Microsoft DirectStorage 1.4 boosts game loading with new compression tech

Microsoft's DirectStorage 1.4 update introduces Zstandard compression and a new asset conditioning library to cut load times and smooth in-game streaming on modern GPUs and SSDs.

Microsoft is rolling out DirectStorage 1.4 with a focus on accelerating game asset streaming by modernizing how data moves from storage to the GPU. DirectStorage was originally built into DirectX 12 to tap into the speed of NVMe SSDs, and its development has removed the CPU as the middle-man in the process of streaming assets from storage to the GPU. Traditionally, this has been done over CPU, causing delays and latencies across teh stack, so the shift enables powerful consumer GPUs to receive data fast enough to keep pace with demanding visuals.

The new release centers on a compression and decompression technology called Zstandard (Zstd), which should improve game loading times and bring much faster game asset streaming than what was previously used. By integrating Zstd directly into the DirectStorage pipeline, Microsoft is aiming to reduce both the time it takes to pull assets from storage and the overhead involved in preparing those assets for the GPU, which in turn targets shorter loading screens and fewer in-game hitches.

To push Zstd even further, Microsoft has developed the Game Asset Conditioning Library (GACL), a companion tool that developers run on their assets before a game ships. The idea is that instead of simply compressing textures, GACL first conditions them to be more compressible, allowing Zstd to squeeze files down by up to 50% more than it otherwise could. Shuffling rearranges data inside texture files so repeating patterns cluster together, giving Zstd more to latch onto. Block-Level Entropy Reduction (BLER) and Component-Level Entropy Reduction (CLER) then reduce texture complexity at the block and color-channel level, using perceptual quality as a guide so any changes remain invisible to the player. CLER takes this a step further by incorporating machine learning to identify exactly where that reduction can be applied without anyone noticing.

52

Impact Score

Hades variant affects 23 PyPI package versions

The Mini Shai-Hulud Hades variant is targeting PyPI packages tied to bioinformatics and Artificial Intelligence themes. Socket researchers say the malware uses Python startup hooks and compiled extensions to run a JavaScript stealer.

DiffusionGemma rethinks text generation with diffusion

DiffusionGemma applies diffusion-style denoising to text, trading autoregressive token-by-token decoding for iterative canvas refinement. Its design combines encoder guidance, bidirectional denoising, scheduling, and entropy-based sampling.

NVIDIA shows RTX Spark platform at Computex 2026

NVIDIA presented RTX Spark in Taipei as a Windows on Arm platform spanning gaming, creator, and Artificial Intelligence workloads. Microsoft also detailed Windows 11 optimizations built specifically for the new NVIDIA silicon.

AWS enterprise processor targets Artificial Intelligence inference

AWS’s Annapurna Labs-designed enterprise server processor uses a chiplet architecture for cloud infrastructure and Artificial Intelligence inferencing. The design combines Arm compute resources, cache coherency, and high-bandwidth interconnects for AWS deployments.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.