What's new in Linux 7.0: key changes in the new kernel

Last update: 18/04/2026
Author Isaac
  • Linux 7.0 inaugurates a new kernel series with stable Rust, support for next-generation hardware, and multiple internal optimizations.
  • The file systems, memory management, scheduler, and swap have been improved to deliver a faster and more efficient Linux.
  • Security is reinforced with post-quantum cryptography, improvements to SELinux and BPF filtering for io_uring, targeting demanding environments.
  • The kernel incorporates advanced support for networking, virtualization, NPUs, and local AI, establishing itself as the foundation for modern clouds, servers, and desktops.

What's new in Linux kernel 7.0

The result is a much more rounded kernel.With improvements in virtually every area: support for current and future hardware (Intel Nova Lake, AMD Zen 6, Snapdragon X2 Elite, ARM64, RISC-V, LoongArch…), faster and more resilient file systems, meticulously polished performance, enhanced security with post-quantum cryptography, and clear policies on the use of AI tools in development. It's not just a simple change of version: on a practical level, you'll notice a more agile, modern Linux, ready for what's to come.

Why does Linux 7.0 exist and not Linux 6.20?

The origin of Linux 7.0 is as pragmatic as it is peculiarLinus Torvalds decided that the 6.x series was already stretching too far, with versions like 6.19 just around the corner, and that adding more sub-versions would become a mess. True to his now-famous joke about "running out of fingers" to count minor versions, he opted to reset the counter and jump straight to 7.0.

There is no philosophical break behind this change in number.but rather a very strong tradition in kernel development: every time the second digit approaches 20, a new, larger series begins. This keeps the numbering manageable and, incidentally, serves to group large blocks of work historically: 6.x remains as a very mature and stable stage, and 7.x starts prepared for new architectures, standards, and ways of programming the kernel.

There is also a point of technical strategy.: close stage 6.x with the kernel 6.19 This allows that branch to solidify as a stability benchmark, with accumulated patches, server improvements (such as Live Update Orchestrator for hot updates), and advancements in areas like HDR and passkeys. Version 6.19 remains the "trusted" core for many distributions and critical environments, while version 7.0 assumes the role of the future foundation, with more ambitious changes.

Linux 7.0, moreover, fits into a very specific scheduleDevelopment of the branch began after the release of stable version 6.19, following the classic cycle of approximately ten weeks from the first Release Candidate. The stable release is slated for mid-April 2026, with tentative dates such as April 5, 12, or 19 depending on the number of Release Candidates required. All of this aligns with the expectation that 7.0 will be the flagship kernel for Ubuntu 26.04 LTS and many rolling release distributions that will adopt it almost immediately.

On a symbolic level, the jump to 7.0 also marks the end of an era.Years of debates, mailing list discussions, subsystem refinement, and quiet preparation for profound changes (like Rust or NPUs) converge in this version. It's not the "Linux desktop revolution" some dream of every year, but it is the point where next-generation hardware, advanced security, and a cleaner codebase align.

Rust consolidates its place in the kernel: goodbye to the experiment label

Rust support on Linux 7.0

One of the big headlines of Linux 7.0 is the integration of Peace as a stable part of the kernel. After several years of work, numerous patches and a prolonged testing phase marked with the "experimental" label, the Rust-for-Linux project ceases to be an experiment to become an official part of development.

For more than three decades, the heart of Linux has been written almost exclusively in CJavaScript, a powerful language, is also prone to memory errors that are an attacker's dream. Rust comes into play precisely to address these kinds of structural flaws: its memory ownership model prevents many classic bugs and drastically reduces a significant portion of the serious vulnerabilities we see year after year.

Work to bring Rust to the kernel started about three years agoWith Miguel Ojeda leading the initiative, the symbolic change is complete in Linux 7.0: the patch that closes the testing phase and marks Rust as "stable" bears his signature, and the consensus of the maintainers was made clear at the 2025 Linux Kernel Maintainers Summit. The message is resounding: the experiment is over, Rust is here to stay.

That doesn't mean C will disappear from the mapThe bulk of the kernel will remain in C for many years, and no one plans to rewrite the kernel from scratch. But starting with Linux 7.0, new drivers and subsystems can be born directly in Rust, taking advantage of its memory safety. For those who want to delve deeper into the challenges, there are articles that analyze the disadvantages and challenges of using Rust in the kernel.

In parallel, the project has adopted an official policy on tool-generated contentincluding AI-powered coding assistants. The kernel opens the door to contributions that have used AI, but establishes clear rules to ensure that the code is understandable, reviewable, and legal. You can read more about it. policy on AI-generated codeAll help is welcome, yes, but subject to a framework that protects the quality and integrity of the project.

File systems: Btrfs, XFS, EXT4, EROFS and more

File system improvements in Linux 7.0

File systems are once again the main focus in Linux 7.0With changes that focus on performance, reliability, and flexibility, this update offers significant benefits if you work with large volumes of data, file servers, or simply want a more responsive desktop.

  How to restore the original Windows bootloader after installing Linux or another operating system

Btrfs gains direct I/O support when the block size exceeds the page sizeThis is key for certain workloads where reducing the overhead of the kernel page cache is a priority. This improvement makes it more competitive in environments where other systems were previously favored simply because of their direct access performance.

XFS debuts a very interesting autonomous self-correction capabilityA background daemon, integrated with systemd, is introduced, capable of automating repairs even with the file system mounted and in use. In practice, this means fewer service outages and greater resilience to errors that previously required a manual fsck or a maintenance window.

EROFS, the read-only file system optimized for immutable distributionsIt enables LZMA compression by default. This choice combines a very high compression ratio with reasonable performance, ideal for system images where saving space is paramount without sacrificing too much speed. To better understand immutable images, see what an immutable image is. immutable Linux distribution.

EXT4, the trusted veteran in millions of installationsIt also receives some love: the handling of concurrent writes is refined, making scenarios where multiple processes write simultaneously more robust. Per-CPU cache management is improved, reducing bottlenecks, especially on modern SSDs, resulting in lower latencies and faster application startups.

NTFS3 and exFAT continue their path to maturity within the main kernelNTFS3 gains internal changes focused on stability and performance, improving interoperability with partitions used in Windows. exFAT, widely used on SD cards and USB drives, continues to refine its behavior to be a reliable option for removable devices.

Furthermore, Linux 7.0 introduces NULLFS as a minimal “pseudo-system”This is interesting from a technical standpoint, although it won't have a direct impact on most users for now. Work also continues on F2FS and other flash memory systems, aiming to optimize device performance and lifespan.

Performance: swap, zram, scheduler and a more agile system

Performance and swap in Linux 7.0

In terms of performance, Linux 7.0 doesn't boast a single magic changeRather, it's a combination of optimizations that, together, make the system feel faster and smoother. Many are invisible to the user, but they are noticeable in response times, resource consumption, and stability under load.

The swap table is undergoing its second major reorganization after the work started in Linux 6.18This redesign improves read speeds from swap to RAM, with up to a 20% improvement in scenarios where multiple processes share swap space. On systems with limited RAM or heavily loaded systems, this change can make the difference between a sluggish system and one that performs well.

zram users also particularly benefitThe kernel no longer needs to decompress pages before writing them to disk, as it can do so directly in compressed format. This reduces CPU load and improves overall efficiency when combining zram with persistent storage devices.

The task scheduler receives a key upgrade with the Time Slice Extension (TSE)For years, many users complained about the infamous micro-stutter: brief interruptions or stutters when the scheduler cut off CPU time to a task at a critical moment. With TSE, relevant tasks can retain the processor a little longer before being interrupted, smoothing the experience on desktops, in games, and in interactive applications.

Work has also been done on the creation and destruction of threadswhich is now between 10% and 16% faster, and file opening and closing operations, improved between 4% and 16%. These percentages, in isolation, may seem modest, but added to everything mentioned above, they make the system react better to intensive loads.

Regarding power, the old laptop_mode function is permanently removed.Designed for mechanical hard drives, this mechanism offered little benefit in the SSD era and complicated kernel maintenance, leading to a preference for simplification and modern power management strategies tailored to current hardware.

In modern Intel CPUs, TSX comes back into play for certain generationsThis feature is reinstated in 10th generation processors and later when the kernel determines that the model is not vulnerable to known vulnerabilities. Activation is managed by the kernel itself, which automatically decides based on the CPU's characteristics, thus allowing for multithreaded performance improvements without compromising security.

Networks, virtualization and cloud: the path to Wi-Fi 8 and more secure enclaves

The networking and virtualization field also receives a significant boost in Linux 7.0, with changes that matter as much to someone who has a server at home as to large cloud companies like Meta or Amazon.

One of the notable new features is general support for AccECNAccECN is a more precise congestion notification mechanism than previous systems. It allows connection endpoints to react more precisely to network congestion, improving latency and performance on complex connections without requiring unusual application-level workarounds.

In the world of virtual machines, support for network namespaces in VSOCK sockets is introduced.This facilitates more secure and isolated communication between guest and host. In parallel, KVM adds support for ERAPS (Return Address Prediction Technology) on AMD Zen 5 processors and higher, improving virtualization performance on the latest hardware.

  Full tmux integration in Arch-based Omarchy

Linux 7.0 also lays the first bricks of the Wi-Fi 8 (802.11bn) standardintegrating initial support that will be expanded in future versions. This ensures that, when compatible routers and cards become more widespread, the kernel will already be prepared to take advantage of them.

In the cloud, the focus is on extreme data isolation.Large platforms are striving to make the information traveling through their infrastructure opaque, even to administrators. Along these lines, Linux 7.0 advances in isolating virtual machines through encrypted memory enclaves, so that sensitive payloads (such as banking data, private documents, or personal photos) can be processed without anyone outside the authorized code having access to their contents.

Small but important adjustments in connection management complete the picture.This includes improvements to network stacks and drivers, and preparation for emerging technologies. All of this keeps Linux as the preferred choice for data centers, containers, and high-traffic services.

Enhanced security: post-quantum cryptography, SELinux, and BPF

Security is another pillar in which Linux 7.0 is stepping on the gaspointing to both current threats and medium and long-term scenarios where quantum computers come into play.

One decisive step is the elimination of SHA-1-based module signature schemesThis algorithm is already considered insecure. Instead, the kernel adopts post-quantum ML-DSA signatures, designed to withstand attacks even from powerful future quantum systems. This aligns the kernel with the industry trend toward quantum-resistant cryptography.

SELinux, the most widespread mandatory access control system in the Linux worldIt is further strengthened with support for BPF tokens as an additional control mechanism. This allows for the definition of more sophisticated and dynamic policies, tailored to complex scenarios where traditional rules fall short.

Another key change is the BPF filtering applied to io_uringThe asynchronous I/O interface has become increasingly widespread due to its ability to reduce system calls and improve performance. With Linux 7.0, administrators can sandbox and filter which io_uring operations are allowed, without having to disable all functionality when security needs to be strengthened in demanding environments.

In terms of user authentication, the 6.19 series had already laid the groundwork with a standard API for Passkeys.This allows for a passwordless experience, eliminating traditional text-based passwords and managing cryptographic keys directly from the operating system. Linux 7.0 builds upon and consolidates this previous work, facilitating the use of more robust authentication and better integrating biometrics and external security devices.

Cleaning up old code also contributes to security.: to the remove obsolete drivers and subsystems that supported virtually extinct hardware, reduced attack surfaces and complexities that barely added value, but did add potential risks.

Hardware support: from legacy to next generation

If there's one thing that characterizes Linux 7.0, it's its ambition in hardware support., ranging from veteran equipment that can still benefit from a second life to platforms that have not yet reached the market.

In CPUs, the kernel comes ready for the new generations of Intel and AMD.Initial support for Intel Nova Lake, Diamond Rapids, and Panther Lake, along with compatibility for AMD Zen 5 and Zen 6, ensures that many next-generation processors will be able to boot Linux from day one, without having to wait for specific patches.

The ARM64, RISC-V, and LoongArch architectures also receive improvementsThis reinforces the role of the kernel in mobile phones, ultralight laptops, development boards, and embedded devices. In RISC-V, for example, extensions such as Zicfiss and Zicfilp are incorporated, which expand capabilities and optimize performance in certain flow control and interrupt handling operations.

In the area of ​​SoCs and peripherals, Linux 7.0 adds support for new chips from Qualcomm and MediaTekas well as improvements to the Google Tensor's USB PHYs and USB-C support in Apple Silicon, making power, data, and video management more reliable in modern laptops and tablets.

Also included are some interesting compatibility improvements.The kernel now recognizes Rock Band 4 Bluetooth guitars for PS4 and PS5, as well as the Logitech K980 solar keyboard. These aren't critical devices, but they demonstrate the Linux community's obsession with making "everything work," from mission-critical servers to niche peripherals.

For platforms like Rockchip RK3588 and RK3576Common in mini PCs and boards like the Orange Pi or Radxa, hardware video decoding (H.264/H.265) is integrated directly into the main kernel. This allows for smooth 4K playback without relying on manufacturer-modified kernels, improving both the multimedia experience and long-term maintenance.

Graphics, HDR, and gaming: AMD, Intel, and a touch of nostalgia

In terms of graphics, Linux 7.0 continues its effort to catch up with the latest GPUs.without forgetting older hardware that can still put up a fight.

For AMD, the AMDGPU driver adds support for new hardware blocksincluding GFX 12.1 (targeting a generation beyond RDNA 4) and GFX 11.5.4, a refreshed variant of RDNA 3.5. The AMDKFD compute driver is updated in tandem, allowing these GPUs to be used for both gaming and compute-intensive tasks.

At Intel, the Xe driver continues to matureWith advancements in SR-IOV (GPU-level virtualization) and shared virtual memory across multiple devices, display support for Nova Lake processors is now fully functional—a vital detail considering the arrival of Ubuntu 26.04 LTS and new laptops that will use these CPUs.

  Complete guide to installing DEB and RPM packages without complications

For fans of backward compatibility, the work already started with kernel 6.19 is noticeable.The modern amdgpu driver is now available for older AMD GPUs with GCN 1.0 and 1.1 architectures, which were previously limited to outdated drivers without Vulkan support. Now, many older systems can run current software with significantly improved stability and performance, extending their lifespan.

HDR support continues to advanceThe 6.19 kernel laid the groundwork for decent high dynamic range in Linux, an area that has historically been a headache. Linux 7.0 continues to refine integration with Intel, AMD, and Nvidia drivers, aiming for a modern visual experience on compatible monitors—key for content creators and demanding gamers.

Small details that enhance quality of life complete this section.More temperature sensors available, better compatibility with laptops from brands like ASUS, HP and Lenovo (fans, backlighting, sensors), improvements for GPUs like Intel Battlemage and small steps forward in projects like Nouveau and PowerVR.

NPUs, local AI and the future of accelerated computing

A less visible, but very powerful, change is the arrival of a revamped computational acceleration subsystem. which allows the kernel to talk directly to the NPU (neural processing unit) without strange intermediaries.

This direct integration with NPUs opens the door to running AI workloads locally. With much greater efficiency: according to the shared tests, certain processes can consume up to 80% less battery power when run by the NPU instead of the CPU. This means more applications will be able to perform AI inference on your own device, without relying on the cloud.

In addition to the improved battery life, there is a direct impact on privacy.If your team can process AI models locally, less sensitive data needs to be sent to external servers. For tasks like speech recognition, image classification, smart filters, or local assistants, this is a significant leap forward.

Linux 7.0, with its focus on NPUs and emerging hardwareThis anticipates a scenario where almost every laptop or desktop will include some kind of dedicated AI accelerator. By having this support in the core kernel, distributions and applications can rely on a common foundation, instead of depending on proprietary solutions or manufacturer-specific patches.

How to install Linux 7.0 and whether it's worth upgrading now

If you're curious to try the new kernel, you have several options.But it's best to proceed with caution. The community itself warns that, although version 7.0 is released as stable, it's a very large update and it's normal for minor bugs to appear as more people use it.

The easiest way in Debian/Ubuntu based distributions To use the Mainline Kernel Installer tool, add the PPA and update it. You'll see a list of available kernels in a graphical interface, select Linux 7.0, and install it with a couple of clicks. Upon restarting, from GRUB You'll be able to boot with the new kernel and check with uname -r that everything has gone well.

For advanced users, there's always the classic option of compiling the kernel by hand.Download the linux-7.0.tar.gz file from the official website, extract it, and configure the options with make menuconfig and install the necessary build dependencies. Then, launch the process with make -j $ (nproc)Followed by make modules_install y make installIf there are no errors, after the reboot you will be able to boot with the new kernel compiled to your specifications. If you need a practical guide, see how download, configure and compile the kernel.

However, the general recommendation for most users is to be cautious.If your system is running smoothly and you don't need a specific improvement from 7.0 (for example, support for a very new chip or some critical feature), it's best to wait until your distribution officially adopts it once the first maintenance patches arrive.

Rolling release distributions like Arch Linux or Fedora They tend to incorporate new kernel versions quickly, so if you use one of them, you'll likely have Linux 7.0 available shortly after its release. In more conservative distributions, such as Ubuntu LTS or Debian Stable, Ubuntu 26.04 LTS and future versions will include 7.0 as the default kernel.

In any case, it's always a good idea to note down the current kernel version. (With uname -rand make sure the boot manager allows you to select older kernels. That way, if Linux 7.0 gives you any unexpected problems on your specific hardware, you can easily revert to the previous version without any issues.

Overall, Linux 7.0 marks the beginning of a new era For the kernel: the commitment to Rust is solidified, key file systems and subsystems are refined, the groundwork is laid for Wi-Fi 8, NPUs, and next-generation CPUs, and security is strengthened with post-quantum cryptography and finer controls over I/O and permissions. Whether you decide to install it now or prefer to wait for your distro to deliver it automatically, this is a release that will set the tone for the Linux ecosystem for years to come.

linux 7.0
Related articles:
Linux 7.0: What's new in the kernel and a generational leap