concern around the future of desktop graphics cards in the consumer marketplace.

You might also like

• Uh oh... Zotac just leaked Nvidia’s next-gen launch line-up, including RTX 5090 GPU with 32GB of VRAM
• We pick the best processors
• These are the best GPUs

Countries must collaborate if we’re going to create a less fearful future for artificial intelligence.

• SK Hynix's 321-layer NAND targets AI-driven data storage needs
• 321-layer NAND flash improves data transfer speed by 12%
• AI storage demands drive innovation in high-capacity NAND solutions

Samsung and SK Hynix have continued their head-to-head battle in the NAND flash memory market with the latter taking the lead with a new launch.

SK Hynix, the world’s second-largest memory chipmaker, recently became the first to mass-produce triple-level cell (TLC) NAND flash with over 300 layers.

The company’s new 321-layer, 1-terabit TLC 4D NAND flash, announced recently, is set to revolutionize the data storage industry, paving the way for more affordable ultra-high-capacity solid-state drives (SSDs) that exceed 100TB in capacity.

SK Hynix 321-layer NAND

The NAND industry is racing to push the limits of data storage technology, and SK Hynix's achievement is a major milestone.

After the launch of its 238-layer NAND flash last year, SK Hynix’s latest 321-layer NAND flash sets a new industry standard. The company plans to supply these chips to customers starting in the first half of 2025, targeting the booming artificial intelligence (AI) market, which demands high-performance, energy-efficient storage solutions.

The 321-layer NAND was made possible by SK Hynix’s "Three Plugs" process technology, which involves electrically connecting three plugs through an optimized follow-up step, significantly improving the speed, power efficiency, and overall performance of the chips.

SK Hynix also developed a low-stress material and introduced technology that automatically corrects alignment among the plugs to further optimize the manufacturing process.

The 321-layer product offers a 12% increase in data transfer speed and a 13% improvement in reading performance relative to the previous 238-layer NAND. Furthermore, it reduces power consumption by over 10%. With a 59% boost in productivity, SK Hynix’s new NAND offers an enhanced storage solution for AI data centers and on-device AI applications.

While SK Hynix has achieved this historic feat, its main competitor, Samsung, is not far behind. Samsung is reportedly working on a 400-layer NAND flash chip, which it plans to release by 2026.

The company’s roadmap includes developing bonding vertical NAND (BV NAND) technology, which will allow for even greater storage density and minimized heat buildup. Samsung’s long-term goal is to introduce NAND chips with over 1,000 layers by 2030, potentially breaking the 200TB storage barrier for AI-driven SSDs.

"SK Hynix is on track to advancing to the Full Stack Al Memory Provider by adding a perfect portfolio in the ultra-high performance NAND space on top of the DRAM business led by HBM," noted Jungdal Choi, Head of NAND Development at SK Hynix.

Via KEDGlobal

You may also like

• These are the best portable SSDs
• Here is our roundup of the best business laptops
• The Vodafone and Three merger could come sooner than thought after major step forward

The only game Player 456 wants to partake in is vengeance. Sweet, sweet vengeance.

• One processor architecture for AI, embedded systems, and robotics
• RISC-V scalability ensures seamless expansion across diverse applications
• Seed funding boosts Ubitium’s drive to redefine chip technology

For over 50 years, the semiconductor industry has relied on the Tomasulo algorithm, introduced by IBM in 1967, to build specialized CPUs, GPUs, and other chips tailored to specific computing tasks.

Now, Ubitium, a hardware startup founded by semiconductor veterans, has developed a universal RISC-V processor that consolidates all computing workloads onto a single, affordable chip.

This technology is particularly significant for embedded systems and robotics, where the cost of hardware often limits the deployment of advanced computing solutions.

Erasing the boundaries between specialized computing tasks

Ubitium’s universal processor is designed to be scalable, supporting a portfolio of chips that vary in size but share the same microarchitecture and software stack, ensuring customers can expand their applications without altering their development processes.

The processor’s workload-agnostic design makes it suited to any computing task and helps to simplifying hardware requirements.

Ubitium has raised $3.7 million in seed funding, which will accelerate the development of prototype chips and initial development kits, with plans to launch the first commercial processors by 2026.

"The $500 billion processor industry is built on restrictive boundaries between computing tasks," noted Hyun Shin Cho, CEO of Ubitium.

"We're erasing those boundaries. Our Universal Processor does it all - CPU, GPU, DSP, FPGA - in one chip, one architecture. This isn't an incremental improvement. It is a paradigm shift. This is the processor architecture the AI era demands."

Cho further stated the company envisions a future where a single processor design can handle tasks ranging from small embedded systems to high-performance computing without specialized hardware modifications.

“For too long, we’ve accepted that making devices intelligent means making them complex. Multiple processors or processor cores, multiple development teams, endless integration challenges—today, that changes," he added.

You may also like

• What is an NPU: the new AI chips explained
• These are the best AI tools
• EU reveals sites for major AI factories across Europe

Here are the answers for The New York Times Mini Crossword for Dec. 26.

Here are some hints and the answer for Wordle No. 1,286 for Boxing Day, Dec. 26.

Here are some hints -- and the answers -- for the Dec. 26 Strands puzzle, No. 298.

Here are some hints — and the answers — for Connections No. 564 for Thursday, Dec. 26.

Here are some hints -- and the answers -- for Connections: Sports Edition No. 94 for Dec. 26.

Looking for the best internet provider in Sebring? CNET’s broadband experts will guide you to the right plan for your needs.

Quordle was one of the original Wordle alternatives and is still going strong now more than 1,000 games later. It offers a genuine challenge, though, so read on if you need some Quordle hints today – or scroll down further for the answers.

Enjoy playing word games? You can also check out my NYT Connections today and NYT Strands today pages for hints and answers for those puzzles, while Marc's Wordle today column covers the original viral word game.

SPOILER WARNING: Information about Quordle today is below, so don't read on if you don't want to know the answers.

Quordle today (game #1067) - hint #1 - Vowels How many different vowels are in Quordle today?

• The number of different vowels in Quordle today is 4*.

* Note that by vowel we mean the five standard vowels (A, E, I, O, U), not Y (which is sometimes counted as a vowel too).

Quordle today (game #1067) - hint #2 - repeated letters Do any of today's Quordle answers contain repeated letters?

• The number of Quordle answers containing a repeated letter today is 1.

Quordle today (game #1067) - hint #3 - uncommon letters Do the letters Q, Z, X or J appear in Quordle today?

• No. None of Q, Z, X or J appear among today's Quordle answers.

Quordle today (game #1067) - hint #4 - starting letters (1) Do any of today's Quordle puzzles start with the same letter?

• The number of today's Quordle answers starting with the same letter is 0.

If you just want to know the answers at this stage, simply scroll down. If you're not ready yet then here's one more clue to make things a lot easier:

Quordle today (game #1067) - hint #5 - starting letters (2) What letters do today's Quordle answers start with?

• T

• O

• F

• S

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

Quordle today (game #1067) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle, game #1067, are…

• TATTY
• OVERT
• FLOUR
• SWEPT

With a little more luck – or skill – I would have finished a bit quicker today, but got stuck on the many possible words that ended up being SWEPT.

With just a handful of unpicked letters left I knew that my final word was likely to have a double TT and gambled that it was TATTY.

A very satisfying game of Quordle.

How did you do today? Send me an email and let me know.

Daily Sequence today (game #1067) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle Daily Sequence, game #1067, are…

• BUGLE
• HATER
• RARER
• CANON

Quordle answers: The past 20

• Quordle #1066, Wednesday 25 December: LIVER, CRIER, SHORE, MURKY
• Quordle #1065, Tuesday 24 December: TIPSY, SCALE, FLYER, VALVE
• Quordle #1064, Monday 23 December: SNAKE, SAVVY, ELBOW, TWEED
• Quordle #1063, Sunday 22 December: DOUBT, SLURP, PAPER, CIVIL
• Quordle #1062, Saturday 21 December: STRIP, BUDGE, SAPPY, SHELF
• Quordle #1061, Friday 20 December: PENAL, EAGLE, DILLY, QUITE
• Quordle #1060, Thursday 19 December: JEWEL, GNOME, TRAMP, IDLER
• Quordle #1059, Wednesday 18 December: LITHE, BIRTH, HUMID, ROBOT
• Quordle #1058, Tuesday 17 December: DOING, ENJOY, SNAKY, AMPLE
• Quordle #1057, Monday 16 December: HYDRO, CREAM, CHIDE, SLIME
• Quordle #1056, Sunday 15 December: DROVE, STILT, LINEN, GIANT
• Quordle #1055, Saturday 14 December: SPIEL, ONSET, TIGER, DITTO
• Quordle #1054, Friday 13 December: ERECT, REBAR, MOIST, IDIOM
• Quordle #1053, Thursday 12 December: MILKY, CRIER, STORK, DRIER
• Quordle #1052, Wednesday 11 December: BRIAR, NASAL, SHARD, FLUFF
• Quordle #1051, Tuesday 10 December: BEGAN, TRIAL, IDEAL, GUIDE
• Quordle #1050, Monday 9 December: TITAN, FORTE, SPEED, BRIAR
• Quordle #1049, Sunday 8 December: BUNCH, SHINY, MOULT, INTRO
• Quordle #1048, Saturday 7 December: SHAKY, CROOK, GHOUL, VERSE
• Quordle #1047, Friday 6 December: BRAWL, UNTIL, DOWRY, STING

Good morning! Let's play Connections, the NYT's clever word game that challenges you to group answers in various categories. It can be tough, so read on if you need clues.

What should you do once you've finished? Why, play some more word games of course. I've also got daily Strands hints and answers and Quordle hints and answers articles if you need help for those too, while Marc's Wordle today page covers the original viral word game.

SPOILER WARNING: Information about NYT Connections today is below, so don't read on if you don't want to know the answers.

NYT Connections today (game #564) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• TIDY
• ALA
• HANDSOME
• ELDER
• SPRUCE
• LIKE
• PINE
• MISS
• HEALTHY
• ROUGHLY
• ARK
• PALM
• ABOUT
• RESPECTABLE
• ORE
• AROUND

NYT Connections today (game #564) - hint #1 - group hints

What are some clues for today's NYT Connections groups?

• YELLOW: Generally similar
• GREEN: Evergreens
• BLUE: A decent portion
• PURPLE: Four of the 50

Need more clues?

We're firmly in spoiler territory now, but read on if you want to know what the four theme answers are for today's NYT Connections puzzles…

NYT Connections today (game #564) - hint #2 - group answers

What are the answers for today's NYT Connections groups?

• YELLOW: APPROXIMATELY
• GREEN: TREES
• BLUE: SIZABLE, AS AN AMOUNT 
• PURPLE: U.S. STATE ABBREVIATIONS, PER AP STYLE GUIDE 

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Connections today (game #564) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #564, are…

• YELLOW: APPROXIMATELY ABOUT, AROUND, LIKE, ROUGHLY
• GREEN: TREES ELDER, PALM, PINE, SPRUCE
• BLUE: SIZABLE, AS AN AMOUNT HANDSOME, HEALTHY, RESPECTABLE, TIDY
• PURPLE: U.S. STATE ABBREVIATIONS, PER AP STYLE GUIDE ALA, ARK, MISS, ORE

• My rating: Moderate
• My score: 2 mistakes

I wasted time today thinking HANDSOME, HEALTHY and SPRUCE were somehow connected – I was imagining some shaving gel commercial featuring some buff man with a chiseled jaw admiring himself in the mirror and thinking how, erm, ARK he is. Sometimes when you can’t find the fourth word it’s for a good reason.

After seeing the tree link, I got there in the end – although those state abbreviations were far from obvious and I would never have got them if it wasn’t for the fact that they were the final Purple four. We still click on those last four words like we knew all along though, right? Yeah, yeah U.S. STATE ABBREVIATIONS, PER AP STYLE GUIDE, yeah, yeah, I knew that.

Yesterday's NYT Connections answers (Wednesday, 25 December, game #563)

• YELLOW: CELESTIAL OBJECTS COMET, MOON, PLANET, STAR
• GREEN: ARCHERS CUPID, HAWKEYE, ROBIN HOOD, SAGITTARIUS
• BLUE: FEMALE ANIMALS JENNY, NANNY, QUEEN, VIXEN
• PURPLE: "S.N.L." CAST MEMBERS FEY, RUDOLPH, SHANNON, STRONG

What is NYT Connections?

NYT Connections is one of several increasingly popular word games made by the New York Times. It challenges you to find groups of four items that share something in common, and each group has a different difficulty level: green is easy, yellow a little harder, blue often quite tough and purple usually very difficult.

On the plus side, you don't technically need to solve the final one, as you'll be able to answer that one by a process of elimination. What's more, you can make up to four mistakes, which gives you a little bit of breathing room.

It's a little more involved than something like Wordle, however, and there are plenty of opportunities for the game to trip you up with tricks. For instance, watch out for homophones and other word games that could disguise the answers.

It's playable for free via the NYT Games site on desktop or mobile.

Strands is the NYT's latest word game after the likes of Wordle, Spelling Bee and Connections – and it's great fun. It can be difficult, though, so read on for my Strands hints.

Want more word-based fun? Then check out my NYT Connections today and Quordle today pages for hints and answers for those games, and Marc's Wordle today page for the original viral word game.

SPOILER WARNING: Information about NYT Strands today is below, so don't read on if you don't want to know the answers.

NYT Strands today (game #298) - hint #1 - today's theme What is the theme of today's NYT Strands?

• Today's NYT Strands theme is… Relative conjunction

NYT Strands today (game #298) - hint #2 - clue words

Play any of these words to unlock the in-game hints system.

• TOWN
• WHEN
• RENEW
• TUNA
• SWAN
• CLUE

NYT Strands today (game #298) - hint #3 - spangram What is a hint for today's spangram?

• Get together

NYT Strands today (game #298) - hint #4 - spangram position What are two sides of the board that today's spangram touches?

First side: left, 3rd row

Last side: right, 4th row

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Strands today (game #298) - the answers

(Image credit: New York Times)

The answers to today's Strands, game #298, are…

• AUNT
• INLAW
• GRANDCHILD
• NIECE
• COUSIN
• UNCLE
• NEPHEW
• SPANGRAM: REUNION

• My rating: Easy
• My score: Perfect

I’m sure I’m not the only one who started looking for nouns that introduce a relative clause in grammar (that, which, whom, etc), but that would have been a little too easy perhaps.

That said, today’s relative hunting wasn’t taxing – although even the thought of entertaining this many relations at a REUNION was enough to make me a little stressed.

How did you do today? Send me an email and let me know.

Yesterday's NYT Strands answers (Wednesday, 25 December, game #297)

• COAL
• SOCKS
• PENS
• SCARF
• TOYS
• CANDY
• ORANGE
• PLUSHIE
• SPANGRAM: STOCKING

What is NYT Strands?

Strands is the NYT's new word game, following Wordle and Connections. It's now out of beta so is a fully fledged member of the NYT's games stable and can be played on the NYT Games site on desktop or mobile.

I've got a full guide to how to play NYT Strands, complete with tips for solving it, so check that out if you're struggling to beat it each day.

• Lightweight 563g tablet sports MediaTek Kompanio 520 processor and LTE option
• Up to 15 hours of battery life with 4GB/8GB memory and 32GB/64GB storage
• Dual cameras, USB 3.2 Type-C ports, and Wi-Fi 6 connectivity

Dynabook has unveiled the Dynabook Chromebook C70 - a 10.1-inch convertible rugged tablet tailored specifically for education environments, and is designed with regular drops and knocks in mind.

This isn't to say its internals have been neglected however, with the C70 sporting a MediaTek Kompanio 520 CPU, up to 64GB of storage, and a range of ports.

Under the hood, the Chromebook C70 is powered by the MediaTek Kompanio 520 processor, offering reliable performance for typical educational applications. It comes with customizable memory options of 4GB or 8GB and storage capacities of 32GB or 64GB, providing adequate resources for classroom tasks.

Convertible design for enhanced flexibility

The Chromebook C70 has detachable WXGA touchscreen (1,280 x 800 resolution), which can function as a standalone tablet or be docked into a full-featured keyboard. The tablet weighs approximately 563g (575g for the LTE model), while the complete setup with the keyboard dock comes in at about 1,092g.

As for its rugged features, the outer casing is made from non-slip thermoplastic polyurethane, and includes rounded corners that help absorb impacts. The keyboard is also equipped with reinforced keycaps to prevent them from falling off when dropped. The Dynabook Chromebook C70 complies with MIL-standard durability tests.

The device includes a dedicated rechargeable USI active electrostatic pen, stored conveniently in a built-in slot that charges it while not in use. With a battery life of up to 15 hours, the device is well-suited for full-day school use without frequent recharging. Additionally, its dual cameras, 2MP front and 5MP rear, cater to video conferencing and casual photography needs.

The device offers a versatile range of connectivity options, including one USB 3.2 Gen 1 Type-A port, two USB 3.2 Gen 1 Type-C ports with support for Power Delivery and external displays, a microSD card reader, and a microphone/headphone combo jack. In addition, it also supports Wi-Fi 6 (IEEE802.11ax) for fast and reliable internet access, along with Bluetooth.

The Dynabook Chromebook C70 is currently available exclusively in Japan, but hopefully there will be a global launch soon.

You may also like

• These are the best Chromebooks
• Take a look at the best laptops for programming
• Plucky French startup takes on Apple with cheap upgradable storage for the Mac Studio

Stay cool in the heat by using a cooling tower fan -- no air conditioning necessary!

• Nvidia CPUs designed for scalable solutions with PCIe connectivity
• Delivers 144 cores and four GPUs for AI acceleration
• 1.3 TB memory makes NVL4 ideal for data-heavy applications

Nvidia has unveiled the GB200 NVL4, an advanced platform designed to meet the needs of modern data centers and computational workloads.

The company's decision to introduce the GB200 NVL4 comes after Nvidia recently restructured its product lineup, setting aside other NVL platforms to favor newer options like the NVL4.

The GB200 NVL4 combines two Arm-based Grace CPUs with four Blackwell GPUs. It is part of Nvidia’s Grace Blackwell lineup, which aims to provide flexible and scalable solutions for diverse deployment scenarios.

High-speed and scalable

The GB200 NVL4 is designed as a mid-range option in the Grace Blackwell family, sitting below high-performance configurations like the NVL72 and NVL36. Each unit features two Grace CPUs, each equipped with 72 Arm Neoverse V2 cores, for a total of 144 cores. The addition of four Blackwell GPUs ensures robust acceleration for AI, HPC, and other compute-intensive tasks.

The six MCIO connectors beneath each CPU provide PCIe connectivity, allowing for high-speed data transfer. It also supports the inclusion of NICs, SSDs, and other essential components while maintaining efficiency.

Power consumption is a critical factor in modern data centers, but Nvidia estimates the GB200 NVL4 will consume just over 6 kW per server when fully configured. While this represents a substantial power draw, it is a marked improvement over larger platforms, with earlier systems like the Nvidia DGX-1 or HGX-1 consuming around 3.5 kW.

Furthermore, this device also supports up to 1.3 TB of combined memory, ensuring efficient data handling and processing while making it a strong candidate for memory-intensive applications. The NVL4 is expected to fill a niche for organizations seeking powerful yet relatively lower-power solutions compared to the higher GPU count NVL platforms.

The GB200 NVL4 and the higher-end NVL72 may appear similar at first glance, however, there are notable differences. The NVL72 features edge connectors for its spine configuration and focuses on larger-scale deployments with double the compute resources. In contrast, the NVL4 is designed as a more compact and power-conscious node while still delivering significant performance for its size.

You may also like

• UK data centers ready for expansion
• These are the best Nvidia graphics cards around today
• The race to build data centers is on - here’s how we keep them secure

So Google wants me to start saying ‘Hey Gemini’ now, huh? No thanks, you can get in the sea with that nonsense. I’m not having it. Call me a Luddite, call me a curmudgeon, tell me to get with the times; I couldn’t care less, I’m not going to talk to my tech.

Now, before I get into the meat and potatoes of this article, I’d like to preface it by saying that I’m not against the existence of voice control features on the whole. They’re actually an extremely vital accessibility feature that many disabled tech users rely on to get the full experience from their hardware. But for those who don’t actually need it, like myself - what the hell is wrong with just pressing some buttons or tapping a touchscreen?

I get annoyed if someone is talking too loudly on their phone on public transport. When tech companies like Google tell me that voice control is the future of how we interact with our tech, I’m immediately filled with horror at the idea of traveling through a city where everybody is constantly barking commands at their phones and tablets.

How many people really use voice controls?

I did some research into the actual statistics behind voice control use, and was surprised at the results. I’ve literally never seen a single person use their phone to search the web for something using a voice command; sure, I’ve seen people ask their Alexa smart speaker to play music or turn off a light, something I will probably also never do because I always have a phone in my pocket that can do those things, but web searches? Really?

Apparently so: according to a 2018 study by PWC, 32% of voice assistant users ask their chosen digital helper at least one thing they’d normally use a search engine for on a daily basis, with 89% doing so at least once a month. Of course, that’s only people who already use a voice assistant, but analysis from Statista claims that almost half of Americans talk to their phones or smart speakers at least semi-regularly (though that figure reduces to about 1 in 5 on a global scale).

Speak to the orb, says Amazon. The orb is always listening. The orb hears all. (Image credit: Amazon)

The thing is, as I dug further and further into these statistics, I became less and less convinced by them. For starters, the very first set of stats I came across (which I won’t link here) claimed that “8.4 billion people worldwide are estimated to use voice assistants” - that’s… more than the current total human population. I started noticing more discrepancies in the data, as well as having to discard some sources for obvious pro-tech-marketing bias.

More confused than enlightened, I had to eventually conclude that much of the statistical research into this area of tech has been based more heavily on product sales than actual unbiased polling of the population: and that’s a serious flaw, because a person who owns one piece of voice-controlled hardware is likely to own more. I have a friend who has three identical Echo Dot smart speakers positioned in different rooms around her home, and she uses Siri on her iPhone to make music requests while in the car. Me? I just have a driving playlist that I shuffle before I start the engine.

Voice control is getting better - slowly

I will admit that my usual excuse for why I abhor voice-controlled tech doesn’t hold as much weight as it used to. That excuse was, in short: it’s crap. The early days of Siri, Cortana, and their ilk were plagued by a constant refrain of “I’m sorry, I didn’t quite understand that”, but with the dawn of AI, things are starting to improve.

Tools like Apple Intelligence and Google Gemini offer multimodal input, allowing them to understand vocal requests as well as text prompts. The large language model AIs of today do a far better job of parsing spoken words than older voice-recognition software, even able to adapt to an individual user’s speech patterns over time to provide more accurate responses.

Google Gemini has far more potential as a pocket-sized voice assistant than Cortana ever did. (Image credit: Google)

However, there are still stumbling blocks to be overcome. While voice recognition typically supports multiple languages, it frequently struggles with strong accents and speech impediments (I myself have a lisp, which doesn’t help matters). This can be due to unnoticed biases in the training data used: if an American company uses recordings of Americans speaking English to train its speech recognition AI to understand spoken English, it’s unsurprisingly going to struggle when it hears a Japanese or Swedish person speaking that language.

I do genuinely hope that one day voice controls work perfectly because the people who really need them deserve a service that works as well as simply typing a query into Google. But I won’t be using it, and I don’t want to live in a future where everybody is - you can bet I’ll be first in line to dunk on any tech company that tries to make voice commands the default mode of interacting with their product.

NFL action continues on Netflix on Christmas Day with the second game featuring the Ravens and Texans and a Beyoncé halftime show.