I interviewed two Assassin’s Creed Shadows devs and asked what they’re excited for fans to experience
One dev focused on the recruitment in-game
Another highlighted the new way that Shadows is telling its story
Excitement for Assassin’s Creed Shadows is growing as we race toward the game’s March 20 release date.
I recently visited Ubisoft Quebec to go hands-on with the game and get a behind-the-scenes look at development. During my visit, I spoke to key developers and quizzed them on their favorite elements of the game that have not been spoken about much up until now but they were most excited for fans to see.
I first talked to associate game director Simon Lemay-Comtois, who identified something about the people you’ll find out in the world in Assassin’s Creed Shadows.
“I think the allies that you can recruit, like their individual stories and personalities, and the actors who play some of them is the coolest bit that we have not yet revealed fully.”
This relates to the party of extra characters you can collect in Shadows who will take up residence in the Hideout area of the game after you meet them.
Advertisement
Game Director, Charles Benoit, then looked towards the narrative when I asked him the same question, saying the most exciting thing we’re yet to see more of is “the way the stories unfold.”
“I think players will really like the story structure”, referring to the game’s use of flashbacks for both characters to understand their backstories, as well as the dual perspective on the world and narrative more generally.
Sign up for breaking news, reviews, opinion, top tech deals, and more.
“So you have a mix of exploring and doing the open world and then leaving [and seeing] past events of their life. And I think the way it’s structured is pretty cool and pretty different from other Assassin’s Creed games, so I can’t wait to see what people will think about that”.
Advertisement
The story is something I’m really keen to learn more about myself. Even after several hands-on hours with the game, and experiencing plenty of intrigue and interesting hooks, I’m still in the dark about what’s going on in the game.
Recent advances in generative AI and machine learning have evolved technological communication to levels beyond what was previously comprehensible. The recent and rapid rise of Large Language Models (LLMs) which are now analysing and generating huge volumes of text outputs, prompts us to consider our progress so far and what this means for the future of human communication alongside the new agentic AI layer.
The third wave of AI will bring us to a pivotal moment in the evolution of work, where the emergence of autonomous agents promises to transform how humans and machines collaborate. For us, agentic AI represents the next frontier in this progression; a limitless digital workforce built on AI agents that can reason, take action, and integrate deeply within enterprise workflows. This innovation not only builds on the foundation of LLMs, but introduces a new agentic layer, enabling machines to move from generating insights to orchestrating and executing actions.
In order to fully grasp the potential of this technology, it is helpful to look back at the origins of language and the emergence of technical languages, where computers interrupt instructions to take action.
Paul O’Sullivan
Advertisement
Prehistoric language and the roots of human communication
Language that we know and use today is the product of a long and intricate evolutionary journey. For early humans, language was believed to be primitive and simple, mostly involving gestures, facial expressions, and sounds to convey meaning.
The evolution to symbolic thinking, which refers to the ability to use concepts, symbols and signs in order to think about things not immediately present, marked a significant step change. Early evidence of symbolic thinking includes ancient cave art, which suggests a time where sounds and symbols were associated with abstract ideas.
Next came anatomical development, the evolution of vocal cords and the ability to produce complex sounds which paved the way for speech. Spoken language is believed to have emerged 50,000-100,000 years ago, giving humans valuable and unique skills such as sharing stories, preserving history, and ultimately, building complex societies.
Advertisement
The evolution of language to the rise of technology
The next transformative step in language development was the invention of writing. Around 3100 BCE, humans began using written symbols to document trade, laws, and ceremony. These early writing systems evolved into alphabets, such as the Phoenician alphabet, which influenced Greek and Latin scripts and laid the foundation for many modern languages.
Languages became more diversified as the human population spread and interacted globally. In today’s interconnected world, the technological revolution has introduced a new form of language: computer programming.
Early machine languages, such as binary and assembly code, required programmers to communicate directly with machines in highly technical terms. The advent of high-level languages, such as Fortran, made programming more accessible by using syntax that mirrored human language. Today, modern languages like Python and Java, along with domain-specific languages like SQL and HTML, cater to diverse needs, from app development to database management.
Looking to the future, natural language programming promises to bridge the gap between human and machine communication. By enabling users to write instructions in plain language, this innovation could democratize coding, making it accessible to those without traditional programming expertise.
Advertisement
The role of AI agents in shaping language
AI represents a definitive chapter in the evolution of language, where models are trained on vast datasets, generating human-like text and even introducing new expressions.
AI raises the possibility of machine-to-machine communication, with AI agents likely to be able to develop machine-specific languages for communication efficiency. This reflects a broader shift in how humans work alongside generative AI, with agents becoming part of the digital workforce re-emphasizes the need for AI practices grounded in trust and security.
Another impact of the interplay between human and machine communication is the emergence of hybrid languages, blending human syntax with machine-friendly structures. These changes could enhance accessibility, as we’ve already seen with the introduction of AI tools for real-time translation for those that are visually and hearing impaired.
In more speculative scenarios, AI might accelerate the creation of entirely new languages tailored to specific purposes, such as quantum computing and alternatively, humanity adopting intermediary language developed through AI.
Advertisement
Language is ever changing, regardless of technological impacts
Language is a living, evolving system that reflects the needs, culture, and technology of its time. From the primitive gestures of early humans to the sophisticated algorithms of modern AI, our modes of communication have always adapted to the challenges and opportunities of the era.
The rise of AI and machine learning represents a pivotal moment in the history of language, with the potential to accelerate its evolution in ways we can’t even imagine. While these advancements offer exciting possibilities, they also compel us to preserve the cultural and human dimensions of language.
As we stand at the threshold of a new era, our agentic AI technology, exemplifies how language and technology continue to converge, creating systems that not only understand but act on our behalf. By enabling a seamless collaboration between humans and autonomous agents, this evolution signals a future where communication becomes not just a medium for expression, but a catalyst for action, driving innovation and redefining the boundaries of what is possible in work and beyond.
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
U.S. authorities have indicted five people over their alleged involvement in a multi-year scheme that saw them obtain remote IT employment with dozens of American companies.
The Department of Justice on Thursday announced the indictment of North Korean citizens Jin Sung-Il and Pak Jin-Song; Pedro Ernesto Alonso De Los Reyes of Mexico, and U.S. nationals Erick Ntekereze Prince and Emanuel Ashtor.
The DOJ said the FBI arrested Ntekereze and Ashtor, and a search of Ashtor’s home in North Carolina found evidence of a “laptop farm” that hosted company-provided laptops to deceive organizations into thinking they had hired workers based in the U.S.
Alonso was also arrested in the Netherlands after a U.S. warrant was issued.
Advertisement
According to the indictment, Ntekereze and Ashtor allegedly installed remote access software, including Anydesk and TeamViewer, on the company-provided devices, allowing the North Koreans to conceal their locations. The two Americans also provided Jin and Pak with forged identity documents, including U.S. passports and U.S. bank accounts.
The indictment alleges that the defendants gained employment from at least 64 American organizations over the course of the multi-year scheme, which ran from April 2018 through August 2024. These included a U.S. financial institution, a San Francisco-based technology company, and a Palo Alto-headquartered IT organization.
According to the Justice Department, payments from ten of those companies generated at least $866,255 in revenue, most of which was laundered through a Chinese bank account.
“The Department of Justice remains committed to disrupting North Korea’s cyber-enabled sanctions-evading schemes, which seek to trick U.S. companies into funding the North Korean regime’s priorities, including its weapons programs,” Devin DeBacker, supervisory official with the Justice Department’s National Security Division, said in a statement.
Advertisement
Alongside Thursday’s indictments, which come just days after the Treasury Department sanctioned two individuals and four entities for allegedly engaging in similar behavior, the FBI released an advisory warning that North Korean IT workers are increasingly engaging in malicious activity, including data extortion.
The agency said it has observed North Korean IT workers leveraging unlawful access to company networks to “exfiltrate proprietary and sensitive data, facilitate cyber-criminal activities, and conduct revenue-generating activity on behalf of the regime.”
Tata Electronics has acquired a 60% controlling stake in the Indian arm of Apple assembly partner, Pegatron, as the conglomerate expands its iPhone manufacturing capacity in the country.
Taiwan-based Pegatron operates an iPhone production plant near Chennai in India’s sourthern state of Tamil Nadu. The deal comes less than a year after Tata Electronics acquired smartphone assembly company Wistron’s Indian business.
Pegatron and Tata did not disclose the financial terms of the deal, but Tata said the acquisition fits into its strategy of growing its manufacturing footprint in the country.
“We look forward to a new era of AI, digital and technology-led manufacturing as we bring up these new facilities and expand our operations in India,” Randhir Thakur, CEO & MD of Tata Electronics, said in a statement.
Advertisement
Tata, which began assembling iPhones in India just last year, is quickly emerging as one of Apple’s most important partners in Asia as the tech giant works to expand its manufacturing base outside of China.
According to recent data from the International Energy Agency, industrial sectors, such as chemical manufacturing and mining, currently contribute 25% of all global CO2 emissions and 37% of all global energy consumption.
Industries face increasing pressure to develop net-zero roadmaps. As sustainability rises on the corporate agenda, digitalization contributes to a clear, strategic path to achieving commercial, operational, and sustainability goals for today’s industrial enterprises.
Rolf Albrecht
Europe Head of Enterprise Campus Sales at Nokia.
Industry 4.0: Accelerating Sustainability in Industrial Enterprises
Industrial enterprises are increasingly implementing Industry 4.0 to achieve greater operational efficiency and sustainability across their operations. The Internet of Things (IoT) stands out as a significant enabler of sustainability within industrial enterprises. IoT technology combined with seamless connectivity facilitates the continuous flow of real-time data by linking devices, machines, and sensors across the industrial site, which helps reduce fuel, waste, energy consumption and carbon emissions.
For instance, IoT sensors capture real-time energy data from machinery and equipment across entire facilities, allowing enterprises to identify inefficiencies and make precise adjustments to reduce energy waste and lower utility bills. Additionally, IoT technology tracks production outputs and material usage in real-time, helping companies better align supply with demand and swiftly address issues such as overproduction and misallocation of resources. This minimizes the environmental impact of enterprise operations while fostering a more efficient operational model, delivering significant commercial benefits for businesses.
Industrial Edge and Private Wireless Networks Strengthening Sustainability
Industrial enterprises increasingly recognize digitalization’s sustainability benefits, with Industry 4.0 applications powered by private wireless networks driving significant operational improvements and footprint reductions. In fact, according to the 2024 Industrial Digitalization report, by deploying private wireless networks, 79% of enterprises surveyed by GlobalData, reported a reduction in carbon emissions.
Advertisement
Industry 4.0 use cases require the kind of high bandwidth, low latency connectivity provided by 4.9G/LTE and 5G private wireless networks to ensure uninterrupted operations to realize the sustainability and efficiency benefits that come with Industry 4.0.
For example, with reliable worker communication, workers have instant access to the right expertise to repair malfunctioning machinery – this not only reduces truck rolls and fuel use, but also ensures quick repairs to lower unnecessary energy consumption by faulty equipment.
Drones can be used for remote inspections of dangerous or hard-to-reach locations, also reducing vehicle-based inspections and identifying potential issues early before major breakdowns and expensive repairs are required, which often consume more resources and produce more waste. Remote-controlled machines, such as port cranes or robots in chemical plants, can significantly improve the efficiency and safety of operations, while reducing energy consumption and associated carbon emissions.
As global ESG commitments become more stringent, industrial enterprises must invest in the right tools to build towards their sustainability goals. The Private Wireless Sustainability Calculator provides enterprises with a clear digitalization path to reach their goals. Combining handprint and footprint analysis, the calculator helps enterprises estimate how new Industry 4.0 use cases, powered by high-performance private wireless and industrial edge, are helping to reduce greenhouse gas emissions as well as worker safety incidents.
Advertisement
Currently available for mining, ports and chemical manufacturing, enterprises can use insights from the calculator to make informed decisions and guide operational planning and investment choices. The calculator shows that digitalization and use cases like critical worker communication and environmental monitoring are critical tools to support their sustainability goals while also driving operational efficiencies and profitability.
A Data-Driven Approach to Sustainability
Reliable data lies at the heart of a successful sustainability strategy – without it, enterprises are operating in the dark, using inaccurate views of their footprint and the impact operational changes have on their ESG targets, skewing reporting, decision-making and long-term planning. IBM reports that 41% of executives name inadequate data as a top barrier to ESG progress. Yet, despite this, KPMG found that almost 50% of companies still manage ESG data manually in spreadsheets, making reporting extremely time-consuming and prone to human error.
Digitalization allows enterprises to collect comprehensive real-time operational technology (OT) data from all systems, machines and sensors. With an on-premises industrial edge, that data gets aggregated into a single data hub, eliminating data siloes to enable a single unified view on sustainability performance across the operation. Access to all operational data in real-time enables teams to align their efforts with sustainability goals and support ESG reporting, leading to more efficient resource management, optimized production schedules, and minimizing waste. Having the right digital foundation enables industrial enterprises to make immediate progress while setting the stage for future sustainability and increased profitability.
In asset-intensive environments such as manufacturing or port operations, all processes are interlinked and dependent on one another. Any interruption or issue in this chain will result in interruptions and losses for the entire process. This means that there is a need for continuous operations, underpinned by redundancies and mission-critical performances with reliability of 99.99% or more across connectivity, and operational technology (OT).
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
Epic Games is adding third-party games to its mobile app store and helping developers cover associated iOS fees. The company is launching a free games program and will cover the cost for all titles from developers that participate in the program on iOS or iPadOS devices.
With this program, Epic will roll out new free games in its store each month. The games will be available worldwide on Android and in the European Union on iOS.
The company’s mobile store launched in August with its own titles, including Fortnite and Rocket League. With this expansion, Epic Games is adding the first batch of mobile games from outside developers.
Epic CEO and founder Tim Sweeney says the company has invested more than $1 billion in its digital storefront.
That’s it for me today, folks. At the moment, there’s little movement on the availability front; I’ll continue checking retailers to monitor any changes in stock or pre-order procedures so you can be the first to know when it changes.
Be sure to check back in with this guide to see how the situation develops! Think of it as practice for getting your hands on a 5080 come launch day if you like.
Advertisement
Diving into the tech specs, the RTX 5080 has a max temperature rating of 88 degrees Celsius (190.4 degrees Fahrenheit), whereas the 4080 maxes out two degrees higher at 90 Celsius (194 Fahrenheit).
This means that ideally, despite churning out (presumably) a significant amount more performance, the 5080 will run a bit cooler than the RTX 4080. However, if we find the 5080 runs particularly hot under load (stay tuned for our review next week), that lower max temperature could be an issue for graphically demanding titles on builds that have poor ventilation.
Advertisement
Speaking of the monster power draw I mentioned earlier, 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.
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.
Advertisement
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.
All that power means the 5000 series isn’t just for gamers, either.
Advertisement
If you’re a creative professional who relies on a quality GPU to render large amounts of 3D art for editing, or you’re a programmer who works with large-scale ML models, you’ll see some significant improvements over other cards (the 5090 scored “between 35% and 49.3% faster in [our] Blender Benchmark 4.30 tests” for 3D rendering over the 4090, as well as “up to 34.3% faster in Geekbench 6 OpenGL compute test and 53.9% faster in Vulcan”).
Interestingly enough, video editing only saw a small bump in performance, though our Components Editor believes CPU bottlenecking may be at fault here.
With great power comes great…power draw. We just put our much-anticipated Nvidia RTX 5090 review live earlier today, and one of the main ideas I took from it was just how much of a high-end card it really is.
Advertisement
The 5090 isn’t just for the average gamer looking to squeeze out a few extra frames per second. This card regularly pulls down 550W or more of power, and is capable of throwing 200+ fps at 8K (yes, 8K) resolution on Ultra settings in games like Cyberpunk 2077, Hogwarts Legacy, and Star Wars: Outlaws using Multi-Frame Generation.
If you’re thinking about upgrading to a 5090, it’s important to also keep in mind the hidden costs that come with that upgrade; do you have a sufficient power supply? Are your motherboard and RAM up to snuff, and will your current CPU be a massive bottleneck?
That’s why I have my eyes set on the RTX 5080 (or hopefully a 5080 Ti down the road) as a realistic upgrade choice; we can’t talk about the specific performance of this card just yet (stay tuned for that review next week), but its more palatable price tag and the fact that it can also take advantage of Multi-Frame Generation makes me think it could be the best value play of the Blackwell lineup.
Advertisement
(Image credit: Nvidia)
You can’t talk about Nvidia’s new graphics cards without talking about AI and DLSS, can you?
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.
Advertisement
If you suddenly decide you actually have more cash to throw at your PC build (or perhaps you made a typo while Googling) and you’d rather spring for Nvidia’s upcoming flagship monster GPU the RTX 5090, be sure to go and have a gander at our where to buy the RTX 5090 page. I’ll be keeping both pages updated regularly up til launch day and beyond to help you find the card you want – and if you change your mind and want to opt for a more affordable card, we’ll be doing the same for the mid-range RTX 5070 and RTX 5070 Ti in February too.
According to recent reports and rumors, RTX 5000 stock shortages are likely – but I could’ve told you that without needing to read anything. This happens every single time; why would this generation be any different?
Advertisement
Also as inevitable as death and taxes are the scalpers. Sadly, we’re almost certainly going to see some particularly scummy (sorry, ‘enterprising’) individuals snapping up these cards and trying to resell them at inflated prices. Even if you’ve got the cash, steer clear. These sellers are also known to pull scams on desperate shoppers, and if I’m being honest, we shouldn’t support this sort of behavior anyway. Stay strong, friends!
One question many eager PC gamers will no doubt be asking right now is this: where are the pre-orders? And that’s a good question, since it appears that no retailers are offering any sort of pre-order system whatsoever, and it seems less and less likely that we’ll see any at all as launch day creeps closer.
This is a little unusual, but there’s a potential simple explanation: Nvidia and its retail partners are perfectly aware that the RTX 5090 and 5080 are likely to sell out near-instantly with demand remaining high for a while (as was the case with the RTX 4090), so implementing pre-orders systems might just not be worth the time and effort. If you know they’re going to sell like hot cakes anyway, why bother?
Advertisement
One thing we can already talk about is the new physical design of Nvidia’s next-gen graphics cards. The high-end Founders Edition cards of the RTX 4000 generation rightly received some flak from gamers and reviewers alike for being… a little chunky, to say the least. Notably, only the 4070 FE and 4070 Super FE actually qualified for Nvidia’s own Small Form Factor certification.
That’s not the case this time around: every RTX 5000 card all the way up to the RTX 5090 will be small enough to fit inside a (relatively) compact case, something we’re very excited about. A smaller footprint also means a lighter GPU, since the previous RTX 4080 and RTX 4090 were very heavy and carried the risk of damaging your motherboard if not properly supported inside your PC.
Advertisement
While our review of the RTX 5080 (and its big brother, the 5090) won’t be going live for a few hours – we’d rather not upset Nvidia today of all days – you can check out our RTX 5090 unboxing right now to see what’s included in the box.
Most notably, Nvidia has moved away from the problematic 12VHPWR power connector in favor of a new connector with additional sensor pins and a bundled in-box adapter, which should make fitting these new GPUs inside smaller cases easy – and prevent any danger of cable-melting fiascos.
(Image credit: Future)
It’s that time of year again; Nvidia is taking the gaming world by storm with the release of its new RTX 5000 series GPUs, and I’ve got my eye set on the higher mid-range RTX 5080. I’m currently rocking an RTX 3080 Ti in my gaming rig, and while it’s served me well, the 5000 series looks to be an interesting value proposition – so I’m excited to see whether the late TI card release trend continues.
Advertisement
But for now, we have a release date of Thursday,January 30 for the Blackwell lineup, including the Nvidia RTX 5080 and RTX 5090 (check out our hub on where to find RTX 5090 stock if you’re after peak performance). We’ve already gotten our hands on a Blackwell card, but we can’t let you know what we think until the review embargo lifts this Friday, January 23 (keep an eye out for TechRadar’s in-depth review later today!)
That said, those release dates haven’t stopped loads of retailers from teasing their 5080 stock early. Some storefronts like Newegg have already created landing pages that group all of its 5080 stock in one place – perfect for comparing prices and delivery dates.
Make sure to bookmark this page for more updates – I’ll be scouring retailers for any updates as stock and availability moves, so you can be the first to know.
President Donald Trump isn’t happy with how European Union regulators have treated U.S.-based tech companies, including Google, Meta, and Apple.
Onstage at the World Economic Forum in Davos, Trump put EU regulators on blast. He said that the EU’s cases against these U.S. tech companies are “a form of taxation,” according to Bloomberg reporting.
Trump specifically pointed out a court case that Apple lost last year regarding a €13 billion tax bill in Ireland. In 2024, the EU’s Court of Justice in Luxembourg found that Apple had benefited from illegal tax breaks in Ireland between 1999 and 2014 and was forced to pay back taxes.
Bloomberg predicted that this is just the start of a fiery back-and-forth between the Trump administration and the EU that could span the entirety of his second term.
Advertisement
Google, Meta, and Apple are all currently being investigated for potential violations of the EU’s Digital Markets Act.
I found it interesting to check Steam’s Hardware Survey and see that of all participating players in December 2024, 19.56% game at 1440p on their main monitor.
2K is the sweet spot between visual fidelity and performance strain, so it makes sense that a sizable chunk of Steam players hang out at this resolution. The 5090 absolutely crushes most titles at 1440p and high settings, though it does underperform compared to some of its predecessors in certain situations.
Advertisement
If you haven’t already checked out Newegg’s RTX 5090 stock, do give it a look. It’s running an interesting deal on GPU trade-ins right now, meaning you can clean out your old GPU and upgrade to the 5090 in one go.
Just for fun to see how much I’d get, I plugged my GeForce RTX 3080 Ti Gaming X Trio 12G in and found Newegg would pay me $419 towards a new graphics card. That’s a 20% discount on the 5090’x $1,999 price tag – not bad at all.
Advertisement
(Image credit: CD Projekt RED)
The RTX 5090 is the first graphics card on the market to feature GDDR7 video memory (32GB of it at that), and when combined with the improvements to DLSS, as well as Tensor and Ray Tracing Cores, it’s positioned as the first GPU viable of capably running games at 8K resolution.
Our Managing Editor, Core Tech Matt Hanson took the 5090 for a spin in graphically-demanding titles like Cyberpunk 2077, Hogwarts Legacy, and Star Wars: Outlaws to test exactly this, and the results are pretty amazing. The 5090 was able to handle Cyberpunk 2077 at 8K (7,680 x 4,320) resolution, with graphics set to RT Over Drive (essentially the highest they can go), at 148.89 frames per second.
With our Nvidia RTX 5090 review live and its benches marked, we finally have some numbers to share with the more quantitatively inclined. The graphs below show how the new RTX 5090 stacks up against cards like the RTX 4090 and 4080, as well as the RX 7900 XTX and 7900XT on AMD’s side.
Advertisement
Yes, there’s a boost where DLSS is considered; but there’s also a markable upgrade over the previous generation without upscaling in the frame. As per our 5090 review, “With ray tracing and no upscaling, the difference is even more pronounced with the RTX 5090 getting just over 34% faster average framerates compared to the RTX 4090 (with a more modest 7% faster average minimum/1% fps).”
These are some exciting performance bumps, and I have a feeling the 5000 series will fly off the shelves particularly quickly.
When it comes to availability, I’ve done my best to provide a breakdown of the key retailers to watch out for higher up in this article, but be aware that stock is extremely likely to all but evaporate on launch day. Some sites, including Best Buy, Scan, and Nvidia’s own site are offering a ‘Notify Me’ feature which can alert you via email or text message when stock drops on January 30.
Advertisement
It’s worth bearing in mind, however, that Nvidia’s own site will often redirect shoppers to other retailers, especially if you’re looking to buy a third-party version of the RTX 5090, so it’s a good idea to keep tabs on actual seller sites instead.
The real question for any shoppers out there is whether or not they should spring for the RTX 5090, or its little sibling the RTX 5080 – after all, the 5080 is quite literally half the price of the 5090 at $999 / £979 compared to the 5090’s $1,999 / £1,939 price tag.
Although we haven’t quite finished our testing with the cheaper (imagine me doing air quotes around that particular word) GPU, it’s fair to say that the 5080 won’t offer a pure half of the performance available on the 5090, especially once DLSS 4 and Multi Frame Generation are factored in. Of course, the RTX 5090 will likely be the more future-proofed choice, and it does carry with it certain bragging rights…
Advertisement
If you’re currently rocking an RTX 3000 GPU in your rig and have been contemplating an upgrade once the next-gen cards drop (well, why else would you be here?) – you might want to know that Nvidia is potentially planning some retroactive upgrades to your GPU.
Nvidia VP Bryan Catanzaro recently suggested that it might be possible to bring Frame Generation to RTX 3000 cards, as the new version of Team Green’s frame-gen tech doesn’t rely on the Optical Flow hardware accelerator that enabled the tool in the RTX 4000 generation. Instead, it uses an AI-based solution, something that RTX 3000 cards – with their AI-capable Tensor Cores – could potentially utilize. In order words, that cutting-edge technology might soon be available for users with older GPUs, potentially nixing the need for an immediate upgrade.
Advertisement
Some of you might be sitting there wondering about DLSS 4 and its fancy new Multi Frame Generation tech (the latter of which will be exclusive to RTX 5000 GPUs). The viability of DLSS and other upscaling tools of its ilk has been hotly contested by some sectors of the PC gaming community, some of whom claim that it’s become a crutch – an excuse for Nvidia to dial back generational hardware improvements and for game developers to cheap out on PC optimization.
But if recently released usage data is accurate (and there’s frankly no reason to believe it’s not), it looks like DLSS is here to stay. Thankfully, the new DLSS 4 will be backward compatible with all RTX GPUs back to the 2000 generation – unlike DLSS 3, which was locked to RTX 4000 cards exclusively.
(Image credit: Future)
Our Nvidia GeForce RTX 5090 review is now live! Our components editor, John Loeffler, gave it 4.5 stars, calling it “the supercar of graphics cards”. He did knock it a bit for its “obscene” power consumption, which exceeded 550W in his testing, but praised its redesigned cooling and slimmer form factor.
Advertisement
Overall, it’s a major step up from the RTX 4090 in terms of performance, even without factoring in DLSS 4, so once that upscaling tech rolls out on launch day, you can expect even better performance.
The new, sleeker RTX 5000 design reportedly almost didn’t happen: earlier this week, we spotted a mysterious possible RTX 5090 prototype that was a seriously beefy boy, packing specs beyond the real 5090 and a truly absurd 800W power requirement.
While that prototype remains shrouded in uncertainty, it’s possible that it might rear its head further down the line if Nvidia chooses to resurrect its long-dormant Titan RTX series for professional users.
Advertisement
If you’ve somehow stumbled onto the wrong page and were actually looking for Nvidia’s more affordable (but still definitively high-end) new GPU, you can check out our where to buy the RTX 5080 guide instead. I’ll be making sure both pages stay updated regularly up to launch day and beyond to help you track down stock, and it’s worth bearing in mind that the RTX 5080 might be a little easy to get hold of than the flagship 5090 – so if you’re desperate for shiny new graphics card, that could be a better bet.
Fun fact: the RTX 5090 is going to be quite a bit smaller than its predecessor the RTX 4090, despite ostensibly being a more powerful card. Yes, I have to say ‘ostensibly’ because we’re not past the review embargo yet, but come on, we all know it’s going to perform better.
Advertisement
I’m personally over the moon that Nvidia has opted to slim things down for this new high-end GPU, because quite frankly the RTX 4090 was a comically oversized beast of a card regardless of which model you bought. All of the Founders Edition models of every upcoming RTX 5000 card will be certified for Nvidia’s own Small Form Factor Ready scheme, meaning lovers of compact PCs and living-room builds can rejoice.
There have been rumors about RTX 5000 stock shortages circling for the past few days, but even if they prove to be untrue, I strongly suspect that we’re going to see an absolute sell-out almost immediately on launch day. With no pre-orders in sight to secure you a unit in advance, your best bet is likely going to be camping on multiple retailer sites (which I’ve handily organized for you above).
The scalpers are likely to be out in force again for this launch, even now that crypto-mining isn’t as widespread and prevalent as it was during previous GPU releases. The RTX 4090 never really got over its stock issues, with units still selling above MSRP on sites like Amazon. However, that was arguably down to how much of a letdown the RTX 4080 was – hopefully this time around, Nvidia will get things right with its new not-quite-flagship GPU.
Advertisement
Although you’ll still have to wait a few hours until we’re allowed to publish our review – we really don’t want to hear from Nvidia’s legal team today! – you can check out our RTX 5090 unboxing right now to get an early sneak peek at the contents of the package.
A new type of power adapter is now included in the box, making it easier to install the GPU inside smaller PC cases and hopefully putting to bed any previous thorny issues with melting power connectors…
Advertisement
(Image credit: Nvidia / Future)
Here we go again, folks… Nvidia is about to drop its new next-gen flagship graphics card, the RTX 5090 (as well as its little brother the RTX 5080), and you can bet that hordes of gamers – and scalpers – will be chomping at the bit to get their hands on one.
As someone who’s currently got an RTX 4080 sitting inside my home PC, I’m feeling rather comfortable this time around – not beset with the urgent need to upgrade that I felt during the last big Nvidia launch. But that doesn’t mean I’m not here to help you out: with the January 30 release date closing in and the review embargo lifting today, January 23, I’ll be here for the next few weeks to keep tabs on stock and (hopefully!) point you in the right direction.
You’ll be able to check out our review later today, but right here I’ve compiled all the retailers you’ll want to keep an eye on when it comes to tracking down one of these highly-coveted GPUs. Some storefronts, such as Newegg, already have dedicated landing pages for the RTX 5090, but nowhere seems to be offering any sort of pre-orders yet – it’s possible they won’t at all. However, some retailers – including Nvidia itself – are allowing shoppers to sign up for email notifications.
Be sure to bookmark this page and check back for more updates – I’ll be keeping a close watch on retailers for any updates as stock shifts, so you can be the first to know.
Seattle-based Madrona Capital is celebrating its 30 years in business by raising $770 million in fresh capital. This is the firm’s largest fundraise to date, exceeding $690 million across two funds Madrona closed in 2022.
While an 11% capital pool upsize may not seem significant, any increase at a time when many venture outfits are forced to reduce their fund hauls is a sign that limited partners are excited about the firm’s prospects and recent track record.
Madrona’s managing director, Matt McIlwain, told TechCrunch that it helped that last year — in a market where exits were few and far between — the firm sold a few portfolio companies and distributed capital to its investors. The firm’s recent exits include Lexion, which sold to Docusign for $165 million, and Octo AI, which Nvidia acquired for a reported $250 million.
“The LP community is generally concerned about distributions,” McIlwain said. “I think we stood out as a firm that had done really well on that front, not just this past year, but over many years.”
Advertisement
Madrona started as a group of “super angels” who wrote a check to an online bookseller, Amazon, in 1995. The firm has since evolved into a multi-stage investor that has backed companies like Redfin, Smartsheet, Snowflake and, more recently, AI startups Typeface and Runway.
Although Madrona undoubtedly benefited from being the largest VC firm in the same geographic location as Amazon and Microsoft, it decided to venture beyond Seattle by opening an office in Silicon Valley in 2022.
McIlwain said that the fresh capital will be used to invest in AI applications in domains ranging from travel to life sciences, as well as in infrastructure companies that “can remove friction” between foundational models and users. The firm will back about 30 pre-seed, seed and Series A startups from its approximately $490 million early-stage fund, and the remaining capital will go towards 12 companies raising their Series B or Series C.
As Madrona enters its fourth decade, it is extremely optimistic about what’s ahead in 2025. McIlwain described the current conditions as a ‘risk-on mindset’ that will help foster entrepreneurship and create value.
A security researcher found a hidden unreleased feature in the Waymo app that allowed her to display whatever characters she wanted on the robotaxi’s top display.
Jane Manchun Wong, a well-known security researcher, posted an image on X on Saturday showing the top display of a Waymo car — officially called “dome” — that included her X handle and other strings of characters.
“I hacked my Waymo into showing weird texts like empty string, ‘wongmjane,’ and emojis as the Car ID, pls don’t ban me or patch it @waymo lol,” she wrote.
Wong, who lives in San Francisco, told TechCrunch that she was able to customize the characters on the self-driving Jaguar I-Pace dome by fiddling with the Waymo mobile app on her Android phone as she waited for the robotaxi to show up.
Advertisement
“The good old magic of messing around with the Waymo mobile app. I guess their servers didn’t validate the input for the Car ID from non-employees,” said Wong. “So no ‘jailbreaking’ or ‘rooting’ the car itself. All I did was change the Car ID to something beyond what it’d normally accept. A pretty harmless thing I suppose.”
Despite her pleas, it appears Waymo updated the app to stop riders from customizing the dome like Wong did. On Tuesday, Wong posted an update saying she was not able to change the Car ID anymore.
Waymo spokesperson Sandy Karp confirmed that Wong found a hidden feature, and the company shut it down for regular users like Wong.
“Jane identified an unreleased feature given her advanced Android knowledge,” Karp told TechCrunch. “We have restricted access to the dome display features.”
Advertisement
In 2020, Waymo announced that it added moving LEDs to its dome “so that it can act as a mechanism for riders to identify the vehicle day and night and know which vehicle is theirs when there may be more than one Waymo car waiting.”
The dome is also used to communicate with pedestrians that the car is yielding to them, or to cyclists that the passenger is about to open the car’s door. And the company has also used the display for marketing purposes.
You must be logged in to post a comment Login