Skip to main content

Nvidia RTX 4090 GPU breaks world records with staggering 3.7GHz overclock

Nvidia’s RTX 4090 has been overclocked to a rather staggering 3.7GHz and as you might expect, when ramped up that far, the GPU claimed some new world records.

VideoCardz picked up on this one, reporting that overclocking outfit Team OGS took the Galax RTX 4090 HOF – the graphics card we wrote about yesterday that has two 16-pin power connectors, enabling it to be cranked to 1000W – and cooled it with liquid nitrogen to reach 3705MHz (3.705GHz) in the GPUPI benchmark.

OGS achieved 13th place in the rankings for GPUPI, but was the fastest result for a single GPU – the overclockers ahead of the RTX 4090 used multiple graphics cards (eight RTX 3090 GPUs in the case of the top spot – yes, eight).

Now, GPUPI is not a graphically intensive benchmark. In fact, it’s a compute test (calculating Pi to the nth degree with a GPU as the name implies), but the RTX 4090 also grabbed a world record in 3DMark Port Royal, where clock speeds weren’t quite as fast (as it’s a more intensive benchmark).

OGS still hit 3570MHz with liquid nitrogen, though, to score 31,096, enough to claim fifth place in the all-time rankings, but again the RTX 4090 was the top performing solo GPU – the four better scores were achieved by twin RTX 3090 graphics cards.


Analysis: Bigger things to come, no doubt

This is impressive stuff, with the Galax RTX 4090 HOF beating out the other RTX 4090 card in the top rankings here (a Colorful model from China, also cooled by liquid nitrogen) by just over 1,000 points.

And going forward, we can only expect that bigger feats will be achieved, and more records will fall, as is always the case when overclockers get more time to play with a GPU (or CPU) and push it even harder. Indeed, we’ll likely see other overclockers using the Galax HOF (‘Hall Of Fame’) RTX 4090 soon enough, as we know that Norwegian overclocker Rauf has got one (that’s where those leaked pics came from yesterday).

The HOF graphics cards from Galax have been around for quite some time (since Nvidia’s 900 series, in fact). They are specifically designed to be used in extreme overclocking scenarios, and with ridiculous wattages, like 1000W or more (we don’t actually know what the power usage was in these record attempts, as we’re not told, but you can bet it was seriously hefty). They’re not cards intended to be bought by consumers, with HOF models being pricey and produced in limited numbers for that niche overclocking audience.

Back in the real world, consumers who are well-off enough to afford an RTX 4090 for their PC are still worried about power consumption and reported incidents of melted cables, and we’re waiting to hear from Nvidia on this score.

Meanwhile, later today we can look forward to the launch of AMD’s RDNA 3 flagship and seeing how that will compare to Nvidia’s Lovelace top dog – with power efficiency expected to be much better for Team Red. How overall performance and relative pricing will stack up will, of course, also be key factors.



Source: TechRadar

Popular posts from this blog

Apple and Meta Reportedly Discussed AI Partnership for iOS 18

Apple has held discussions with Meta about integrating the Facebook owner's AI model into iOS 18 as part of its Apple Intelligence feature set, according to a report over the weekend. Meta launched Llama 2, its large language model, in July 2023, and in April, the company released the latest versions of its AI models, called Llama 3 . The Wall Street Journal reports that the two longtime rivals have held talks about offering Meta's model as an additional option to OpenAI's ChatGPT. The paywalled report notes that the discussions haven't been finalized and could fall through. As part of Apple Intelligence, Apple has announced a partnership with OpenAI that will allow Siri to access ChatGPT directly in iOS 18, iPadOS 18, and macOS Sequoia to provide better responses in relevant situations. Using ChatGPT will be optional, so users with concerns about the technology can abstain and still make use of Apple's own new AI features. Speaking at WWDC 2024, Apple's

Here Are the macOS Sequoia Features Intel Macs Won't Support

When Apple released macOS Monterey in 2021, some key features required a Mac with Apple silicon. The same scenario played out with macOS Ventura in 2022, and then again the following year with the release of macOS Sonoma. With macOS Sequoia set to arrive in the fall, which new features can Intel Mac owners expect to be unavailable to them this time around? Apple says that macOS Sequoia is compatible with the same Macs as macOS Sonoma, but Apple's fine print reveals that certain new features won't work on Intel machines. If you're still on an Intel Mac, here's what you won't have access to. Apple Intelligence Apple Intelligence , a deeply integrated, personalized AI feature set for Apple devices that uses cutting-edge generative artificial intelligence to enhance the user experience, won't be available on Intel Macs. Apple says the advanced features require its M1 chip or later, so if your Mac was released before November 2020, you're out of luck. T

iPhone 16 Pro Models to Adopt 'M14' Advanced Samsung OLED Panels for Improved Brightness and Lifespan

The upcoming iPhone 16 Pro and iPhone 16 Pro Max will be the first Apple smartphones to adopt Samsung's high performance "M14" OLED display panel, claims a new report coming out of South Korea. According to ETNews , Samsung's "M" series of OLED panels are made for flagship smartphones, while "14" refers to the number of high-performance materials used to produce them. "M14" is the first series of its kind, and the panel is said to have been developed to deliver superior brightness and longevity. Samsung has reportedly placed orders for the M14 materials and is preparing to mass produce the displays in the second half of the year for Apple's iPhone 16 Pro models. Google's Pixel 9 smartphone is the only other device that is expected to adopt the high-performance displays in 2024. A previous report out of China claimed that this year's ‌iPhone 16 Pro‌ models will feature up to 1,200 nits of typical SDR brightness – a 20%

Apple Boosts A18 Chip Orders in Anticipation of High iPhone 16 Demand

Apple is said to have upped its order of next-generation chips from TSMC to between 90 million and 100 million units, following heightened demand expectations for its iPhone 16 series. Last year's initial chip order volume for the iPhone 15 series launch is believed to have been in the region of 80-90 million units, suggesting Apple is anticipating higher demand for its 2024 devices, according to Taiwanese outlet CTEE . The arrival of Apple Intelligence in iOS 18 is expected to boost initial sales of the devices. One of the reasons is that Apple Intelligence requires at least an iPhone 15 Pro to run, which means owners of last year's iPhone 15 and iPhone 15 Plus will miss out on Apple's new AI features unless they upgrade to an iPhone 15 Pro or plump for one of the iPhone 16 models. Last year, the iPhone 15 and iPhone 15 Plus were equipped with the A16 Bionic chip – the same chip that was in the iPhone 14 Pro models – whereas the iPhone 15 Pro and iPhone 15 Pro Max f