Tag:

Nvidia

Nvidia DGX
Product Reviews

Nvidia to boost AI server racks to megawatt scale, increasing power delivery by five times or more

by admin



Nvidia is developing a new power infrastructure called the 800V HVDC architecture to deliver the power requirements of 1 MW server racks and more, with plans to deploy it by 2027. According to Nvidia, the current 54V DC power distribution system is already reaching its limit as racks begin to exceed 200 kilowatts. As AI chips become more powerful and demand more electricity, these existing systems would no longer be able to practically keep up, requiring data centers to build new solutions so that their electrical circuits do not get overwhelmed.

For example, Nvidia says that its GB200 NVL72 or GB300 NVL72 needs around eight power shelves. If it used 54V DC power distribution, the power shelves would consume 64 U of rack space, which is more than what the average server rack can accommodate. Aside from this, it also said that delivering 1 MW using 54V DC requires a 200 kg copper busbar — that means a gigawatt AI data center, which many companies are now racing to build, would need 500,000 tons of copper. This is nearly half of the U.S.’s total copper output in 2024, and that’s just for one site.

So, instead of using the 54V DC system, which is installed directly at the server cabinet, Nvidia is proposing to use the 800V HVDC, which will connect near the site’s 13.8kV AC power source. Aside from freeing up space in the server racks, this will also streamline the approach and make power transmission within the data center more efficient. It will also remove the multiple AC to DC and DC to DC transformations used in the current system, which added complexity.


You may like

The 800V HVDC will also reduce the system current for the same power load, potentially increasing the total wattage delivered by up to 85% without the need to upgrade the conductor. “With lower current, thinner conductors can handle the same load, reducing copper requirements by 45%,” said the company. “Additionally, DC systems eliminate AC-specific inefficiencies, such as skin effect and reactive power losses, further improving efficiency.”

According to Digitimes [machine translated], the AI giant is working with Infineon, Texas Instruments, and Navitas to help develop this system. Furthermore, it’s expected that they will deploy wide-bandgap semiconductors like gallium nitride (GaN) and silicon carbide (SiC) to achieve the high power densities needed by these powerful AI systems. The 800V HVDC is a technical challenge that data centers must solve for power efficiency, especially as they start to breach 1 GW capacity and more. This solution should help them reduce wasted power, which, in turn, would reduce operating costs.

Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.



Source link

0 comments
0 FacebookTwitterPinterestEmail
A modded GTX 970 graphics card with 8GB VRAM
Gaming Gear

Nvidia GTX 970 memory mod boosts performance with 8GB VRAM upgrade

by admin



The Nvidia GeForce GTX 970 was a promising graphics card at launch offering performance similar to AMD’s Radeon R9 290X at a more affordable price range. Ten years later, the GPU has been given a new life thanks to Brazilian modder Paulo Gomes and his team who managed to add more VRAM, doubling the original from 4GB to 8GB.

To increase the memory, the mod involved replacing the original 512MB, 7 Gbps GDDR5 memory modules with 1GB, 8 Gbps chips along with a resistor so that the GPU can recognize the higher density VRAM. By having access to 8GB of upgraded VRAM, the modder noted an improvement in Unigine Superposition benchmark showing nearly double the score.

This one-of-a-kind Asus Strix-branded RX 970 was then handed to Youtuber Peperaio Hardware to evaluate whether increasing VRAM on an older GPU translates to any improvement in gaming. Testing was done using a mix of old and new games where titles like Red Dead Redemption 2, GTA V Enhanced, and Plague Tale Requiem didn’t show any signs of improvement. The results were compared to a Gigabyte Windforce GTX 970 with 4GB of memory.


You may like

In Cyberpunk 2077, the results varied depending on the graphics settings with an uplift of 5-15% in performance on the 8GB modded GPU. The two games that gained noticeable benefit from the increased VRAM were The Last of Us Part II Remastered with a 24% increase in frame rate and Horizon Forbidden West offering as much as 40% improvement.

Back in February last year, the same team of modders managed to repair and upgrade a defective RTX 3070 by replacing its original 8GB of GDDR6 memory with 12GB. The team soldered in new 2GB memory chips and modified the GPU’s BIOS to correctly detect and utilize the additional VRAM. After the upgrade, the graphics card operated stably and passed all stress tests, showing that the mod was technically viable and functionally solid.

The upgrade had led to a performance gain in games including Resident Evil 4 Remake, where the modified RTX 3070 delivered up to a 66% performance increase compared to the standard 8GB version. In The Last of Us Part I, performance saw a boost of around 25%, while Hogwarts Legacy ran 20% better with the extra VRAM.

Both Nvidia and AMD have continued to claim that 8GB of VRAM is sufficient for most gamers. However, real-world mods like the ones mentioned above, suggest otherwise. Even as modern titles become increasingly demanding, Nvidia went on to launch the RTX 5060 and 5060 Ti in 8GB variants, while AMD has announced that its new Radeon RX 9060 XT will also be coming with 8GB of VRAM.

Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Nvidia CEO Jensen Huang Next to a B200 Node
Product Reviews

Nvidia RTX PRO 6000D (B40) Blackwell GPUs reportedly set to supersede banned H20 accelerators in China

by admin



Following the ban of its Hopper H20 accelerations in China, Nvidia is reportedly planning on launching new Blackwell-based solutions at a lower price this year, per Reuters. With mass production anticipated by June, we can expect these solutions to be widely available in the Chinese market by Q3 or Q4. While technical details are still emerging, we can already discern some important details and specifications.

Due to stringent U.S. export policies targeting China, the Hopper family has largely been a cat-and-mouse chase between U.S. regulators and Nvidia. Even before their official debut, the flagship H100 and H200 accelerators were already subject to export bans. Nvidia introduced the H800 to circumvent these regulations, which eventually faced a similar fate in October 2023. The cut-down H20 served as Nvidia’s primary AI solution for the Chinese market in the interim until its recent ban under the current administration last month, which forced Nvidia to write off $5.5 billion in GPU supply.

Reuters reports that Nvidia’s follow-up to the H20 will be based on the Blackwell architecture, more specifically, the RTX Pro 6000D. Further clarification by tipster Jukanlosreve at X, citing a report from China’s GF Securities, suggests the RTX Pro 6000D will be dubbed B40 (likely a successor to the Ada Lovelace L40). Reuters classifies this as a server-class GPU which uses traditional GDDR7 memory instead of HBM, and notably avoids the use of TSMC’s CoWoS packaging technology, likely signaling at its monolithic nature.


You may like

Silicon possibilities and B40 pricing

There are two possibilities based on the available data. This GPU can either be based on datacenter grade GB1XX Blackwell or consumer-grade GB2XX Blackwell silicon. The former is unlikely, as it only features HBM controllers at the silicon level. If the B40 utilizes GB2XX dies, it would be a derivative of the GB202 chip (found in the RTX 5090 and RTX Pro Blackwell 6000) and would lack NVLink support. The report estimates the price of the B40 between $6,500 and $8,000, which is less than the H20 and comparable to Nvidia’s global RTX Pro 6000 workstation models.

The HGX H20 could be configured in an 8-GPU configuration, but without NVLink, the B40 would likely face challenges in multi-GPU setups. Nvidia’s latest RTX Pro Blackwell servers employ up to eight RTX Pro 6000 GPUs, connected via ConnectX-8 SuperNICs with integrated PCIe 6.0 switches, for GPU-to-GPU communication. This setup is likely what we’ll see for the B40, with scaling beyond eight GPUs expected to be handled by Nvidia’s Spectrum-X networking platform. Since details are scarce, this is just speculation on our part, so please don’t read it as gospel.

Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Nvidia rtx 50 series
Product Reviews

Nvidia releases emergency RTX 5060-series firmware to fix blank screens on reboot

by admin



Many RTX 5060 and RTX 5060 Ti users encounter blank screens when restarting their systems. Nvidia has released an update to resolve this problem, not through a new driver, but as a firmware update for the GPU, also known as the vBIOS. The report strictly advises that only users experiencing this problem should update their firmware.

The problem likely stems from how Nvidia’s vBIOS communicates with the system’s BIOS/UEFI. If your motherboard does not support UEFI boot mode, Nvidia recommends contacting the customer service of your GPU manufacturer for a legacy vBIOS update. The exact culprit hasn’t been specified, but the fact that Nvidia is providing a fix strongly implies that part of the problem is within their domain.

Nvidia’s RTX 50 GPUs have struggled with stability and compatibility, starting with the problematic R570 branch of drivers, which even affected older RTX 20/30/40 series GPUs. Having experienced these issues first-hand, it took Nvidia multiple weeks to address most of them with the 576.02 update. If your RTX 5060 or RTX 5060 Ti throws a black or blank screen on reboot, you may need to update its vBIOS. An abrupt power outage during firmware updates can potentially brick your GPU, so proceed cautiously.


You may like

Fortunately, Nvidia’s provided utility can verify compatibility with your system’s firmware and determine if a vBIOS update is necessary. Nvidia recommends only installing this update if you’re encountering blank screens. To apply the fix, you must first boot into the Operating System, which might be a hassle given the current predicament. Nvidia suggests the following temporary workarounds to boot with a display:

  • Completely power down your system before booting.
  • Ensure you are using the latest SBIOS from your motherboard vendor.
  • Ensure you are in UEFI boot mode and not Legacy/CSM.
  • Boot using an alternate graphics source (secondary card or integrated graphics).
  • After powering on your system, wait for your operating system to load with the graphics driver installed.

Download the NVIDIA GPU UEFI Firmware Update Tool v2.0 to apply the update, run the executable, and follow the on-screen instructions. Interestingly, this issue only impacts the RTX 5060 and RTX 5060 Ti family of GPUs, both based on GB206 silicon. This raises the question of why older RTX 50 cards are unaffected. What specific changes in these GPUs led to this problem?

Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Oracle logo on building
Product Reviews

Oracle has reportedly placed an order for $40 billion in Nvidia AI GPUs for a new OpenAI data center

by admin



Oracle has reportedly purchased about 400,000 Nvidia GB200 AI chips, worth about $40 billion, for deployment in Abilene, Texas. According to the Financial Times, this site will be the first to host the U.S. Stargate project —a $500 billion investment in AI infrastructure by OpenAI, Oracle, SoftBank, and Abu Dhabi sovereign wealth fund MGX that President Trump announced earlier this year.

Upon completion, the site is estimated to deliver up to 1.2 gigawatts of computing power, making it one of the most powerful data centers in the world and competing against Elon Musk’s Colossus in Memphis, Tennessee.

The project’s ownership is a bit complicated, though. The Abilene site is owned by Cruso, an AI infrastructure company, and Blue Owl Capital, a U.S. investment firm. Both have poured over $15 billion into the site through debt and equity financing.


You may like

The AI data center campus is approximately 875 acres and will have eight buildings. Construction began in June last year. Oracle will lease the Abilene site for 15 years, where it will deploy these AI chips. It will then lease this massive computing power to OpenAI, allowing it to use the AI chips to train its next-generation AI LLMs.

OpenAI has solely relied on Microsoft’s data centers for its computer power needs. However, it has since become frustrated with the setup, because the former’s demand was greater than what the latter could supply. So, the two decided to terminate their exclusive contract, allowing the AI pioneer to partner with other organizations for its processing power demands.

Interestingly, all these talks on equity and investments are commitments by the individual companies—the Financial Times said that Stargate itself has yet to commit any amount to a data center. Nevertheless, OpenAI recently also announced Stargate UAE, where it plans to deploy a 1GB cluster in Abu Dhabi in coordination with the U.S. government. The 5 GW data center will be built by G42 and is envisioned to utilize more than 2 million Nvidia GB200 chips.

All these AI data center deployments have helped Nvidia become one of the most valuable companies in the world, with Team Green hitting the number one spot (again) in late 2024. Aside from this $40 billion Oracle order, there has also been news that Elon Musk is raising cash to build Colossus 2, featuring a million GPUs. Other tech giants and even governments are trying to get in on the bandwagon, pushing the demand for the latest, most powerful AI chip through the roof.

Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Nvidia GeForce RTX 5060 review: better than console performance - but not enough VRAM
Game Reviews

Nvidia GeForce RTX 5060 review: better than console performance – but not enough VRAM

by admin


The RTX 5060 is here, finally completing the 50-series lineup that debuted five months ago with the 5090. The new “mainstream” graphics card is far from cheap at $299/£270, but ought to offer reasonable performance and efficiency while adding the multi frame generation feature that’s exclusive to this generation of GPUs. However, the 5060 also ships with just 8GB of VRAM, which could be a big limitation for those looking to play the latest graphics showcases.

Before we get into our results, it’s worth mentioning why this review is a little later than normal, coming a few days after the cards officially went on sale on May 19th. Normally, Nvidia or their partners send a graphics card and the necessary drivers anywhere from a couple of days to a week before the embargo date, which is typically a day before the cards go on sale. That’s good for us, because it allows us to do the in-depth analysis that we prefer and still publish at the same time as other outlets, and it’s good for potential buyers, as they can get a sense of value and performance and therefore make an informed decision about whether to buy a card or not – from what is often a limited supply at launch.

For the RTX 5060 launch, Nvidia – via Asus – delivered a card in good time ahead of its release, but the drivers weren’t released to reviewers until the card went on sale on May 19th, coinciding with Nvidia’s Computex presentation. Without the drivers, the card is a paperweight, so any launch day coverage is necessarily limited – and in many cases, graphics cards went out of stock before the usual tranche of reviews went live from the tech press. It’s a frustrating situation all around, and I doubt that even Nvidia’s PR department will be thrilled that most reviews start with the same complaint.

Nvidia’s GeForce RTX 5060 gets the Digital Foundry video review treatment.Watch on YouTube

Following the public release of the drivers, we’ve been benchmarking around the clock to figure out just how performant the new RTX 5060 is, where its strengths and weaknesses lie, and where it falls compared to the rest of the 50-series line-up, prior generation RTX cards and competing AMD models.

Looking at the specs, you can see that the RTX 5060 is based around a cut-down version of the same GB206 die that powered the RTX 5060 Ti. The 5060 has 83 percent of the core count and rated power of the full-fat 5060 Ti design, with an innocuous three percent drop to boost clocks and the same 448GB/s of memory bandwidth.

Unlike the 5060 Ti, however, which debuted in 8GB and 16GB models, the 5060 is only available with 8GB of frame buffer memory – a limitation we’ll discuss in some depth later. For your 16.6 percent reduction to core count and TGP versus the 5060 Ti, you pay around 20 percent less – so the 5060 ought to be slightly better value.

Marvel’s Spider-Man 2 and Monster Hunter World – 1440p resolution. We aren’t at native resolution. We aren’t on ultra settings, but both 8GB RTX 5060 and RTX 5060 Ti see performance collapse. The 16GB RTX 5060 Ti works fine and delivers good performance – proof positive that 8GB is too much of a limiting factor for these cards. | Image credit: Digital Foundry

RTX 5070 Ti
RTX 5070
RTX 5060 Ti
RTX 5060

Processor
GB203
GB205
GB206
GB206

Cores
8,960
6,144
4,608
3,840

Boost Clock
2.45GHz
2.51GHz
2.57GHz
2.50GHz

Memory
16GB GDDR7
12GB GDDR7
16GB GDDR7
8GB GDDR7
8GB GDDR7

Memory Bus Width
256-bit
192-bit
128-bit
128-bit

Memory Bandwidth
896GB/s
672GB/s
448GB/s
448GB/s

Total Graphics Power
300W
250W
180W
150W

PSU Recommendation
750W
650W
450W
450W

Price
$749/£729
$549/£539
$429/£399
$379/£349
$299

Release Date
February 20th
March 5th
April 16th
May 19th

There’s no RTX 5060 Founders Edition, as you’d perhaps expect for a mainstream model, with various third-party cards available in a range of sizes. The RTX 5060 model we received is the Asus Prime model, an over-engineered 2.5-slot, tri-fan design that is nonetheless described as “SFF-ready” due to its relatively modest 268mm length. On top of the robust industrial design, the card features a dual BIOS with “quiet” and “performance” options – always useful. In this case however, the cooler is so large that even the “performance” option is very, very quiet. The card ships with this preset and we recommend it stays there.

Hilariously, the manufacturer product page recommends a 750W or 850W Asus power supply, though the specs page for the same model makes a more sane 550W recommendation. Regardless, you’ll be good to go with a single eight-pin power connector. In terms of ports, we’re looking at the RTX 50-series standard assortment, including one HDMI 2.1b and three DisplayPort 2.1b.

Like the RTX 5060 Ti – but not AMD’s just-announced Radeon RX 9060 XT – the RTX 5060 uses a PCIe 8x connection. That’s perfectly fine on a modern PCIe 5.0 or 4.0 slot, but potentially problematic on earlier motherboards with PCIe 3.0 slots – something we’ll test out in more detail on page eight.

For our testing, we’ll be pairing the RTX 5060 with a bleeding-edge system based around the fastest gaming CPU – the AMD Ryzen 7 9800X3D. We also have 32GB of Corsair DDR5-6000 CL30 memory, a high-end Asus ROG Crosshair X870E Hero motherboard and a 1000W Corsair PSU.

With all that said, let’s get into the benchmarks.

Nvidia GeForce RTX 5060 Analysis

  • Introduction and test rig [This Page]
  • RT benchmarks: Alan Wake 2, Avatar: Frontiers of Pandora, Cyberpunk 2077
  • RT benchmarks: Dying Light 2, F1 24, Hitman: World of Assassination
  • RT benchmarks: Indiana Jones and the Great Circle, Metro Exodus Enhanced Edition, A Plague Tale: Requiem
  • Game benchmarks: Alan Wake 2, Black Myth: Wukong, Cyberpunk 2077
  • Game benchmarks: F1 24, Forza Horizon 5, Senua’s Saga: Hellblade 2
  • Game benchmarks: Hitman: World of Assassination, A Plague Tale: Requiem
  • PCIe 3.0 vs PCIe 5.0: Black Myth: Wukong, F1 24, Indiana Jones and the Great Circle
  • PlayStation 5 comparisons and DLSS 4 multi frame generation
  • Conclusions, value and recommendations

Manage cookie settings



Source link

0 comments
0 FacebookTwitterPinterestEmail
An Nvidia GeForce RTX 5080 on a desk with its retail packaging
Gaming Gear

Rumored Nvidia RTX 5080 Super specs disappoint some gamers, but I don’t think there’s anything to worry about with this GPU

by admin



  • A leak has detailed the claimed specs of Nvidia’s RTX 5080 Super
  • Some gamers might see this refreshed GPU as underwhelming – it doesn’t add any extra cores into the mix, notably
  • However, there are robust upgrades elsewhere with the video memory and also clock speeds

Another rumor about Nvidia’s RTX 5080 Super has been aired and we’ve got a look at what are supposedly the full specs of this GPU.

As VideoCardz pointed out, leaker Kopite7kimi has posted the claimed specs for the rumored graphics card on X, and that may mean Nvidia has just provided said details to its graphics card making partners (and they leaked from there). Or, it might mean precisely nothing, because as ever, rumors, much like demons, need considerable salting.

GeForce RTX 5080 SuperPG147-SKU35GB203-450-A110752FP32256-bit GDDR7 24G 32Gbps400+WMay 20, 2025

The key parts of the specifications are that the RTX 5080 Super will supposedly use the same GPU as the RTX 5080, which is the GB203 chip. As the RTX 5080 has already maxed out the cores on that chip, the core count will be the same with the Super version of this graphics card – there’s no room to maneuver to increase it.


You may like

The big upgrade comes from the leap from 16GB to 24GB of video RAM (VRAM), and as well as that 50% uplift, the leaker believes Nvidia is going to use faster memory modules here (32Gbps rather than 30Gbps).

We’re also told that the TDP of the RTX 5080 Super is going to sit at 400W, or it might use even more power than that.

Analysis: Crunching the specs and not forgetting about clocks

(Image credit: Future / John Loeffler)

Looking at those specs, you might think: how is the RTX 5080 Super going to be a tempting upgrade on the vanilla version of the GPU? It has the same CUDA core count, and somewhat faster video memory, but only around 7% more VRAM bandwidth than the RTX 5080. So, what gives?

Well, don’t forget that added to that VRAM boost, the RTX 5080 Super is expected to have considerably faster clock speeds. Pushing those clocks faster is why this incoming GPU is going to chug more than 400W (perhaps a fair bit more) compared to 360W for the plain RTX 5080.

So, if you’re worried that the RTX 5080 Super may represent an underwhelming prospect in terms of an upgrade over the RTX 5080, don’t be. (Although you may have concerns about your PC’s power supply instead). All this is in line with previous speculation that we’ll see something like a 10% performance boost with the RTX 5080 Super versus the basic version of the GPU, or maybe even slightly more (up towards 15%, even).

Plus that much bigger allocation of 24GB of VRAM is going to make a difference in some scenarios where 4K gaming coupled with very high graphics settings gets more demanding with certain games. (A situation that’s only going to get worse as time rolls on, if you’re thinking about future-proofing, which should always be something of a consideration).

On top of this is the fact that Nvidia is falling out of favor in the consumer GPU world, with AMD’s RDNA 4 graphics cards making a seriously positive impact on Team Red’s chances – and sales. The latest RX 9060 XT reveal has pretty much gone down a treat, too, so I don’t think Nvidia can risk damaging its standing with PC gamers any further, frankly, by pushing out subpar Super refreshes.

Speaking of refreshes – with the emphasis on the plural – previous rumors have also theorized an RTX 5070 Super graphics card with 18GB of VRAM is on the boil, but that’s notably absent from Kopite7kimi’s post here. That doesn’t mean it isn’t happening, but it could be read as a sign that the RTX 5080 Super is going to arrive first.

Again, previous spinning from the rumor mill indicates a very broad 2025 release timeframe for the RTX 5080 Super, but if the specs really are decided on at this stage – and it’s a huge if – that suggests Nvidia intends to deploy this GPU sooner, rather than later, this year.

You might also like





Source link

0 comments
0 FacebookTwitterPinterestEmail
Nvidia's Jen-Hsun Huang on stage during the GTC 2025 keynote
Gaming Gear

Nvidia’s CEO says attempts to control chip exports to China are a failure: ‘If they don’t have enough Nvidia, they will use their own.’

by admin



Attempts by the US government to put a cap on China’s development of AI technologies by limiting exports of GPUs has been a “failure”. So says no less an authority on the subject than Nvidia CEO, Jensen Huang.

The New York Times quotes Huang at the ongoing Computex show in Taipei, Taiwan denouncing GPU export controls. “AI researchers are still doing AI research in China,” Huang said on Wednesday. “If they don’t have enough Nvidia, they will use their own,” he said. All of which means, “the export control was a failure.”

He may have a point. But then Nvidia does rather have a dog in this fight. Huang himself says that restrictions on Nvidia’s H20 GPU will cost the company $15 billion in sales. So, it’s not hard to understand why he might prefer those limitations to be lifted.


You may like

Just for context, back in 2022 the former Biden administration imposed limits on the export of the most powerful GPUs from the US into China. Into the void left by restricted Nvidia exports has moved local outfit Huawei, whose GPUs currently do not match those of Nvidia for AI prowess. However, the fear is that the GPU export restrictions have only encouraged Huawei to put even more effort into closing the gap.

Indeed, according to the New York Times, Nvidia is concerned about just that, with an adjacent worry that, “any advantage gained by Huawei in China could eventually spread into other markets, helping Huawei build a stronger foundation from which to compete around the world.”

Computex 2025

(Image credit: Jacob Ridley)

Catch up with Computex 2025: We’re on the ground at Taiwan’s biggest tech show to see what Nvidia, AMD, Intel, Asus, Gigabyte, MSI and more have to show.

Meanwhile, it’s a little difficult to gauge Jensen Huang’s strategy and loyalties in all this. He recently appeared with other business leaders as a guest of the Trump administration in Saudi Arabia. But Nvidia has also just unveiled what will be a new Global headquarters in Taiwan, which doesn’t entirely square with the broader push to reshore tech manufacturing to the US.

Likewise, the New York Times reports that, “the day after the US government opened an investigation into whether Nvidia’s previous sales to China had violated its rules, Mr. Huang met with top economic and trade officials in Beijing.”

The plot, as they say, thickens. At the very least, it seems Huang and Nvidia are keeping their options fully open.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Nvidia GeForce RTX 5060 review: passable GPU, shame about the drivers
Game Reviews

Nvidia GeForce RTX 5060 review: passable GPU, shame about the drivers

by admin


Zotac Gaming GeForce RTX 5060 Solo specs:

  • CUDA Cores: 3840
  • Base Clock Speed: 2.28GHz
  • Boost Clock Speed: 2.49GHz
  • VRAM: 8GB GDDR7
  • Power: 145W
  • Recommended System Power: 550W
  • Price: From £270 / $299

I’d so desperately like to do a graphics card review without the fug of a wider controversy (or cacked-up market conditions), but the RTX 50 series hasn’t been particularly cooperative in that regard, so why should the RTX 5060 be any different? This time, the sadness cloud comes wafting from Nvidia themselves, amid accusations of engineering dodgy RTX 5060 previews and attempting to trade access for greater coverage of its Multi Frame Generation (MFG) capability.

Such scheming, if true and intentional, would suggest a remarkable lack of faith in the RTX 5060’s core, un-frame-genned performance. Yet now that I’ve spent some quality time with the card myself – independent of any tit-for-tat preview shenanigans, obviously – it really isn’t that bad, on pure hardware terms. It’s not equipped for cut-price 1440p but as an affordable 1080p pusher, it’s fine. Adequate. Reasonable. Hardly some catastrophe that needs a thunder-running PR offensive to cover up with MFG figures.

The drivers, though? Now there’s a disaster. And not because of the convention-breaking lack of early review software for press hacks – I got access to the RTX 5060’s Game Reader Driver 576.52 update at the same time everyone else did, when both it and the GPU released on May 19th. The real problem is that it’s the latest in a series of GeForce driver updates that have invited all manner of unforced errors upon games old and new, including several of my benchmarking regulars.

Image credit: Rock Paper Shotgun

Some, like Horizon Forbidden West and Dragon Age: The Veilguard, began suffering sustained framerate drops that they’ve never exhibited before. Meanwhile, Metro Exodus would crash on startup, and F1 24 would perform worse with Nvidia’s precious DLSS frame generation than without. Even if these aren’t the fault of the RTX 5060 hardware, these driver problems are just yet more bad vibes around a graphics card that should be – like the RTX 3060 and, after a while, the RTX 4060 before it – a people’s champion.

This is, after all, the more affordable RTX 50 GPU of the bunch, and thus the least taxing entry point (literally, if you’re over the pond in Tariffs Land) into full-spectrum DLSS 4 support. In fact the very model you see here, the nicely compact Zotac GeForce RTX 5060 Solo, is one of several that are actually selling at RRP/MSRP, or £270 / $299. That’s a snip, by 2025 standards, especially when most RTX 5060 Ti models have already gained a few quid. If nothing else, then it’s at least worth looking for some upsides.

Nvidia GeForce RTX 5060 review: 1440p benchmarks

Granted, 1440p maybe isn’t the best place to start the search. After various restarts and reloads, I did eventually get some usable data from the 576.52 drivers, and at native resolution the RTX 5060 does make for visibly smoother framerates than the RTX 4060 in most games – Metro Exodus and Total War: Warhammer III especially. But then it only produced a single extra frame in Forbidden West, forcing it to drop behind the older and cheaper Intel Arc B580. That’s an underdog GPU that the RTX 5060 could also only hold to a draw in both Cyberpunk 2077 and F1 24.

Click to embiggen! | Image credit: Rock Paper Shotgun

The RPS test PC:

  • CPU: AMD Ryzen 7 9800X3D
  • RAM: 32GB Trident Z5 Neo RGB DDR5
  • Motherboard: MSI MPG X870E Carbon WiFi
  • PSU: NZXT C1000 Gold

Points for coming close to the Radeon RX 7700 XT, a potential challenger from the second-hand market, but if I was speccing a 1440p rig on a budget, I’d still save up for an RTX 5060 Ti 16GB. In part because the VRAM difference does become evident at this kind of rez, particularly in Forbidden West, which had a certain jittery quality that I don’t think is entirely explained by the lower frame output alone. Playable evidence suggests 8GB is okay for most 1080p games, despite recent grumbles from tech enthusiasts, but an extra 8GB on top of that likely will help cope with the rigours of Quad HD.

You’ll also need to invest more if you want to partake in path tracing. Even with Quality DLSS upscaling, neither of my path traced test games could reach 30fps on the RTX 5060, again making an argument for the RTX 5060 Ti. MFG could get the numbers up, but only Cyberpunk 2077 at 2x felt remotely playable: at 4x, input lag went off the charts, and Alan Wake II had a similarly sludgy feeling (along with noticeable blurring on camera movements).

Image credit: Rock Paper Shotgun

DLSS 4 is still the best overall upscaler/frame gen package in the biz, but as I seem to say every time it comes up, it just doesn’t work as a means of smoothening out low performance. It’s great at taking quite-fast games and making them look even slicker, but that really needs a foundation of ordinarily rendered frames for DLSS to generate new ones from; without that, it’s the gaming hardware equivalent of sitting in a rusty wheelbarrow with a Ferrari livery. You’ll pick up some good speeds on the right hill, but won’t enjoy the sensation.

Nvidia GeForce RTX 5060 review: 1080p benchmarks

Life is much better at 1080p. The B580 still beats the RTX 5060 in Horizon, but only by a few frames, and it’s practically on par with the RTX 5060 Ti 16GB in Metro and Cyberpunk. Not far behind in Warhammer III or Assassin’s Creed Mirage, either.

Image credit: Rock Paper Shotgun

Vitally, there’s also – more often than not – a decent improvement on the RTX 4060, especially in Metro and Cyberpunk. This is sustained with the application of regular ray tracing, too. Adding Ultra-quality RT effects to Metro only brought the RTX 5060 down to 77fps, while the RTX 4060 managed 62fps.

Will it consistently fill out a 165Hz monitor on max quality? No, but then for less than £300 it doesn’t need to. It’s fine. Adequate. Reasonable, I remember someone saying. VRAM-wise, you should probably think about whether you might like to upgrade to 1440p within this card’s shelf life, but for the time being it does look like you can get away with 8GB at 1080p. I didn’t see much more of that jittering in Forbidden West, for one thing, and a side-jaunt into Doom: The Dark Ages – with its always-on ray tracing and Ultra Nightmare settings – produced a smooth, stutter-free 73fps, once again besting the RTX 4060 at 60fps.

Image credit: Rock Paper Shotgun

Path tracing remains a questionable endeavour, mind. Although the RTX 5060 could crawl to an ostensibly playable 30fps-plus in both games at this lower rez, this wasn’t enough to avoid an offputting deluge of input lag once MFG tried to make up the difference. Alan Wake 2 wasn’t as blurry as at 1440p, but still, it felt sharper to just have Ultra-quality ray tracing at 44fps instead.

In fairness, I did find a use case for MFG in Dragon Age: The Veilguard. With Ultra settings, DLSS on Quality and all ray tracing effects enabled, Nvidia’s tech turned 53fps into 88fps on 2x and 150fps with a 4x override. Crucially, neither of the heightened results came with excessive latency, thanks largely to the fact that the RTX 5060 was already running the game acceptably without them.

Image credit: Rock Paper Shotgun

It’s just unfortunate that to get these numbers, I had to re-run the test every time one of those newfound framerate collapses took place, on top of having to restart the game after every settings change because otherwise they’d kill performance for no apparent reason. That’s not because of a recent bad patch on Bioware’s part, and it’s certainly not a problem with how the RTX 5060 itself is engineered, with Zotac’s single fan keeping peak GPU temps to a sensible 68°c. Nope, this was the fault of my lifelong enemy for the past two days: those 576.52 drivers.

A possibility exists that they’ll be fixed, and might not even be replaced by something worse, but at this point, there have been enough faulty Game Ready drivers – whose faults are usually specific to the RTX 50 series – that it’s become a problem for the entire GPU family. Sadly, that has to include the RTX 5060. By Nvidia’s own hand, this puts it in the unenviable position of being the most powerful and flexible 1080p graphics card in its price range yet also one that makes the words “Yes, you should buy this” disproportionately difficult to say out loud. Why would, or should, someone invest in a component when its requisite drivers have such a high chance of breaking their games?

This review is based on a retail unit provided by the manufacturer.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Radeon Rx 9060 Xt From Amd
Gaming Gear

AMD’s Radeon RX 9060 XT Could Do Budget GPUs Better Than Nvidia

by admin


In the battle of the low-end, 60-class graphics cards, AMD wants to see if it can pull off the same sucker punch of price and performance it gave Nvidia during the launch of its mid-range GPUs. The graphics card maker offered the first, sparse details on its Radeon RX 9060 XT graphics processors late Tuesday at Computex. The card may offer enough power for your PC to hit solid gaming performance at 1440p resolution, similar to the $450 Nvidia GeForce RTX 5060 Ti, on cheaper gaming rigs. The real inflection point of this latest card will be whether you can actually buy it for its base price.

The Radeon RX 9060 XT is the step down in GPU performance from the RX 9070 that AMD launched back in March. It’s based on the same RDNA 4 microarchitecture of the mid-range cards, but with 32 of the company’s latest compute units compared to the 56 on the higher-end card. The GPU comes with two options: one with 8 GB and another with 16 GB of GDDR6 VRAM. The version with more memory will be better for your rig long-term, especially if you plan to hook your PC up to a 1440p monitor and run the latest, more graphically intensive games.

AMD did not offer us the full range of specs, which makes it hard to pin down just where this GPU will land in terms of raw performance compared to Nvidia’s latest cards. While the number of RDNA 4 compute units—the core clusters on AMD cards that process the thousands of calculations necessary for graphically intensive tasks—offers a vague impression of performance compared to the RX 9070, AMD didn’t provide any charts to compare FPS between games. The GPU runs on a 3.13GHz boost clock and has between 150W and 182W of board power compared to the 2.54 GHz clock and 304W board power on the company’s Radeon RX 9070 XT.

Without a price tag, it’s impossible to judge how much of a step down the latest card is compared to the RX 9070. AMD didn’t offer any word on a non-XT variant, either. The card will require a PCIe 5.0 x16 interface, the same as its other cards. AMD doesn’t craft its own GPUs and instead relies on AIC (add-in card) makers to produce its cards. We’ll update this article if AMD announces details on price or availability during its Computex keynote.

The crown jewel of AMD’s current lineup of graphics cards is the RX 9070 XT. AMD made headlines when it set the suggested sale price of the GPU at $600, only $50 more than the 9070, but it packs enough performance to get playable framerates out of multiple intensive games at 4K with a fair amount of ray tracing settings turned up. Unfortunately, because of a combination of tariffs and stock woes, the 9070 XT ended up priced at over $800 and as high as $1,000 at some online retailers.

We’ve seen prices fluctuate regularly over the past several months, but a near 20% price inflation to what should be a mid-range card is simply too much to stomach. However, the lower-end GPUs are faring better. The RTX 5060 Ti MSRP is set at $450, and the lowest price we’ve seen so far is $480. The $300 RTX 5060 is sitting closer to $320 from some AIC makers like Gigabyte. A fair number of Nvidia’s lowest-end GPUs are currently listed as “Out of Stock” or “Coming Soon” on sites like Newegg and Best Buy. Those buying a lower-end GPU are more price sensitive than people who can drop $2,000 on an RTX 5090 without blinking. AMD has even more impetus to set a price people can afford, and make sure it can keep costs level when the card finally hits store shelves.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Close