What's Up With Error Correcting Memory on AM5 in 2024?

Summary

In this detailed guide, the speaker discusses AMD's launch of its AM5 platform, exploring its evolution from server-grade EPYC CPUs to consumer-focused Ryzen chips. Focusing on the state of ECC (Error-Correcting Code) memory in 2024, the speaker highlights the challenges and improvements in AM5, comparing it to the previous AM4 platform. Issues like BIOS bugs, the complexities of DDR5, and compatibility with error-correction are addressed, while also providing recommendations for various motherboards, emphasizing reliability for home server setups. The video offers insights into AMD's progress in ECC support, user experiences, and practical advice for configuring a stable and powerful system.

Takeaways

• 😀 AMD has launched EPYC CPUs for AM5, a move that repurposes high-performance desktop processors like the 7950X for server use.
• 😀 The shift of desktop chips like the 7950X to EPYC branding reflects the growing demand for powerful, small-form-factor server solutions.
• 😀 DDR5 memory, especially error-correcting DDR5 (ECC), is now far more reliable in 2024 compared to its early struggles in 2023.
• 😀 The primary challenges with ECC on AM5 in 2024 stem from BIOS bugs and incomplete implementation, though things are improving over time.
• 😀 AM5 platforms support unregistered ECC memory (not to be confused with server-grade registered ECC), which offers some error correction capabilities.
• 😀 DDR5 memory's error correction is better supported for server use, but for desktop platforms, issues like DIMM slot compatibility and kernel recognition remain problematic.
• 😀 While unregistered ECC memory works on AM5, there's still a lack of clear error reporting, leading to some uncertainty in system stability diagnostics.
• 😀 AM5 motherboards, particularly from brands like ASRock and ASUS, have made substantial improvements in error correction functionality, but some issues persist with specific models.
• 😀 AMD's EPYC platform offers superior error detection and correction compared to desktop AM5 setups, with more advanced management tools for system administrators.
• 😀 Reliability in AM5 ECC setups can be improved by using two DIMMs per channel, avoiding the instability of using all available DIMM slots at high speeds.

Q & A

What is the significance of AMD's EPYC CPUs for AM5?

AMD's EPYC CPUs for AM5 are a crucial development, offering server-class CPUs with up to 16 cores in a compact form factor. These processors are suitable for small business or home server use, and the availability of these CPUs on AM5 platforms means that error-correcting memory (ECC) support is much improved, making them an excellent choice for those needing reliable performance.

What challenges did error-correcting memory (ECC) face on AM5 platforms in 2023?

In 2023, ECC support on AM5 platforms had issues, particularly with improper error logging and inconsistent functionality. However, AMD's efforts to improve this with firmware updates (such as AGESA 1.0.0.5c) and motherboard updates by manufacturers like ASRock and Asus have helped resolve many of these problems by 2024.

How does error-correcting memory differ between DDR4 and DDR5 platforms?

DDR5 introduced changes to error-correcting memory, particularly through the transition from registered to unregistered ECC DIMMs. Registered ECC memory is designed for server use, while unregistered ECC in DDR5 is designed for desktop platforms. DDR5's error correction is less mature compared to DDR4, with some challenges still present in error reporting and handling.

What were the main issues with the B650 motherboard's ECC functionality?

The B650 motherboard faced issues where it would post with error-correcting memory but not properly report error correction events. Initially, it would not log ECC errors to the IPMI (Intelligent Platform Management Interface), although this was eventually corrected with BIOS updates. However, there were still occasional issues with error reporting in the kernel.

What is the impact of BIOS updates on ECC functionality for AM5 systems?

BIOS updates are crucial for improving ECC functionality on AM5 systems. Updates often fix bugs that prevent error correction from being properly reported or used, as seen with various motherboards like the ASRock B650 and Asus X670. Manufacturers are constantly refining their BIOS to ensure better support for ECC memory as the platform matures.

How does ECC error reporting work on Epic CPUs compared to AM5?

Epic CPUs feature advanced error detection and reporting facilities, which allow detailed error logs that specify the exact location of errors in the system. In contrast, AM5 platforms initially struggled with reporting such detailed error logs, with issues often requiring manual intervention to log errors in operating systems like Linux.

What is the recommended configuration for memory in AM5 systems to ensure stability?

For stability and reliability, it's recommended to use two DIMMs per channel on AM5 systems. Using four DIMMs per channel, particularly at high data rates, introduces electrical challenges that can affect performance and stability. Keeping memory configurations to two DIMMs helps avoid noise and instability.

Why is it important to use ECC memory for home server or small business server setups?

ECC memory is critical for home or small business servers as it helps prevent data corruption by automatically correcting errors in memory. This is particularly important in environments where uptime and data integrity are crucial, and ECC ensures that the system can detect and correct memory errors without crashing.

How does the ASRock Taichi motherboard perform in terms of ECC support?

The ASRock Taichi motherboard has been praised for its good ECC support, with error correction working as expected after firmware updates. It was one of the first to provide stable ECC functionality on AM5 platforms, making it a reliable choice for users who require error correction in their systems.

What are the advantages of using Epic CPUs on AM5 platforms for home server setups?

Using Epic CPUs on AM5 platforms for home server setups offers several advantages, including powerful multi-core performance, improved ECC support, and cost-effective server-class processors. These CPUs are ideal for users who need to run reliable servers with robust error correction in a compact and energy-efficient form factor.