Microsoft announced that the Xbox dev kit price is being increased from the original $1,500 to $2,000, marking a 33% jump in cost. This price hike, according to an official statement from Microsoft to The Verge, is a reflection of the macroeconomic developments. The price hike comes during a time when hardware and electronics prices are increasing as well. 

The Xbox dev kit is a combination of specialized hardware and software used by game devs to create and test games on Xbox. It comes with a Software Development Kit (SDK), a development console, and debugging features. 

The price increase saw criticism online. One Reddit user labeled the increase a “terrible business strategy.”

The highest subscription tier of Xbox Game Pass also recently saw price increases; some regions in the EU are reportedly exempt from the price increase. In addition, Microsoft previously announced the price increase of its flagship console, which now costs $599.99 from the prior $499.99. Sara Bond, president of Xbox, confirmed that the company is working on creating the next-generation console. 

These price increases affect not only RAM and SSD in retail, but OEMs, too. DRAM and NAND prices are climbing due to the higher demand for memory upcycle, attributed to AI’s increasing prevalence in the global economy. 

Qualcomm announced that it will acquire Arduino, and alongside the deal, unveiled the Arduino UNO Q. This $44 board computer pairs a Qualcomm CPU with a real-time microcontroller. Arduino states that the brand, open-source approach, and multi-vendor chip support will remain unchanged.

On October 7, 2025, Qualcomm announced that it had reached an agreement to acquire Arduino, the Italian open-source hardware platform used by a community of over 33 million active users, which “will gain access to Qualcomm Technologies’ powerful technology stack and global reach”.

However, financial terms were not disclosed. Both companies state Arduino will keep its name, tools, and mission, and will continue to support chips from multiple suppliers. The transaction is pending regulatory approval.

Arduino UNO Q

Alongside the announcement, Arduino introduced the UNO Q. The board combines a quad-core Cortex-A53 Qualcomm Dragonwing QRB2210 CPU running Debian Linux with a real-time STM32U585 microcontroller. The base model features 2 GB LPDDR4 and 16 GB eMMC, as well as Wi-Fi 5 and Bluetooth 5.1.

Additionally, The PCB ships with the new Arduino App Lab environment, which enables developers to combine Python apps, Arduino sketches, and lightweight AI models. The price starts at $44, with shipping available from October 24.

In short, Arduino gains Qualcomm’s resource and ecosystem reach, while Qualcomm gets a direct path into a massive DIY user base. However, the reaction of the Arduino user base is mixed for now, primarily due to the platform’s open-source nature, and partly due to the AI angle of the deal.

Qualcomm designs Snapdragon processors and platforms for smartphones, PCs, VR headsets, and other devices. For gaming, Qualcomm offers the Snapdragon Elite Gaming feature set and purpose-built chips like the Snapdragon G3x, which are used in various Android handhelds (Razer Edge, Logitech G Cloud).

Moreover, Meta’s Quest 3 runs on Qualcomm’s XR2 Gen 2 platform. Tying Arduino’s creator base to Qualcomm’s hardware strengthens its reach from mobile and VR into hobbyist controllers, DIY handheld devices, and, as seen in the announcement, cutting-edge AI projects.

Arduino is an open-source hardware and software company/platform for rapid prototyping, used in education, hobby projects, and industry, with a vast, active, and versatile community.

For gaming, this ecosystem underpins a significant share of DIY controllers and other peripherals, as many boards can present themselves as standard USB HID devices (gamepads and keyboards) and are supported by extensive community guides and libraries.

ASUS officially revealed its 30th anniversary edition ROG Matrix GeForce RTX 5090 GPU, which comes in red and black. Only 1,000 total units were produced, and they reportedly cost a little bit more than $2500, according to WCFTech.

The GPU uses fifth-generation tensor cores, maximizing AI performance with FP4 and DLSS 4. The hardware uses a four-fan cooling system, marketed to increase air flow and air pressure by 20%. 

Three of the fans face forward while the fourth fan faces backward, performing a push-pull air cooling. The 30th anniversary edition ROG Matrix GeForce RTX 5090 GPU offers BFT-supplied power delivery and a 16-pin 12V-2X6 connector, allowing the hardware to draw up to 800W of power, which is a marketed 10% performance boost from what is presumably Nvidia’s Founders Edition, according to PCGuide’s speculation.

ASUS’s premium PCB design uses a 3 Oz Copper PCB layer, facilitating heat dissipation and reducing voltage drops. The hardware features a 323 MHz factory overclock with a max clock speed of 2730 MHz.

The Graphics Engine is Nvidia’s GeForce RTX 5090, which features 32GB GDDR7, supporting an AI performance of 3593 TOPs and 21760 CUDA core, 28 Gbps memory speed, and 512 bit memory interface. The full specs can be found on the official ROG ASUS posting.

A multiplayer backend solution, Quark, that claims to remove hard player caps for online instances — allowing up to 100,000 players within the same shard — has entered early access for developers. The tech was developed by MetaGravity and was first used to build the upcoming large-scale PvP game Edge of Chaos.

In traditional online infrastructure, shards refer to isolated server instances that typically max out at 100–200 players. MetaGravity says Quark removes that ceiling entirely. The company stated that there was no actual hard limit, suggesting that games with 10,000 players per shard — and even 100,000 — were still possible.

In most modern online multiplayer games, the typical player limits per server instance, or shard, tend to vary by genre and infrastructure. Many battle royale and open-world PvP games such as PUBG, DayZ, Rust, Ark, and Escape from Tarkov usually support around 100 to 200 players per instance. Large-scale shooters like the Battlefield series typically cap out between 64 and 128 players. Some MMORPGs, including World of Warcraft and EVE Online, allow up to 250 players per zone or shard in typical engagements. Instances supporting over 1,000 players are rare and generally require heavy optimization or specialized infrastructure, as seen in games like Planetside 2 or EVE Online during large-scale battles where time dilation is used to manage server load.

Aside from removing the playercount cap, Quark also promises major cost savings for live service games. Quark developers claim that the traditional equation of one server per group of players no longer applies. In many cases, MetaGravity said, one virtual machine may be sufficient to run a large game instance, significantly reducing hosting and maintenance expenses.

Quark comes with official plugins for both Unity and Unreal Engine; developers interested in experimenting with Quark can apply for access on MetaGravity’s official site.

AMD denied rumors that the MSRP of the Radeon RX 9070 series would increase following its launch and very positive reviews from multiple publications and hardware-oriented YouTube channels.

The launch of AMD’s Radeon RX 9070 series GPUs was met with overwhelming demand, selling out within just 10 minutes of release. However, rumors surfaced that the launch MSRP was only temporary and that prices would rise after the initial batch sold out.

AMD has officially denied reports that the MSRP of the Radeon RX 9070 series was a “launch-only” price. According to Frank Azor, AMD is working closely with its partners to ensure that pricing remains stable despite the unprecedented demand. The company is also coordinating with AIBs (add-in board partners) to replenish inventory as quickly as possible.

The RX 9070 series has been praised for delivering strong price-to-performance value, making it a sought-after option for gamers looking for high-end performance at a reasonable cost. While AMD remains committed to keeping its MSRP intact, regional pricing fluctuations may still occur. It is worth noting that European markets are experiencing lower stock levels.

Game engine Unity reported a 25% decline in revenue in the fourth quarter of 2024, totaling $457 million compared to $609 million from the fourth quarter of 2023. The majority of its revenue stats appear to be lower than 2023’s figures, with Unity attributing the drop to their “portfolio reset.” 

Despite the evident drop in revenue, Matt Bromberg, President and CEO of Unity, stated “the Company’s fourth quarter results meaningfully exceeded expectations on both revenue and profit, underscoring our progress in building a new Unity.”

“The successful launch of Unity 6, the appeal of our new pricing model, and the progress we’re making in AI for our advertising customers are providing a lot of optimism for the future,” he continued. 

Unity underwent a major restructuring to its business model – which resulted in 1,800 lost jobs in early 2024. The revenue for the company supposedly reflected the restructuring efforts. The company is relying heavily on its newest version of Unity to keep the company’s revenue steady in the upcoming quarter. Unity 6 is reportedly the most stable the game engine has ever been, earning critical praise in the industry.

Unity is a game engine primarily used by a variety of game developers. The engine has user-friendly UI and cross-platform compatibility, making it a great engine for both indie and triple A developers alike.

Microsoft has unveiled Muse, a generative AI model designed to assist game developers and enhance gameplay experiences. Developed by Microsoft Research in Cambridge in collaboration with Ninja Theory (Enslaved, DmC, Hellblade franchise), Muse was trained on the multiplayer battle arena game Bleeding Edge.
A scientific article titled “World and Human Action Models Towards Gameplay Ideation” detailing how Muse works has been published in Nature.

According to the press-release, the model aims to “revolutionize game development” by understanding 3D game environments, physics, and player inputs, allowing AI to generate gameplay in a way that maintains consistency with established mechanics.

Microsoft claims Muse represents a “breakthrough” in prototyping gameplay ideas, introducing new content into existing games, and even player-driven AI-assisted content creation. Another use case mentioned in the anouncement is the game preservation—the company is exploring Muse’s ability to revive older games, potentially optimizing them for modern hardware without traditional remastering. This could allow classic titles, previously limited by outdated technology, to be playable on any device that supports Xbox services.

Microsoft has made Muse and its research openly available on Azure AI Foundry, allowing developers to experiment with the technology. The company also plans to launch interactive AI-driven game experiences on Copilot Labs in the near future, letting players experience AI-powered mechanics firsthand.

Microsoft’s Muse AI is not the only major advancement in generative AI for game development. In 2023, Google DeepMind introduced SIMA (Scalable Instructable Multiworld Agent), an AI system capable of learning and playing various video games across different genres. Google also showcased AI-generated game prototypes through Google Research’s GamenGEN, a model that recreated games without any traditional game engine, using only observational data.

Valve has released a massive update to the Source SDK, allowing content creators to build completely new games based on Team Fortress 2 (TF2). From now on, modders have access to the full client and server game code of TF2. 

The Source SDK is designed to let creators do much more than just tweak things. With access to the full game code, modders can change, extend, or even rewrite TF2 entirely, from tiny adjustments to fully-fledged game conversions.

It is worth mentioning that Source SDK is being released on a non-commercial basis, meaning all mods created with it must remain free to play while published TF2 mods can actually show up on the Steam Store, appearing as new games in the Steam library. While mods can’t profit off of the work created by the Steam Workshop community, Valve is keen to ensure that modders continue to respect the contributions of the original creators.

On top of that, Valve also brings a big update to all our multiplayer back-catalog Source engine titles like Day of Defeat: Source, Half-Life 2: Deathmatch, Counter-Strike: Source, and Half-Life: Source. Key improvements include 64-bit binary support, scalable HUD/UI, prediction fixes, and other optimizations.

Unity has laid off the team responsible for Unity Behavior, a visual tool that can be used to control NPCs and objects. The number of laid off individuals are currently undertermined.

First spotted by Gamigion, the layoffs were announced by the employees themselves. Nikita Zatsepin, a Senior Product Marketing Manager on Unity posted on LinkedIn that Unity has laid off “many talented and highly experienced people”. Soon enough, another post on LinkedIn gained traction, from a Senior Software Developer.

“After nearly 5 years @ Unity Technologies, I’m saddened to share that I am part of the recent layoffs.” read the post by Coline. It was only found out after ShaneeNishry, a Senior Software Engineer, posted about the layoffs on Unity’s forums.

The forum post is a farewell post saying that they cannot continue the development of the Behavior tool as the whole team is being laid off. According to Shanee, they also asked if the project could be open-sourced, however the company still hasn’t replied.

The Behavior tool can be used by developers to program NPCs or objects. However, the tool works on a graphical interface which means that it eliminates the need to meticulously type code to program each NPC to do certain tasks. Simply put, it saves a lot of time.

In January 2024, Unity implemented its largest round of layoffs to date, cutting approximately 1,800 jobs—roughly 25% of its workforce. The decision was described as a “company reset” by Unity chairman Jim Whitehurst, who framed it as a necessary restructuring effort. The layoffs came as part of a broader corporate shift following internal challenges and declining financial performance.

On November 7, 2024, Unity released their financial results for the third quarter. The company’s overall revenue was down 18% year-over-year, however their net loss went down from $126 million to $125 million quarter-over-quarter.

Unity’s layoffs are just one example of the wave of job cuts that have hit the gaming industry in recent months. A week ago, BioWare (Mass Effect, Dragon Age series) and Hi-Rez Studios (Smite 2) also announced staff reductions, with BioWare losing more than half of its workforce. Meanwhile, Microsoft—one of the industry’s biggest players—revealed its own layoffs less than a month ago, adding to the growing uncertainty within the sector.

Epic Games pinpointed the root cause of a widespread issue with stuttering affecting titles developed in Unreal Engine 5: there are too many shaders to load in real time, overwhelming the GPU.

When describing the reason for stuttering, Epic Games said, “a render engine discovers that it needs to compile a new shader right before it uses it for drawing something, so everything stops while waiting for the driver to finish compilation. To understand why this can happen we need to take a closer look at how shaders are translated into code that runs on the GPU.”

A GPU binary compiled for AMD is incompatible with NVIDIA, and even within the same vendor, instruction sets can vary between different hardware generations. In essence, Epic games cannot simply make a single code and distribute it out to fix the issue – if the problem was in relation to the CPU and not GPU, they could do that. 

Instead, Epic Games is introducing and recommending Pipeline State Objects precaching and Asynchronous compilation to resolve the issue. 

Pipeline State Object precaching is “a technique to determine potential PSOs at load time.” It essentially estimates a subset of possible, then attempts to precompile the necessary shaders in a cache, rather than relying solely on real time rendering alongside gameplay. 

If the game requires a previously compiled PSO (Pipeline State Object), the game pulls it from the cache rather than fully rendering it from scratch. 

This solution is limited in application, since the bundled cache is essentially useless if the environment changes according to the players’ actions. 

“Gathering it is resource-intensive and it must be kept up to date when content changes. The recording process might not be able to discover all the PSOs in games with very dynamic worlds: for example, if objects change materials based on player actions.

“The cache can become much larger than what’s needed during a play session if there’s a lot of variation between sessions, e.g. if there are many maps, or if players can choose one skin out of many,” said Epic Games’ representative.

Their workaround was to introduce PSO precaching, which determines potential PSOs at load time – this is not a perfect fix, as the subset is larger than what ends up being used, but it allows shaders to be compiled during loading.