Latest news with #ComputerBase
Digital Trends
20-07-2025
- Digital Trends
GPUs with 8GB VRAM are 2025's real budget trap
It has become increasingly hard to justify buying a graphics card with only 8GB of VRAM. Once the standard for midrange GPUs, it is now a hard bottleneck in modern games and the cracks are showing more than ever. Whether you're planning to upgrade or building a new budget PC, it's time to reconsider what you actually get with that 8GB label. Let's be clear, this isn't about raw performance alone. As we pointed out during the launch of the RTX 5060, the problem lies less in core count or clock speeds, and more in how 8GB of VRAM caps your long-term usability. Even if the card is decent at 1080p right now, it's the kind of hardware decision you'll regret two years from now. Recommended Videos Take Indiana Jones and the Great Circle, a 2024 release that made headlines for its high system requirements. As per testing done by Computer Base, the game chews through VRAM at 1080p using ultra textures, forcing cards like the RTX 4060 and 4060 Ti 8GB to drop frames or crash entirely. That's not just limited to a particular title, it's becoming the new norm. Games like The Last of Us Part I, Hogwarts Legacy, and Alan Wake 2 are similarly harsh on VRAM, especially at higher settings. And no, simply lowering textures doesn't always 'fix' the problem as it can still result in texture pop-in, long asset loading, and a generally compromised experience. VRAM is the real limiting factor The conversation around GPU performance often fixates on frames per second, but in 2025, VRAM capacity is increasingly the thing separating playable experiences from broken ones. When modern game engines request more than your card can handle, it results in poor performance including stutters and crashes. It doesn't matter if your 8GB GPU technically has enough shader power, it becomes a glorified bottleneck. Worse still, newer cards with 8GB are often misleadingly marketed. Nvidia's RTX 5060 and 5060 Ti (8GB variants) look appealing on paper, with Blackwell efficiency and support for DLSS 4 with multi-frame generation. But when they choke on big textures or fail to keep up in open-world games, the real-world experience often falls short. AMD hasn't been much better. Recent cards like the RX 9070 XT and 9060 XT do bump up to 16GB, but the 8GB RX 9060 XT and even older options like the RX 7600, 6600 XT, and 6650 XT still populate store shelves with 8GB VRAM and tempting discounts. False economy in 2025 It's easy to think you're saving money by buying an 8GB card, but that short-term gain quickly erodes. As games become more demanding and AI workloads become more memory-intensive, you'll end up running into performance walls sooner than you'd expect. That leads to either compromises like dropping settings or resolution, or spending more to replace your GPU sooner than planned. If you're buying a card in 2025, aim for at least 12GB of VRAM, preferably 16GB if you want the system to stay relevant for 3–4 years. Cards like the RTX 5060 Ti (16GB), RTX 5070 (12GB), RX 9060 XT (16GB), and even some budget 7700 XT (12GB) models offer a far better long-term experience, even if they're a bit more expensive. Exceptions to the rule? Despite the growing irrelevance of 8GB GPUs in modern AAA gaming, there are still select scenarios where these cards make sense, provided the buyer understands their limitations. Esports titles like Valorant, League of Legends, and Counter-Strike 2 remain light on VRAM requirements and are designed to run at high frame rates even on modest hardware. For gamers who stick to 1080p resolution and play older or well-optimized games, 8GB cards can still deliver decent results. Similarly, budget-constrained builders, those with less than $300 to spend on a GPU, may find that 8GB cards are their only option unless they opt for second-hand GPUs with higher VRAM but lower efficiency and weaker features. There are also workloads where VRAM isn't the primary bottleneck. Media-centric systems, such as HTPCs or dedicated streaming rigs, can benefit from the video encode/decode capabilities of modern 8GB cards, especially if AV1 support or low power draw is a priority. Small form factor (SFF) builds or compact office PCs often can't accommodate large or high-wattage GPUs, and in such contexts, a compact 8GB card may be the most practical choice. Some users also rely on cloud services like GeForce Now or Adobe's AI-based rendering tools, where the heavy lifting is offloaded to remote servers. In these hybrid workflows, the local GPU serves more as a bridge than a workhorse, making an 8GB card a tolerable, if not an ideal solution. These use cases won't apply to everyone, but they do highlight that there's still a narrow but valid market where 8GB GPUs haven't been completely left behind. Final thoughts The writing is on the wall: 8GB GPUs are no longer a smart buy in 2025. Between games that already exceed that memory envelope and hardware cycles moving quickly toward more demanding workloads, buying an 8GB card today is like buying a smartphone with 64GB of storage. It technically works, but you'll regret it the moment you use it in the real world. If you want to build a PC that lasts, gives consistent performance, and doesn't force you to dial back settings in every new game, skip the 8GB options. It's no longer enough.
Yahoo
08-07-2025
- Yahoo
No PC gamer actually needs this 3000 W power supply unit but like Gollum and the one ring, I wants it
When you buy through links on our articles, Future and its syndication partners may earn a commission. It's fair to say that, compared to CPUs and graphics cards, power supply units aren't really PC components that normally get one's hardware juices flowing. They're mightily important, of course, and you should always get the best you can afford, but once they're hidden in the depths of your gaming rig, you'll never pay much attention to them. That said, the new Asus Pro WS 3000W Platinum is so wonderfully over-the-top that I'm drawn to it like a proto-hobbit to a band of gold. I caught wind of the mighty PSU over at ComputerBase and immediately headed over to Asus' site to pore over its specs. Let's begin with the most important aspect here: the Pro WS range isn't aimed at gamers, as the construction and feature set are firmly in the workstation camp. That said, they will fit any decent-sized desktop case, and if you're feeling flush and foolish, you could always pop one in. The range-topper is the Pro WS 3000W Platinum, and as the name clearly states, this is a 3,000-watt PSU. Yep, three thousand watts. It's that high because it's designed to power up to four GeForce RTX 5090 cards. Or just one and never even notice it. As well as supporting a ludicrous number of GPUs, and the obligatory CPU and motherboard requirements, you've also got scope to host a small mountain of other devices via four 8-pin PCIe and six SATA/Molex connectors. You'd probably need all those just to power the sheer volume of fans required to keep four RTX 5090s from melting the motherboard, though. For me, though, I just really like how industrial it looks. There's no mere flimsy metal grill covering the cooling fan here; the PSU's casing itself acts as a barrier to inquisitive fingers, and it just shouts heft and solidness. Now, I should imagine that you'll be thinking that the Pro WS 3000W Platinum is rather expensive. And you'd be right if you are, as ComputerBase says the range-topper has an estimated MSRP of €699 (around $822/£602 with current exchange rates). It's not due to appear until the last few months of 2025, so the price may alter somewhat by then. But while 699 euros is a lot for a PSU, it's actually not that bad considering its capabilities. I'd say that it's better value than the other Pro WS models, but none of them are what I'd call excessive: the 2,200 W one is €599 and the 1,600 W version is a surprising €479. If you head over to Amazon and check out other 1,600 W PSUs and you'll see that they're all just as pricey, if not more. That's a surprise because you'd expect a 'workstation' product to sport a 'workstation' tax, but if Asus does stick to these prices, then fair play to them for not going wild with the price tags. I suppose the specs are ludicrous enough for anyone. One thing that's worth noting about all PSUs is that they typically hit their peak efficiency at around 50% of the maximum rated load. The exact point varies from model to model, but I reckon that this one will run best in the 1,500 W region. That's still more power than most RTX 5090 gaming rigs will draw, but if you have an extreme overclocked one, then this supply could be the perfect match. Alas, I'll never be able to afford one and my house has rather iffy electrics—I'd hate to plug a fully-loaded Pro WS 3000W Platinum into any of my sockets—but one can dream, yes? Famous fictional characters may be driven to a life of desperate yearning by cursed rings of power, but it's chonkmeister power supplies for me.
Yahoo
27-04-2025
- Business
- Yahoo
TSMC discloses N2 defect density — lower than N3 at the same stage of development
When you buy through links on our articles, Future and its syndication partners may earn a commission. TSMC exposed the defect density (D0) of its N2 process technology relative to its predecessors at the same stage of development at its North American Technology Symposium this week. According to the company, the defect density is below that of N3, N5, and N7 manufacturing nodes. In addition, the slide published by ComputerBase reveals that N2 is two quarters away from mass production, which means that TSMC is on-track to start making 2nm-class chips in late Q4 2025, as expected. Although TSMC's N2 is the company's first process technology to adopt gate-all-around (GAA) nanosheet transistors, the node has lower defect density than its predecessors at the same stage of development, two quarters before mass production (MP). The predecessors — N3/N3P, N5/N4, and N7/N6 — all relied on well-known FinFET transistors. So, despite being TSMC's first node using GAA nanosheet transistors, the N2 defect density is getting lower quicker (well, steeper) than that of its predecessors before the high volume manufacturing (HVM) milestone. The chart plots defect density against time, spanning from three quarters before mass production through six quarters after MP. Across all nodes shown — N7/N6 (green), N5/N4 (violet), N3/N3P (red), and N2 (blue) — defect densities drop significantly as production ramps, though at different rates depending on node complexity. Notably, N5/N4 displayed the most aggressive early defect reduction, while N7/N6 showed a more gradual yield improvement. The N2 curve begins with higher initial defect levels than N5/N4 but declines sharply, closely matching the defect reduction trajectory of N3/N3P. The slide emphasizes that production volume and product diversity remain the key drivers for accelerating defect density improvements. Larger production volumes and a wide variety of products using the same process enable faster identification and correction of defect density and yield issues, enabling TSMC to optimize defect learning cycles. TSMC stated that its N2 fabrication technology got more new tape outs than predecessors (as TSMC now risks producing N2 chips for smartphone and HPC customers), so the defect density decrease curve mostly proves that. The fact that N2's defect reduction rate aligns well with previous FinFET-based nodes is particularly significant, given the risk factors associated with introducing an all-new transistor architecture. It suggests that TSMC has successfully transferred its process learning and defect management expertise into the new GAAFET era without major setbacks (at least based on what TSMC discloses). Sign in to access your portfolio
Yahoo
12-03-2025
- Business
- Yahoo
AMD takes lead over Nvidia, but how long will it last?
While both AMD and Nvidia make some of the best graphics cards, pitting the two against each other usually reveals that Nvidia dominates the GPU market with an over 80% share. However, a new survey revealed that, at least in the recent weeks, many gamers preferred to go with AMD when buying a GPU. But how long will this surprising lead even last? This surprising reveal comes from a ComputerBase survey that was later shared by 3DCenter. According to the survey, a whopping 71% of ComputerBase readers bought an AMD RX 9000 series GPU as opposed to buying Nvidia's RTX 50-series. The rest of the participants opted to buy an older GPU instead, taking advantage of falling GPU prices. This is a dramatic flip on the usual state of things, but there are a few reasons why we're seeing AMD dominate the GPU market all of a sudden. First of all, the survey doesn't take into account the RTX 5070, which has recently launched and would likely tip the numbers more in Nvidia's favor (although I can't see it beating the RX 9070 XT in this particular survey). However, due to limited stock, the GPU may not have made as much of a splash as it would've otherwise. That brings me to my second (and main) point: AMD and its partners succeeded in delivering the RX 9070 XT and the RX 9070 to retailers in time for launch. Nvidia, unfortunately, did not. Availability, or lack thereof, likely affected the numbers in a major way here. There's no beating around the bush: Many people default to Nvidia regardless of the type of product that AMD puts out. However, when Nvidia's GPUs launch with nothing available to buy, gamers turn elsewhere, as can be seen in the ComputerBase survey. It's also worth noting that some enthusiast websites favor Team Red over Team Green, so take these results with a grain of salt. AMD also managed to surprise everyone in a good way by not pricing the GPUs too high, making them an affordable alternative to Nvidia equivalents. However, some cards still cost up to $250 more than the MSRP, and some retailers claim that the prices will rise. Meanwhile, Nvidia is slowly filling in the stock of its graphics cards. Many are still hard to come by, and even if they're in stock, they cost more than they should. But, as things begin to even out, we might see the scales once again tipping in Nvidia's direction — as they always do.
