Feed aggregator

Jump to your region

US Nvidia RTX 5080 stock

UK Nvidia RTX 5080 stock

Today's the day! The Nvidia GeForce RTX 5080 goes on sale today, January 30, and stock is expected to sell out fast - and we're here to help you find the best places to buy the Nvidia RTX 5080.

The Nvidia RTX 5080 will go on sale on January 30 at 6AM PT / 9AM ET / 2PM GMT with an MSRP of $999 in the US and £979 in the UK.

In our review of the Nvidia RTX 5080, we were impressed with the performance of the GPU, with it almost hitting RTX 4090 levels of performance (a GPU that cost a lot more money when it launched), though the generational leap from the RTX 4080 and RTX 4080 Super to the RTX 5080 isn't the biggest.

However, with hardware improvements in the form of more CUDA cores, fourth-gen RT cores, fifth-gen Tensor cores, and 16GB of GDDR7 memory, the RTX 5080 is a brilliant graphics card for people looking to play the latest PC games at 4K, especially with the new features included with DLSS 4, Nvidia's AI-powered image rendering software.

There's a lot of hype around the RTX 5080 (and its bigger sibling, the RTX 5090, which we also show you where to buy), and that means there could be a real rush when the RTX 5080 goes on sale, especially as we've been hearing rumors that stock levels could be low.

So, finding out where to buy the Nvidia RTX 5080 during this frenzy could be difficult, which is where we come in. We'll be keeping tabs on all the major retailers in the US and UK and will be updating this page live to point you towards where you can buy the GPU - without being ripped off by scalpers.

WHERE TO BUY RTX 5080: US QUICK LINKS

• Nvidia: Buy direct from Nvidia (including Founders Edition and third-party GPUs)
• Amazon: Major retailer will hopefully have stock
• Best Buy: Founders Edition will be on sale
• Newegg: Founders Edition will go on sale soon
• B&H: Preorders start at 9:00am ET, Thursday January 30
• Micro Center: Retailer will have stock on January 30
• Adorama: Worth checking for stock when sales go live

WHERE TO BUY RTX 5080: UK QUICK LINKS

• Nvidia: Buy Founders Edition and third-party GPUs direct
• Amazon: Should have a wide range of RTX 5080 models
• Ebuyer: Already lists RTX 5080 models but no pre-orders (yet)
• Scan: Get notified when RTX 5080 GPUs go on sale
• Overclockers: Respected retailer should have stock on launch day
• Box: Another well-known store for PC components should have stock
• EE: The mobile network operator is getting in on the act, selling RTX 5080 on a first-come first-serve basis
• CCL: RTX 5080 GPUs from Palit, Gigabyte and more
• Currys: Lots of third party options starting at £1,139.99

WHERE TO BUY RTX 5080 IN THE US

Nvidia RTX 5080 deals at Amazon
Amazon is expected to be a reliable source of RTX 5080 stock as with nearly everything else - that's the good news. The bad news is that everyone else will think to check Amazon, so I expect stock to be dwindling or already sold out very soon after launch. You will likely find some options if you don't mind shelling out a bit extra for a third-party GPU, though.View Deal

Nvidia RTX 5080 deals at Nvidia
Listed for $999, the RTX 5080 FE stock on Nvidia's website is sure to sell out particularly quickly. If your plan was to go directly to the source, I'd get in now before scalpers and resellers take over. However, if you're able to snag one, the FE cards are at a reasonable price, feature a sleek design, and are even small enough to fit into an SFF build.View Deal

Nvidia RTX 5080 deals at Best Buy
Best Buy is similar to Amazon in that you can expect a solid stock of Nvidia's new 5000 series cards, but they'll be flying off the shelves and into carts without a second look at the price tag.View Deal

Nvidia RTX 5080 deals at Newegg
Newegg, a personal favorite tech retailer of mine, is a fantastic location to find upgraded 5080s from Asus, Zotac, MSI, and more. Newegg is also offering trade-in value on your current GPU towards the new card, so if you want to get some value from your 30 series card, this is an excellent opportunity.View Deal

Nvidia RTX 5080 deals at B&H
B&H is another solid retailer to search for third-party RTX 5080 deals. It just added its 5080 stock to pre-order pages within the last few days, so I believe it'll have a decent backlog of MSI, Asus, and Gigabyte 5080s to sling around come January 30.View Deal

WHERE TO BUY RTX 5080 IN THE UK

Not to worry UK folks, Nvidia hasn't forgotten about you - though stock is proving to be just as hard to get there as well. Here's what we dug up from retailers like Ebuyer and Box:

Nvidia RTX 5080 deals at Nvidia
Nvidia has stock for the RTX 5080 on its UK website as well, but like the US counterpart, stock is moving very quickly. Founders Edition cards will be selling like hotcakes, but you can also find third-party GPUs on Nvidia's website.View Deal

Nvidia RTX 5080 stock at Overclockers
Overclockers UK is an excellent site for buying PC components, and it has already listed its 5080 stock, and there are a fair number of cards on offer with prices starting at £1,099.99.View Deal

Nvidia RTX 5080 stock at Ebuyer
Ebuyer is another fantastic site to check for new PC components, and the 5080 launch is no exception. Like everywhere else, be sure to check back in on January 30 to assess the available stock.View Deal

Nvidia RTX 5080 stock at Scan
Scan also just populated its GPU pages with swathes of RTX 5080 cards from Gigabyte, Asus, Zotac, and MSI. For now, the only thing you can do is get notified when they're available, but be sure to check back first thing January 30.View Deal

Nvidia RTX 5080 stock at EE
While EE is better known for selling smartphones and tarrifs, it's also going to be selling the RTX 5080 on January 30 on a first-come-first-serve basis from £1,245.99.View Deal

Nvidia RTX 5080 stock at Currys
Currys has a wide range of RTX 5080 GPUs from third party makers such as Asus and Palit. Currently, the Asus GeForce RTX 5080 16GB Prime OC is the cheapest option at £1,139.99.View Deal

Nvidia RTX 5080 stock at CCL
CCL is a dependable retailer for computing components and is offering plenty of third party RTX 5080 models. Stock will also be limited to one per customer, which will hopefully slow down how fast these sell out today.View Deal

• These are the best graphics cards of 2025

(Image credit: Future)

Launch day is finally here! We'll be here all day, monitoring stock at all the major retailers and beyond to try and help you snag an RTX 5080.

It looks like RTX 5090 stock is going to vanish in a flash, so you might have better luck than those hunting for Nvidia's new flagship GPU, but there's no guarantee this one will be easier to pick up...

We're still hours away from the official stock drop, but Micro Center's RTX 5000 page has seemingly gone down, returning a 403 Forbidden error at the time of writing. Either too many people were visiting the page, or Micro Center has taken it offline intentionally to prep it for sales...

As we approach the hour when the RTX 5080 goes on sale, I have to say that stock levels are looking rather dicey.

MSI has issued a statement revealing that it'll only be opening pre-orders for the RTX 5090 today, with actual cards shipping on February 6 instead. However, the RTX 5080 will be available on MSI's online storefront.

(Image credit: Temporary-Director46)

Some eager shoppers are even camped out in front of physical stores in the US to try and snag one of these highly prized GPUs. The image above is from a Reddit post showing people camping four days before the launch.

While I can't stop you from doing that, I would personally recommend avoiding the cold and just bookmarking this page...

(Image credit: Future)

While we all wait for the Nvidia RTX 5080 to go on sale in a few hours, make sure you read our Nvidia GeForce RTX 5080 review by our components editor John Loeffler. John's been reviewing GPUs for us for many years now, and there's no one's opinions on graphics cards I respect more, so definitely take a look. He was impressed, giving the new GPU a very good four and a half stars.

Speaking of our RTX 5080 review, definitely dive into the relative performance of the best graphics cards on the market and see whether the RTX 5080 is the right card for you.

If you're seriously considering upgrading to Nvidia's RTX 5080, I'd head to one of the websites that lets you get a notification as soon as stock drops in - these being Nvidia, Best Buy and B&H Photo in the US, and Nvidia, Scan, Overclockers, and Box in the UK (links above) - so you can try your best to beat the scalpers and resellers.

If you live near a Micro Center, you might have a good chance of picking a 5080 up in person. It's offering a voucher system on a first-come first-serve basis that guarantees you'll be able to purchase a GPU on launch day (though it doesn't guarantee a specific model or manufacturer). Micro Center stores also have a rule in place that allows only one 5080 or 5090 per customer, meaning if you don't mind waiting in line, this could be a good chance to pick up a Blackwell card the day it comes out.

If you'd prefer to try your luck at buying one online, Newegg is a great place to look. It's offering a trade-in deal on your current GPU towards a new 5000 series card. Just for fun, I looked at how much I'd get if I traded my (fully-functioning, undamaged) MSI Gaming GeForce RTX 3080 Ti in - $419 towards a new GPU isn't shabby at all.

(Image credit: Nvidia)

You can't talk about Nvidia's new graphics cards without talking about AI and DLSS, can you?

DLSS is the tech that allows Nvidia to tout results like the new $549 5070 performing as well as 2022's $1,599 RTX 4090 - it's also cause for heated debate in the PC gaming world, as enthusiasts are divided on whether "fake frames" should be considered in performance metrics.

While all RTX cards (including the 20, 30, and 40 series GPUs) will be able to utilize most of DLSS 4's features, as of right now only the new 50 series cards will benefit from the latest improvements like Multi Frame Generation (though it looks like Nvidia may be bringing MFG to older cards).

If you read through our Nvidia RTX 5090 review, you'll find "In non-upscaled 1440p gaming, you're looking at a roughly 18% better average frame rate and a 22.6% better minimum/1% framerate for the RTX 5090." over the 4090 - so it seems DLSS isn't everything the 5000 series has to offer.

(Image credit: Future)

Oof, we've just published some worrying news - some sources are saying that Nvidia RTX 5090 GPU stock will be ‘basically non-existent’ for launch day, and RTX 5080 doesn’t sound like it’ll be much better.

It seems retailers could be getting less stock from Nvidia than with the RTX 4080 launch. While RTX 5080 GPUs from third party makers like Asus and Gigabyte might be easier to get hold of, this latest news, combined with recent rumors, suggests it could really be a struggle to find stock - which is why this guide will be so useful.

(Image credit: Shutterstock / rudall30 / Asus)

If you're thinking of buying the RTX 5080 today, then you'll also want to make sure that the rest of your PC is able to support it.

One consideration is the PSU (Power Supply Unit), as you need something that will be able to cope with the demands of the RTX 5080, as well as powering the other parts of your PC as well.

Asus ROG's RTX 5080 page lists the ROG THOR III 1600W (or 1200W at least) power supply as the "perfect PSU companion" to the newly-listed ROG Astral GeForce RTX 5080 16GB GDDR7 OC Edition graphics card - now that's a lot of power.

This version of the 5080 also comes stocked with not two, or even three, but four fans along with rows of heatsink fins to keep the heat from slowing down clock speeds, and 16GB of Nvidia's new GDDR7 video memory will definitely be a sight to behold.

Check out our guide on the best PSUs for more options.

If you're in the UK and near Bolton, you could head over to Scan Computers, which is holding a launch day event that will have RTX 5080 and RTX 5090 GPUs for buying in-person.

Sure, not the most convenient way of getting a new GPU, especially if you don't live near Scan Computers (which is also selling the GPUs online), but if you can get there, it could be a good way to make sure you get an RTX 5080.

You can find out more about the Scan Nvidia 50-series launch event here.

(Image credit: Shutterstock)

We're approaching the final countdown, with RTX 5080 stock going on sale in just over an hour's time. Now is a good time to get retailers loaded up.

Posted by OverclockersUK on 

Overclockers UK's Facebook page is hyping up the launch - looks like they will have lots of stock! But how long will it last?

Checking Amazon US and there's no placeholder RTX 5080 GPUs yet. However, there are pre-built PCs with the RTX 5080 in its listings. This might be a better way of getting hold of a GPU, especially if you need to upgrade other parts of your PC.

• CyberPowerPC Gamer Supreme Gaming PC, AMD Ryzen 9 9900X 4.4GHz, GeForce RTX 5080 16GB, 32GB DDR5, 2TB PCIe 4.0 SSD - $2,299.99
• CyberPowerPC Gamer Xtreme VR Gaming PC, Intel Core Ultra 7 265KF 3.9GHz, GeForce RTX 5080 16GB, 32GB DDR5 - $2,319.99
• CyberPowerPC Gamer Xtreme VR Gaming PC, Intel Core Ultra 9 285K 3.7GHz, GeForce RTX 5080 16GB, 64GB DDR5 - $2,699.99
• CyberPowerPC Gamer Supreme Gaming PC, AMD Ryzen 9 9950X 4.3GHz, GeForce RTX 5080 16GB, 64GB DDR5, 4TB PCIe 4.0 SSD - $2,769.99

Over in the UK, we can see Overclockers UK's RTX 5080 models and their prices. While we're still around 55 minutes away from being able to buy them, you can at least get an idea of what kind of price you'll be paying for the new GPU.

(Image credit: Zotac)

The most affordable RTX 5080 model from Overclockers UK looks like it'll be the Zotac GeForce RTX 5080 Solid, which is £1,099.99.

Next cheapest looks like they'll be the Gainward GeForce RTX 5080 Phoenix and Palit GeForce RTX 5080 Gaming Pro which will both go on sale in around 50 minutes for £1,159.99.

For US readers looking to buy the RTX 5080 from Best Buy, you'll need to make sure you have a Best Buy account ahead of the GPU going on sale, including providing location details. It's best to get this sorted now, so you're in with a better chance of getting hold of an RTX 5080 when things kick off in about 45 minutes!

(Image credit: Future)

Speaking of Best Buy, it might be the best place to get hold of an affordable RTX 5080, as it is selling the Founders Edition RTX 5080 for $999.99.

Remember, if you're looking for the RTX 5090, we have a dedicated where to buy the RTX 5090 live guide as well!

We're now 30 minutes away from the RTX 5080 (and the RTX 5090) going on sale, and it looks like there's a lot of excitement building about this launch.

Interesting fact from the Overclockers UK page, which says that "Pre-Orders for this product will begin after 2pm on the 30th January. Models will be put on sale at varying intervals after this time."

So we might see a staggered release of stock, so if things sell out quickly in 25 minute's time, there might still be a chance that some models will be available to buy later on.

Just under 10 minutes to go! Use this time to make sure you're signed into any retailers you're trying to buy from, as this will speed up the process when checking out - and every second could count!

OK stock should now be on sale! Good luck!

Nvidia's page hasn't updated for the RTX 5080 going on sale yet, so I'm keeping an eye on that...

Best Buy has the RTX 5090 Founders Edition in stock for $1,999.99!

Oh! It's now saying sold out, but there's other options like the GIGABYTE - AORUS GeForce RTX 5090 MASTER ICE 32G - White which is 'coming soon'

Looks like Best Buy's stock is all selling out very fast, so best check other retailers.

Looks like Overclockers UK website has crashed under the pressure of people trying to buy the RTX 5080 - I'll keep you posted. Also in the US, it seems stock has already sold out.

OK UK buyers! It looks like Scan has got stock!

All in stock:

• MSI NVIDIA GeForce RTX 5080 16GB VENTUS 3X OC Blackwell Graphics Card - £979
• ZOTAC NVIDIA GeForce RTX 5080 SOLID 16GB Blackwell Graphics Card - £979
• PALIT NVIDIA GeForce RTX 5080 16GB GAMING PRO Blackwell Graphics Card - £979

Microcenter seems to have stock for the RTX 5080 for buying in store, which could be a good way to get hold of the new GPU, though you'll need to be near a store with stock.

If you have the budget, one way of getting an RTX 5080 is to buy a pre-built PC with it in - this can sometimes be a good choice if you need to upgrade lots of components in your PC. Adorama has sold out of standalone RTX 5080s, but it has the CyberPowerPC Gamer Supreme Liquid Cooled Gaming Desktop Computer, AMD Ryzen 7 9800X3D 4.7GHz, 32GB RAM, 2TB SSD, NVIDIA GeForce RTX 5080 16GB, Windows 11 Home, Black for $2,419.99.

You're essentially paying an extra $1,400 for some pretty decent components.

Overclockers UK website is still having issues - sometimes it's up, then it crashes. If you can get through, it might be a good bet as stock sells out elsewhere.

Overclockers also has some pre-built PCs with the RTX 5080 GPU for sale and in stock, if you really want to get hold of the GPU.

In the UK, Novatech will be selling RTX 5080 GPUs, though it doesn't say when. It could be a retailer that gets the next batch of stock when it's made available.

Still no sign of any RTX 5090 on Amazon's US site.

STOCK ALERT!

Overclockers UK is back up and has RTX 5080 GPUs in stock and at decent prices!

• Palit GeForce RTX 5080 GameRock 16GB GDDR7 PCI-Express Graphics Card - £1,299
• MSI GeForce RTX 5080 Gaming Trio OC 16GB GDDR7 PCI-Express Graphics Card - £1,469.99

If the above say out of stock, go back to the RTX 5080 product page and click 'Add to basket' it seems to work!

STOCK UPDATE!

CCL has RTX 5080 GPUs in stock and delivering from tomorrow! Be quick as it might sell out (it's UK only I'm afraid).

• Palit GeForce RTX 5080 GameRock 16GB GDDR7 Graphics Card - £1,349.99
• INNO3D GeForce RTX 5080 X3 16GB GDDR7 Graphics Card - £1,349.99
• INNO3D GeForce RTX 5080 X3 OC 16GB GDDR7 Graphics Card - £1,399.99

• Palit GeForce RTX 5080 GameRock OC 16GB GDDR7 Graphics Card - £1,399.99
• Gigabyte GeForce RTX 5080 WINDFORCE SFF OC 16GB GDDR7 Graphics Card - £1,399.99
• Gigabyte GeForce RTX 5080 AERO SFF OC 16GB GDDR7 Graphics Card - £1,399.99
• Gigabyte GeForce RTX 5080 GAMING OC 16GB GDDR7 Graphics Card - £1,499.99

More RTX 5080 UK stock at CCL:

• MSI GeForce RTX 5080 VENTUS 3X PLUS 16GB OC GPU - £1,499.99
• MSI GeForce RTX 5080 GAMING TRIO 16GB OC GPU - £1,499.99
• MSI GeForce RTX 5080 GAMING TRIO 16GB OC GPU - £1,499.99
• ASUS GeForce RTX 5080 TUF Gaming 16GB OC GPU - £1,499.99
• MSI GeForce RTX 5080 GAMING TRIO WHITE 16GB OC GPU - £1,559.99
• ASUS GeForce RTX 5080 ROG Astral 16GB OC GPU - £1,599.99
• Gigabyte GeForce RTX 5080 AORUS MASTER 16GB OC GPU - £1,699.99
• Gigabyte GeForce RTX 5080 AORUS MASTER ICE 16GB OC GPU - £1,699.99
• Gigabyte GeForce RTX 5080 Xtreme Waterforce 16GB OC GPU - £1,799.99

Take advantage of high APYs while they last. Rate cuts are expected later this year.

Midco offers some of the best fiber broadband in North Dakota, but there are other options too. CNET helps you pick out the best internet service provider for your needs in Fargo.

Augusta Robbins sits on the floor drawing at the temporary Airbnb her family was staying in. Since her family home burned down, she's been drawing a lot of houses.'/>

The historic fires in Los Angeles meant many students lost homes, schools, or both. Now, as some schools reopen, families are trying to find routine.

(Image credit: Jireh Deng for NPR)

One of President Trump's first executive orders impacts transgender inmates incarcerated in federal facilities. It's just one in a series of actions signed by Trump that target the trans community.

(Image credit: Ben Curtis)

There's still time to earn interest, but there's no guarantee that these rates will stick around.

These rumbly motor-powered sex toys—for all bodies and genders—promise good vibes for all.

Reflecting on a transformative residency program, the jazz vocalist now nominated for her first Grammy Award says her album Journey in Black reflects "the expansiveness of the Black experience."

(Image credit: Jared Soares for NPR)

President Trump has charted a new course for electric vehicle policy in the U.S. While the direction is clear, the details aren't.

(Image credit: Bill Pugliano)

Actor, writer and director Jesse Eisenberg says he has had more failures than successes. In this week's Wild Card, he opens up about ambition and his his defense against despair.

(Image credit: Gareth Cattermole)

Dozens of Jan. 6 defendants who received pardons from President Trump had past criminal convictions for charges including rape, manslaughter, domestic violence and drug trafficking.

(Image credit: Julio Cortez)

Among the topics likely to come up at the hearing: her 2017 meeting with Syrian President Bashar al-Assad, her friendly positions toward Russia and her push to drop charges against Edward Snowden.

(Image credit: Kevin Dietsch)

Kash Patel, President Trump's pick to run the FBI, answers questions Thursday in a public hearing before the Senate Judiciary Committee.

(Image credit: Anna Moneymaker)

The rise of Generative AI, from tools like ChatGPT to a new wave of AI tools, is reshaping technology demands, driving a surge in the need for advanced CPUs and GPUs. These powerful processors, with ever-increasing thermal design power (TDP)—like Nvidia’s GB200 Grace Blackwell superchip consuming 1,200W per GPU with racks powering 120kW—are pushing power and cooling requirements to unprecedented levels, making traditional air cooling solutions insufficient.

Enter liquid cooling. With adoption projected to grow at a 45% compound annual growth rate (CAGR) from 2023 to 2028, direct-to-chip liquid cooling solutions, powered by coolant distribution units (CDUs), are transforming how data centers handle the heat generated by high-performance AI workloads. Cooling efficiency isn’t just a consideration—it’s a necessity as AI processors continue to push power and density boundaries.

This article examines how CDUs are supporting thermal heat management, enabling operators to meet the relentless cooling demands of AI and high-performance computing while paving the way for the next generation of data center efficiency.

What is a Coolant Distribution Unit (CDU)?

A CDU is the heartbeat of a liquid cooling system, carefully regulating coolant temperature and flow rates to maintain optimal cooling efficiency. By managing the coolant flow to IT equipment and returning the IT heat to the facility’s water for re-cooling, CDUs help stabilize temperatures and minimize the risk of overheating. In direct liquid cooling, the CDU plays a vital role by allowing for temperature conditioning to prevent condensation and by isolating the IT equipment from harsher facility water, which may contain mineral deposits, particulates, and other impurities that could damage cooling systems or reduce efficiency.

CDUs are also pivotal for increasing system longevity. According to a recent Uptime Institute study, over 70% of unplanned downtime in data centers is linked to power and cooling system failures. With advanced CDUs, operators can reduce such risks, ensuring reliable operations even under heavy workloads.

Different Types of CDUs: Liquid-to-Liquid vs Liquid-to-Air

Liquid-to-Liquid and Liquid-to-Air CDUs both serve the essential function of cooling IT equipment but are suited for different environments based on their cooling mechanisms and efficiency levels. Liquid-to-Liquid CDUs are ideal for facilities with access to facility water, offering high cooling capacity and efficiency due to the superior thermal conductivity of water.

These systems transfer heat from the IT equipment’s coolant loop to the facility’s water loop using a heat exchanger, which is well-suited for high-density environments. In contrast, Liquid-to-Air CDUs use air-cooled radiators and fan systems to dissipate heat into the surrounding air, making them a better choice for locations without access to facility water, though they typically offer lower overall cooling capacity and efficiency.

While both systems include similar components needed to run (such as pumps, temperature control systems, and filtration), the key difference lies in how they transfer heat. Liquid-to-liquid CDUs rely on primary and secondary pumps to circulate coolant and water, while Liquid-to-Air CDUs depend on fans to move air over radiators. Maintenance for both systems is essential, though Liquid-to-Air CDUs may require more frequent attention to air filters and fan components.

In contrast, liquid-to-liquid systems require monitoring the water loop and its cleanliness. Ultimately, the choice between a Liquid-to-Liquid or Liquid-to-Air CDU depends on the available infrastructure and specific cooling and efficiency goals, with Liquid-to-Liquid systems excelling in high-density environments and Liquid-to-Air systems offering flexibility where water access is limited.

Industry Trends and Future Directions

The adoption of liquid cooling is on a sharp upward trajectory. In 2023, only 10% of data centers used liquid cooling, but this figure is expected to reach 50% by 2030. Driving this shift are the increasing thermal demands of AI and HPC and environmental considerations. Traditional air cooling can consume up to 40% of a data center’s energy, with goals for advanced cooling to reduce energy expenditures to as little as 5% of total IT load. This is a significant factor in reducing overall Power Usage Effectiveness (PUE).

CDUs are also addressing water conservation challenges. Records obtained by the Financial Times reveal that data center water consumption has spiked by nearly two-thirds since 2019, with over 1.85 billion gallons consumed in 2023 compared to 1.13 billion gallons in 2019. Technologies that eliminate reliance on evaporative cooling are becoming indispensable. Eco-friendly coolants and closed-loop systems are gaining traction, helping operators reduce their environmental footprint.

Conclusion

As AI and HPC workloads intensify, data centers must adopt cutting-edge solutions like CDUs to stay ahead. By enabling efficient, scalable cooling, CDUs not only support operational demands but also align with sustainability goals. With liquid cooling expected to dominate the market by 2030 and the data center cooling market set to exceed $20 billion by 2028, the time to embrace these technologies is now. CDUs are not just a tool for cooling—they are the foundation for a sustainable, high-performance future.

We list the best web hosting services.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

The cyber threat landscape has reached a critical tipping point.

According to the UK government's 2024 Cyber Security Breaches Survey, a staggering 50% of businesses experienced some form of cyber breach in the last 12 months, with this figure rising dramatically to 70% for medium businesses and 74% for large businesses.

Phishing attacks dominate the threat landscape, accounting for 84% of business breaches, followed by email impersonation (35%) and malware (17%).

The statistics, while alarming, reveal only part of the challenge facing organizations today. The most pressing issue isn't just the increasing frequency of attacks, but also the growing disparity between how quickly attackers can breach systems and how long organizations take to respond.

Contemporary security technologies can detect threats within minutes, yet the average time for organizations to fully identify, contain and restore systems stretches to about 20 days – with recovery times far longer. This extended vulnerability window gives cybercriminals ample time to infiltrate networks, compromise sensitive data and even establish a backdoor for future attacks.

Recent headlines have highlighted the devastating impact of delayed response times across various sectors. From the UK Air Traffic Control's miscommunicated cyber incident last year to UnitedHealth's delayed response to a massive data leak in April this year, as well as ongoing challenges faced by British Ambulance Services and Nuclear plant Sellafield, the impact of inadequate response times continues to be felt.

These incidents underscore a troubling reality. When organizations cannot respond swiftly to cyber threats, the consequences ripple far beyond immediate operational disruption. The financial toll is substantial – IBM reports a 10% increase on the cost of an average data breach in 2024, rising to $4.8million.

The evolution of cybersecurity tools

That said, the cybersecurity industry has made remarkable strides in developing defensive technologies, yet many organizations struggle to maximize their potential.

Modern Extended Detection and Response (XDR) platforms represent a significant advancement, offering sophisticated threat detection and automated response capabilities that can identify and neutralize threats across an organization's entire IT infrastructure.

The latest generation of security tools also incorporates predictive capabilities, leveraging vast databases of threat intelligence to anticipate and prevent attacks before they materialize. These systems can link seemingly unrelated events across different parts of the network – in doing so, subtle patterns that might indicate an emerging threat can be identified – a key part in taking detection timelines from days to hours.

This evolution from reactive to proactive defense represents a crucial step forward in closing the response time gap. However, the form of data remains crucial. Too often, we see organizations dealing with theoretical data as opposed to actual, real-time information. Relying on the former may prove effective in theory, in practice, it’s a different story altogether. No one organization's defense is the same.

Building a cyber safe culture

Indeed, creating an effective cyber defense requires more than deploying the latest security tools – it demands a fundamental shift in organizational culture.

Security posture assessments need to become an ongoing process rather than a periodic checkbox exercise. By continuously evaluating and adjusting defenses, organizations can identify and address vulnerabilities before attackers have the chance to exploit them. The integration of artificial intelligence and machine learning capabilities has become key to this effort, not least because it has reduced time needed to spot and investigate potential threats but also the ability to bring contextual data into play, allowing a more informed response.

Best practices for rapid response

Indeed, a robust cybersecurity strategy must seamlessly integrate people, processes and technology.

Security teams require immediate access to clear and actionable threat intelligence through intuitive interfaces that support rapid decision-making. Protection must extend across the entire attack surface, from cloud infrastructure to remote work endpoints, to create a unified defense against increasingly sophisticated threats.

Modern security platforms can automate initial containment measures, which will buy precious time for security analysts to investigate and respond to incidents. However, technology must be supported by clear protocols for incident communication and stakeholder coordination. While building these defenses requires significant investment, the potential costs of a serious breach can be markedly higher – both in immediate financial terms and long-term reputational damage.

The most effective rapid response strategies now incorporate real-time monitoring of the complete environment. In the most effective cases, this monitoring is bolstered by strong detection and response processes, which provide the correct amount of insights into each individual risk and the damage it is capable of causing. After, cyber teams can quickly understand the scope and nature of any security incident, facilitating faster and more targeted responses.

Looking ahead

A proactive security posture, supported by continuous adaptation and improvement, has become essential for survival. This means not only keeping pace with emerging threats but anticipating and preparing for tomorrow's challenges. It means being cyber safe – not just cyber secure. The current gap between attacker capabilities and defender response times represents one of the most pressing challenges in modern cybersecurity. However, organizations that combine cutting-edge mindset backed by the right technology with robust processes and a cyber safety-conscious culture can work to close this gap. The objective isn't merely to catch up with cyber attackers – it is to stay ahead of them.

Checkout our list of the best identity management software.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

The world of clitoral suction sex toys is getting bigger every day. We’ve tested dozens, and these are our favorites.

The executive order directs government agencies to use all available tools to prosecute or remove perpetrators of antisemitic harassment and violence, especially on college campuses.

(Image credit: Andrew Harnik)

The president says up to 30,000 criminal migrants deported from the United States could be housed at the facility in Cuba, but it wasn't immediately clear how the plan would be implemented.

(Image credit: Ramon Espinosa)

More than one in 10 US adults are already behind on their energy bills. Here's what you can do if you're struggling with high energy costs.

• A new approach called DualPipe seems to be the key to DeekSeek's success
• One expert describes it as an on-GPU virtual DPU that maximizes bandwidth efficiency
• While DeepSeek has used Nvidia GPUs only, one wonders how AMD's Instinct would fare

China’s DeepSeek AI chatbot has stunned the tech industry, representing a credible alternative to OpenAI’s ChatGPT at a fraction of the cost.

A recent paper revealed DeepSeek V3 was trained on a cluster of 2,048 Nvidia H800 GPUs – crippled versions of the H100 (we can only imagine how much more powerful it would be running on AMD Instinct accelerators!). It reportedly required 2.79 million GPU-hours for pretraining, fine-tuning on 14.8 trillion tokens, and cost - according to calculations made by The Next Platform - a mere $5.58 million.

But exactly how DeepSeek's developers managed this feat is likely down to a clever hack.

A virtual DPU on the GPU itself

First, some background. DeepSeek is an advanced Mixture-of-Experts (MoE) language model designed to optimize performance by selectively activating only the most relevant parts of its architecture for each task. The third version of the model, DeepSeek-V3, features a total of 671 billion parameters, with only 37 billion activated for any given token prediction. This selective activation massively reduces computational costs while maintaining high performance and accuracy – which you’ll see if you try it.

It’s easy to be skeptical of DeepSeek and the claims made regarding its training, but the paper reveals some of the magic the developers came up with to make the most of the crippled hardware they had to work with. This includes the creation of the DualPipe algorithm for efficient pipeline parallelism.

According to the information published by DeepSeek, DualPipe overlaps forward and backward computation, reduces latency, and optimizes data movement across GPUs. By efficiently managing communication, it minimizes idle time (pipeline bubbles) and dynamically balances GPU compute cores (Streaming Multiprocessors) between computation and communication, preventing data transfer bottlenecks as the model scales.

A commenter on The Next Platform describes DualPipe as "essentially creating a virtual DPU on the GPU itself to handle all-to-all communication," which highlights its role in optimizing data transfer efficiency.

The paper goes into further detail, "In order to ensure sufficient computational performance for DualPipe, we customize efficient cross-node all-to-all communication kernels (including dispatching and combining) to conserve the number of SMs dedicated to communication. The implementation of the kernels is co-designed with the MoE gating algorithm and the network topology of our cluster. To be specific, in our cluster, cross-node GPUs are fully interconnected with IB, and intra-node communications are handled via NVLink.”

Example DualPipe scheduling for 8 PP ranks and 20 micro-batches in two directions. The micro-batches in the reverse direction are symmetric to those in the forward direction, so we omit their batch ID for illustration simplicity. Two cells enclosed by a shared black border have mutually overlapped computation and communication. (Image credit: DeekSeek) You might also like

• Check out our list of the best AI tools around today
• “This is a wake-up call" - the DeepSeek disruption: 10 experts weigh in
• DeepSeek forced to pause new signups following large scale cyberattack