Goto

Collaborating Authors

Nvidia promises insanely fast 360Hz 1440p eSports monitors

PCWorld

Nvidia has been touting the advantages of incredibly fast displays for a while now, showing off the way that 360Hz panels (showing six times the frames of a standard monitor) can enhance performance for eSports gaming. But those monitors make a sacrifice in resolution: 1080p, something that feels a little stifling when every TV on the Walmart shelf has four times that many pixels. At CES 2022, Nvidia announced a breakthrough for these high-speed monitors, a bump up to 1440p (2560 1440) resolution. Nvidia says that this next generation of 27-inch, 1440p panels can enhance the experience of eSports "athletes" with up to a 3% improvement in aiming due to the extra resolution. Naturally, these 360hz monitors also benefit from the usual Nvidia bells and whistles, including G-Sync to prevent screen tearing and a built-in performance overlay to show framer per second and latency with a keyboard shortcut, which Nvidia calls its "Reflex Analyzer."


The Morning After: Razer's light-up smart face mask finally goes on sale

Engadget

After a long run of teasing the RGB-lit Zephyr mask, Razer is finally ready to sell it to die-hard fans -- or possibly cosplayers. Alas, the Starter Pack is already listed as "out of stock," and the mask by itself is still "coming soon." Yes, you can customize the lighting through a companion app, but the highlight is a dual-fan active air filtration system with N95 filters -- something I'd never thought I'd be writing about a Razer device. Microsoft's dual-screen ambitions continue to struggle. Upgraded hardware and a new Glance Bar don't mean much when the Duo 2 is still plagued with inconsistent, finicky software.


Roku may undercut the new Apple TV with a cheap 4K streaming stick

Mashable

Streaming trailblazer Roku will reportedly challenge Apple, Google, and Amazon's streaming products with its forthcoming 4K Roku Streaming Stick . The site Zats Not Funny reports Roku will unveil a new line of streaming products this fall, including an updated version of its traditional stick design. Roku's current streaming stick, which sells for about $40, can carry a maximum resolution of 1080p. SEE ALSO: Apple's new Apple TV 4K is built to fuel its Hollywood ambitions Streaming sticks -- as opposed to "boxes" -- typically have fewer features than their larger counterparts, although those features tend to migrate downward a generation later. Most streaming boxes (including Roku) have been offering 4K resolution for over a year, so it makes sense that sticks would get the feature this year.


A 4-year-old adorably explains the problem with New Year's resolutions

Mashable

Following through on a big New Year's resolution may feel overwhelming, but this young girl can help with a few words of advice. "Keep your resolutions, but go easy on yourself. But it probably won't happen in one big moment. It'll happen in thousands of little moments." Listen to this wise-beyond-her-years 4-year-old and give yourself the flexibility to make change with small steps instead of big leaps.


Extremely Lightweight Quantization Robust Real-Time Single-Image Super Resolution for Mobile Devices

arXiv.org Artificial Intelligence

Single-Image Super Resolution (SISR) is a classical computer vision problem and it has been studied for over decades. With the recent success of deep learning methods, recent work on SISR focuses solutions with deep learning methodologies and achieves state-of-the-art results. However most of the state-of-the-art SISR methods contain millions of parameters and layers, which limits their practical applications. In this paper, we propose a hardware (Synaptics Dolphin NPU) limitation aware, extremely lightweight quantization robust real-time super resolution network (XLSR). The proposed model's building block is inspired from root modules for Image classification. We successfully applied root modules to SISR problem, further more to make the model uint8 quantization robust we used Clipped ReLU at the last layer of the network and achieved great balance between reconstruction quality and runtime. Furthermore, although the proposed network contains 30x fewer parameters than VDSR its performance surpasses it on Div2K validation set. The network proved itself by winning Mobile AI 2021 Real-Time Single Image Super Resolution Challenge.