Tag: Reviews

Introduction

“Kingdom Come: Deliverance II,” the long-awaited sequel, brings back the medieval intrigue and realism that captivated players in the original. Building on its predecessor’s success, this highly anticipated game expands its vast, historically rich world, placing players once again in the boots of Henry as he navigates a realm of political conflict, brutal combat, and deep personal quests.

With an enhanced open world, refined swordplay, and an even greater emphasis on player choice, the game delivers an immersive medieval experience like never before. Players can enjoy breathtaking visuals, customizable settings, and support for upscalers and improved quality settings vs console, bringing the 15th century to life in stunning detail.

Kingdom Come: Deliverance II is developed by Warhorse Studios, the team behind the original game, known for its commitment to historical accuracy and deep role-playing mechanics. Powered by an upgraded version of CryEngine, the sequel delivers great visuals, realistic melee combat, and a richly detailed medieval world. The PC version offers support for upscalers like NVIDIA DLSS and AMD FSR, but there is no support for ray tracing or frame generation.

This review will evaluate the performance of KCD 2 across a wide range of contemporary graphics cards, compare image quality settings, and analyze the game’s VRAM usage to provide insight into the hardware requirements needed for an optimal experience.

Introduction

Of the many, many audio brands that exist today, very few have impressed me like Dan Clark Audio (DCA) has. The DCA CORINA is an electrostatic set of open-back headphones which, when paired with my Headamp BHSE amplifier, continues to be one of my most preferred listening experiences for music, or even general media consumption. In fact, I had just emailed Dan late last year about how his recommended NOS tubes for the BHSE worked so well with vocals playback on the CORINA. He thanked me in the email and then teased me about something else coming up soon which was hopefully going to shake up the headphones market one more time. It wasn’t long before the brand announced the product in question—the DCA NOIRE X—and thanks to Dan Clark Audio for providing a review sample to TechPowerUp!

The Dan Clark Audio Aeon Closed X, developed in collaboration with Drop, was my first introduction to the brand. The Aeon series showcased DCA’s tuning philosophy, with the use of filters to help achieve a tonality more amenable to the average listener, and the goal of creating a sound signature akin to a set of speakers in a well-treated room. The tuning filters of course had limitations in what they could do, and driver dampening only goes so far either as other brands have found out. However, the industrial design of the Aeon series of headphones was a clear winner, with its human ear-shaped cups and minimal headband that made for some of the most comfortable headphones on the market today—all the harder to achieve with heavier planar magnetic drivers. Dan and his team soon developed the patent-pending acoustic metamaterial tuning system (AMTS) for better control over the headphone tuning, which we saw used in the closed-back flagship STEALTH and the open-back flagship CORINA. The DCA E3 came out earlier this year, bringing AMTS and a lot of what made the STEALTH popular at $2000—half the price of the STEALTH. Today we examine the new Dan Clark Audio NOIRE X, which now brings AMTS to an even more affordable price range to promise a smooth frequency response and detail on par with flagship headphones, and we begin the review with a look at the product specifications in the table below.

Specifications

Introduction

After being a PlayStation 5 exclusive for less than two years, Marvel’s Spider-Man has web-slinged to PC in all its glory. The full-fledged sequel to the 2018 game is bigger and prettier than the original. Instead of controlling just one Spider-Man like in the first game, and the follow-up Spider-Man Miles Morales, players can now sling around NYC as Peter Parker and Miles Morales, with the game supporting seamless transition between the two playable characters.

The open world set in New York City has been greatly expanded compared to the first two games. Aside from Manhattan, you can now freely explore Brooklyn and Queens as well. The PC version of the game comes with a selection of upgrades, including support for ultrawide monitors, a slew of improved and expanded RT effects, featuring all major upscaling technologies.

Marvel’s Spider-Man 2 supports AMD FSR, NVIDIA DLSS, and Intel XeSS and also includes Insomniac’s own upscaling solution called IGTI (Insomniac Games Temporal Injection). While Insomniac had been at the helm during the development of the PS5 version of the game, the PC port duties were bestowed to Nixxes, which have ported pretty much all PC versions of Sony’s first-party PlayStation 4 and PlayStation 5 exclusives.

While the game has relatively modest system requirements (for a PS5 exclusive), the question arises: is it playable on the two most popular PC gaming handhelds: the Steam Deck and the ASUS ROG Ally? In this performance review we will try to answer that question and provide optimized settings for both devices.

Introduction

“Marvel’s Spider-Man 2” swings onto PC, bringing the next chapter in Sony’s acclaimed superhero game series. Following the success of its predecessors, this highly anticipated sequel expands the open-world action of New York City, allowing players to step into the suits of both Peter Parker and Miles Morales as they face new threats, including the formidable Venom.

The game features an expanded open world, enhanced traversal mechanics, and deep character-driven narratives, promising the ultimate superhero adventure. Now optimized for PC after being a Sony PS5 exclusive for more than a year, players can enjoy stunning visuals, customizable settings, and cutting-edge performance, bringing the excitement of being Spider-Man to a whole new level.

Spider-Man 2 is developed by Insomniac Games, the renowned studio behind the previous Spider-Man titles and many other Sony game PC ports. Powered by the company’s own, in-house “Insomniac Engine,” the game delivers breathtaking visuals, fast-paced action, and a richly detailed open-world New York. On the PC version you get support for various levels of ray tracing, but RT can be disabled, too. There’s also support for NVIDIA DLSS Ray Reconstruction, Upscaling and Frame Generation, AMD FSR Upscaling and Frame Generation and Intel XeSS Upscaling.

This review will evaluate the performance of Spider-Man 2 across a wide range of contemporary graphics cards, compare image quality settings, and analyze the game’s VRAM usage to provide insight into the hardware requirements needed for an optimal experience.

Introduction

For the past decade we’ve seen the power envelopes of CPUs and GPUs being pushed further and further to maximize performance, while at the same time the same components have gotten much more effective at conserving power when not fully loaded. This wide power consumption gap between a lightly loaded and a fully loaded system. In combination with intelligent fan controls, this makes it possible to have a near silent machine when doing light office work that turns into something that makes you put on your headphones on when you start your game.

With power supplies getting more and more efficient we have also benefited from semi-passive operating modes. While some models have fairly conservative fan ramp up profiles, others will turn up the fan RPMs quite rapidly. So if noise is something that you absolutely want to minimize, you can either carefully pick the PSU that has the best acoustic real world performance, or to be 100 % sure, pick a power supply that’s completely fanless and will remain silent even delivering its maximum rated power.

Today we’re looking at the Cooler Master X Silent Edge Platinum (850 W)—a fully passive power supply, which, at least on paper looks like a great option for a really powerful single-GPU rig. With advertised 80 Plus Platinum efficiency levels, ATX 3.1 support and an industry leading 15-year warranty, there’s not much more one could ask for. Does that translate into excellent real world performance? We’ll see. But for now let’s start with the vendor-provided specification summary.

Specifications

Introduction

MSI debuts its new Vanguard line of graphics cards with the GeForce RTX 50-series, and we have with us the MSI RTX 5080 Vanguard SOC graphics card, in its Launch Edition collector’s package. The new Vanguard line of graphics cards are positioned between MSI’s SUPRIM line of halo-segment custom design graphics cards, and its more popular Gaming X series. The company retired the Gaming Z series with the advent of SUPRIM, so I guess you could call Vanguard the spiritual successor to the Gaming Z. The MSI Vanguard series combines aspects of both SUPRIM and Gaming X. From the SUPRIM you see hints of titanium gray metal alloy embellishments on the cooler shroud. The backplate has a matching brushed-aluminium appearance, and from the Gaming X series, it has certain sporty elements, such as carbon fiber surfaces and acrylic RGB LED lighting diffusers. Under the shroud, the Vanguard features many of the cooler innovations MSI introduced with the SUPRIM series this generation.

The GeForce RTX 5080 is NVIDIA’s latest enthusiast-segment graphics card, designed for maxed out gaming at 4K Ultra HD, with ray tracing enabled. The flagship RTX 5090 from last week does the same thing, but faster, and has other use cases such as AI development and content creation, thanks to its much larger video memory and memory bus. The RTX 5080 squarely targets gamers looking for a 4K gaming experience, with a few allied workloads on the side, such as game streaming. If compare the specs sheets of the RTX 5090 and RTX 5080, you’ll find that the latter is almost half the card the RTX 5090 is, and is yet squarely an enthusiast-segment GPU. The RTX 5080 debuts the new GB203 silicon, which is the second largest to implement the GeForce Blackwell graphics architecture. With this generation, NVIDIA is holding on to the process node, so these chips are built on the same NVIDIA 4N silicon fabrication node that the company codeveloped with TSMC, based on the foundry’s 5 nm EUV node.

The GeForce Blackwell graphics architecture introduces a revolutionary new concept in consumer 3D graphics, called Neural Rendering. The company figured out a way to combine objects created by a generative AI with conventional raster 3D graphics, just like it discovered a way to combine real time ray traced objects with RTX late last decade. You already know the ability of popular generative AI models to create photorealistic images and video, so you know this technology has immense potential. GeForce Blackwell GPUs have the ability to accelerate generative AI and render 3D graphics in tandem, thanks to a new hardware scheduling component called the AI management processor. The architecture also introduces the ability for 3D applications to directly access the Tensor cores, and NVIDIA worked with Microsoft to standardize this at the DirectX API level.

The new Blackwell SM features concurrent FP32 and INT32 capability on all 128 CUDA cores, the previous generation Ada SM only has INT32 capability on half its cores. The 5th Generation Tensor core puts on FP4 data format capability to step up throughput by trading in precision. The 4th Generation RT core has the hardware groundwork for Mega Geometry, letting ray traced objects have exponentially higher polygon counts. DLSS 4 and Multi Frame Generation are this generation’s other major highlights. DLSS 4 introduces a new Transformer based AI model replacing the CNN-based one, for all its subcomponents, namely, super resolution, ray reconstruction, and frame generation. This model is more accurate, and hence steps up image quality in every performance tier. Multi Frame Generation gives the GPU the ability to use AI to draw not just one succeeding frame to a conventionally rendered one, but four such succeeding frames, effectively quadrupling frame rate. It relies on the new Flip Metering component of the Blackwell Display Engine to achieve this.

Blackwell implements many technological firsts for the company. It’s the first gaming GPU to use the PCI-Express 5.0 x16 host interface, the 12V-2×6 power connector, and feature an ATX 3.1 + PCIe 5 CEM power architecture. It’s also the first generation to implement the new GDDR7 memory standard, which doubles efficiency over GDDR6. The Blackwell display engine implements DisplayPort 2.1b with UHBR20 on all ports. When used with DSC, it enables 8K high refresh-rates with a single cable. You can see where this is going with DLSS 4 MFG.

The GeForce RTX 5080 maxes out the GB203 silicon, enabling all 84 SM, which works out to 10,752 CUDA cores, 336 Tensor cores, 84 RT cores, 336 TMUs, and 112 ROPs. It also features all 64 MB of the on-die L2 cache, and two sets of NVDEC and NVENC video accelerators that support 4:2:2 formats for AV1 and HEVC. The chip features a 256-bit wide GDDR7 memory bus, which drives 16 GB of memory. NVIDIA is running this memory at 30 Gbps on the RTX 5080, yielding 960 GB/s of bandwidth—a 34% increase over that of the RTX 4080.

The MSI GeForce RTX 5080 Vanguard SOC, as we said, combines the cooling innovations of the SUPRIM series, with the aesthetics of both the SUPRIM and Gaming X series. The new Hyper Frozr cooling solution comes with a vapor chamber plate making contact with the GPU and memory, which conveys heat to a cluster of heat pipes that have been flattened to a cuboidal shape for better contact with the vapor chamber. From here, the heat is spread across a large aluminium fin-stack with wavy edges that increase turbulence for better heat dissipation. The cooler relies on a trio of MSI’s StormForce fans, the same exact ones found in the company’s RTX 5080 SUPRIM graphics card which we are also reviewing today. There are two sub-variants of the RTX 5080 Vanguard, SOC and OC, we’re reviewing the SOC, which comes with a higher factory overclock of 2730 MHz on both the Gaming and Silent BIOSes, compared to 2617 MHz reference. The Launch Edition is a special package that includes a puzzle toy based on Lucky, MSI’s dragon mascot. MSI is pricing this card at $1230, a $230 premium over the $999 baseline.

Introduction

NVIDIA only just kicked off its GeForce RTX 50-series Blackwell generation last week, and we already have our second GPU from series, the new GeForce RTX 5080. We have many of these cards with us to review, but today we bring you the NVIDIA GeForce RTX 5080 Founders Edition, their de facto reference design. The RTX 5080 is being launched as the generation’s second-best graphics card, and does many of the same things as the RTX 5090, but at a much lower price-point. NVIDIA recommends the RTX 5080 for maxed out gaming at 4K Ultra HD—the same use-case its predecessor the RTX 4080 was launched at. You get half the amount of memory as the flagship RTX 5090 at 16 GB, even the memory bus width is half of it, at 256-bit, and yet there’s plenty of muscle in this card for gamers to play today’s and tomorrow’s games at 4K UHD.

The GeForce RTX 5080 debuts NVIDIA’s second silicon from the GeForce Blackwell generation, the GB203. This is a physically smaller chip than the GB202 powering the RTX 5090—in fact it’s half its size. What’s interesting though, is that the RTX 5080 maxes out all available SM on the GB203. If NVIDIA has to release an “RTX 5080 SUPER” next year, it would have to tap into the much larger GB202. In case you missed it, NVIDIA is building Blackwell on the same exact foundry node as the RTX 40-series Ada from October 2022, the TSMC 4N, which is a specialized variant of the 5 nm EUV node that the foundry co-developed with NVIDIA. The GB203 has similar transistor counts and die-size to the AD103 silicon powering the RTX 4080.

The new GB203 silicon physically has 84 Blackwell streaming multiprocessors (SM) across 7 GPCs, and the RTX 5080 has all of them enabled. This works out to 10,752 CUDA cores, more than the 10,240 that the RTX 4080 SUPER comes with. It also comes with 336 Tensor cores, and 84 RT cores. The TMU count is 336, while the chip has 112 ROPs. The shared L2 cache size is unchanged from the previous generation at 64 MB. Memory bandwidth is where the RTX 5080 gets a significant upgrade over the RTX 4080. While the bus width is the same 256-bit, it drives newer GDDR7 memory chips, and NVIDIA picked memory speeds of 30 Gbps, yielding 960 GB/s of memory bandwidth (a 34% increase over the RTX 4080). The media and display engines of the RTX 5080 see generational updates over the RTX 4080, too, you get two each of the latest Blackwell NVDEC and NVENC accelerators. The display engine supports DisplayPort 2.1b with UHBR20, and HDMI 2.1a.

The new GeForce Blackwell graphics architecture lays the hardware-level groundwork for neural rendering, a revolutionary new concept in consumer 3D graphics, where generative AI plays a participatory role in the core rendering stack, and not just part of the DLSS Super Resolution feature, where it helps reconstruct details in upscaled frames. Just as NVIDIA discovered a way to combine certain real-time ray traced elements with classic raster 3D, it found a way to combine objects created by a generative AI with raster 3D. The company worked with Microsoft to standardize this in the DirectX 12 API, letting 3D applications directly access Tensor cores. The GPU also runs generative AI and 3D graphics rendering workloads in tandem thanks to a new hardware scheduler component called the AI Management Processor (AMP).

Blackwell also introduces DLSS 4 and Multi Frame Generation. DLSS 4 replaces the CNN (convoluted neural network) based AI models driving Super Resolution, Ray Reconstruction, and Frame Generation, with a new transformer based AI model that improves image quality at every performance tier. The updated AI models are available to even the RTX 40-series, RTX 30-series and RTX 20-series GPUs in games that implement DLSS 4, however Multi Frame Generation (MFG) is exclusive to the RTX 50-series. MFG is a technology that leverages AI to draw not just the frame after a conventionally rendered frame, but up to three succeeding frames. It requires the new Flip Metering hardware of the Blackwell display engine, which is what makes MFG exclusive to this generation.

The NVIDIA GeForce RTX 5080 Founders Edition features an identical-looking design to the RTX 5090 Founders Edition we reviewed last week, down to the same length, height, and 2-slot thickness. You wouldn’t be able to tell the two apart installed until you pay attention to the decals on the backplate. This card features the same Double Flow Through thermal solution where the PCB is shrunk down to the bare minimum size and relocated to the center of the card, relying on breakaway PCBs for the display and PCIe I/O. The design allows for airflow from both fans to flow through heatsink fins of the cooler, venting out the back of the card. This cooler did wonders with the RTX 5090, which has a TGP of 575 W, and so it’s only going to do better with the RTX 5080 and its 360 W TGP. NVIDIA is pricing the RTX 5080 Founders Edition at $999, which is also the starting price for this SKU.

Introduction

Softears, formally known as Softear Acoustics, is a Chinese IEM brand that was conceptualized in 2014, with a studio started in Shenzhen in 2017 before moving to Chengdu in 2019 where it operates an independent R&D lab and its own production facilities—not too different from FiiO’s industrial park that we visited recently. If some of this sounds familiar to you then you are probably too involved with IEMs and may need to take a break for your own mental health. Jokes aside, some of this story is also shared by MOONDROP, arguably the most famous IEM brand among the younger audiophiles today. Both MOONDROP and Softears operate out of the same industrial park and share some resources too. I understand that MOONDROP’s CEO invested in Softears early on with the goal of helping the latter be a higher end, R&D-centric brand that puts out more premium products. In return, MOONDROP might use a Softears-developed driver in its IEMs or have some Softears personnel help with the R&D and manufacturing of some products. Softears is even considered as a sister brand of MOONDROP in this regard, although it operates fairly independently when it comes to its own products. This is also why I’ve had no issues getting MOONDROP samples in for review, yet had to wait until today for our first Softears product review.

One of my goals attending the Shenzhen International Audio Show recently was to finally meet Softears. The brand had a booth at the show—in a completely different section from MOONDROP, for anyone wondering—and had a few different products for people to try out. The Enigma, which is Softears’s new flagship, immediately got my attention and it was easily one of the best IEMs I’ve tried at the show. Understandably, Softears was reluctant to send out the expensive Enigma for review, but they did offer the brand’s previous release in the form of the Studio4. This set released in March 2023, making it almost ancient in a world where IEMs release and get discontinued within a calendar year at times, if not sooner. However, this is part of what makes Softears different from the norm—its IEMs have a longer product life cycle and get supported far beyond your typical IEM brands even contemplate. This also means any Softears product has undergone a longer R&D phase too, with a clear purpose that needs to be fulfilled before it is deemed ready.

The Studio4 is a four balanced armature driver set that is developed with studio recording, music mixing, and stage monitoring in mind. It’s effectively an IEM analog of the Sennheiser HD 490 PRO we saw recently, and has a different role to play. Yet it has quickly become a favorite among audiophiles for its sound signature and tuning to where you will see the Studio4 near the top of many reviewer/influencer ranking lists, and it often gets recommended by the average IEM enthusiast too. I heard the Studio4 at Shenzhen and was immediately interested as a result of its comfortable fit and clean, yet engaging sound. Knowing that show floor conditions can be tricky, I of course agreed to their offer of a review sample to be able to try this out under familiar conditions. Note that there is also a newer “Starry Version” of the Studio4 that released recently, using clear resin shells and a sparkly blue faceplate, which seems to be exclusive to HiFiGo as of the time I write this review. Both versions should be otherwise identical elsewhere, including the sound which matters more, so this review should be representative of either in those metrics. Thanks to Softears for providing a review sample to TechPowerUp, and let’s begin this review with a look at the product specifications in the table below.

Specifications

Introduction

We have with us the ASUS ROG Astral GeForce RTX 5090 OC graphics card, which debuts the new ROG Astral brand for ASUS. Designed to provide the best of ASUS thermal engineering and industrial design, the ROG Astral brand is positioned a notch above the ROG Strix brand, providing the company’s best air cooling solution, a premium set of materials, and the new Quad Fan Force arrangement of no less than four fans. Aesthetically, ROG Astral graphics cards pair well not just with ASUS ROG Strix series motherboards, but also the more premium ROG Maximus and ROG Crosshair lines. The GeForce RTX 5090 is designed to crunch through any of today and tomorrow’s games with maxed out settings, at 4K Ultra HD.

The Blackwell graphics architecture powering the RTX 5090 introduces neural rendering, a breakthrough new concept where generative AI works more collaboratively with classic raster 3D graphics. You’ve had a taste of generative AI, and its ability to conjure up photorealistic images and videos. Now imagine AI drawing parts of your 3D scene in real time, complete with geometric detail and ray tracing effects. Making this possible is API-level standardization that allows games to access Tensor cores, and the ability for the GPU to accelerate a generative AI model and render games in tandem, thanks to a new hardware scheduling component called AMP.

Perhaps the most interesting aspect about Blackwell is that GPUs in its generation do not introduce a new foundry node, all chips in the RTX 50-series are built on the existing TSMC 4N process, which is an NVIDIA-specific variant of the 5 nm EUV node. Whatever generational improvements in efficiency you see are purely thanks to advances made by the graphics architecture itself, and a re-architected power management system.

The RTX 5090 is based on the GB202 ASIC, a massive 750 mm² slab of silicon featuring over 92 billion transistors, 192 Blackwell streaming multiprocessors, a PCI-Express 5.0 x16 interface, and a mammoth 512-bit GDDR7 memory interface, belting out 1.79 TB/s of memory bandwidth. The trend across the RTX 50-series is large increases in bandwidth thanks to GDDR7, because neural rendering and the new DLSS 4 Multi Frame Generation are memory sensitive technologies. The RTX 5090 is carved out of the GB202 by enabling 170 out of those 192 SM, and enabling 96 MB out of the 128 MB on-die L2 cache available. This results in 21,760 CUDA cores, 680 Tensor cores, and 170 RT cores, across 11 GPCs, and this doesn’t even max out the GB202. The card comes with 32 GB of 28 Gbps GDDR7 memory.

The Blackwell graphics architecture introduces a new generation streaming multiprocessor with concurrent FP32 and INT32 capability across all its 128 CUDA cores, shader execution reordering with awareness for neural shaders, and the new 5th Gen Tensor core that’s capable of FP4 data formats for 32x the throughput of the original Tensor core. The new generation RT core has the hardware groundwork for Mega Geometry, the ability to give ray traced object significantly higher poly counts, or those many triangles onto which rays should interact with. DLSS Multi Frame Generation is a technology that uses AI to predict not just every other frame following a conventionally rendered one, but up to three frames following it, effectively quadrupling frame-rates (or at least the smoothness of output).

The ASUS ROG Astral RTX 5090 OC doesn’t just come with a heavy cooling solution that fits into the dimensions of the previous generation RTX 40-series ROG Strix graphics cards, but also dials things up a notch with the introduction of a fourth fan arranged along the tail end of the backplate where you normally expect a cutout to be, for airflow from the third fan to go through. This fourth fan acts as a “pull” fan, increasing the overall airflow volume of the heatsink by around 20%. There’s plenty of tastefully executed dual-tone surfaces, rich metal alloy textures, and ARGB LED lighting. Innovations you expect from ROG Strix series cards are also here, such as dual-BIOS, case fan headers, and ARGB headers. ASUS is giving the RTX 5090 a factory overclock of 2580 MHz compared to the 2410 MHz reference GPU Boost frequency. The cooling solution is tasked with ensuring higher boost frequency residency and lower noise. Since this is the company’s most premium air-cooled custom design RTX 5090, ASUS is pricing it at USD $2,800, an astounding 40% premium over the NVIDIA baseline price.