AMD News Archives

Last week I wrote about the AMD Secure Processor support for Linux 5.6 being queued as part of the cryptography subsystyem work with supporting the PSP / Secure Processor of Raven Ridge APUs. That AMD Secure Processor support is now rounded out with the Trusted Execution Environment (TEE) driver being queued for wiring into that subsystem.

While we have found recent Linux kernels paired with latest motherboard BIOS releases to work out generally well for recent AMD APUs, not everyone has been having a trouble-free experience on recent kernels. But an affected user has discovered a possible workaround if hitting stability issues.

With the recently launched Threadripper 3960X / 3970X processors there was a workaround needed to boot them on Linux until an AMD MCE driver issue was resolved. That patch was upstreamed last week into the Linux 5.5 development kernel while now is getting ready to make its debut into supported Linux stable release branches.

With the Wraith Prism heatsink fan included with many modern AMD Ryzen processors there is configurable RGB lighting, which unfortunately AMD hadn't publicly documented or offered a Linux utility for manipulating the RGBs under Linux. Fortunately, there is now a straight-forward solution for dealing with those Wraith Prism RGB LEDs thanks to the open-source and independent CM-RGB project.

Last month I wrote about AMD working on TEE driver support to load "trusted applications" onto the AMD Secure Processor under Linux. That work is now queued for introduction with Linux 5.6 and wired through for Raven Ridge APUs.

For those looking for some fresh reference numbers on the impact of using GCC's Profile Guided Optimizations (PGO), here are some benchmark runs looking at the GCC 10 PGO performance on an Ubuntu 19.10 workstation built around the Ryzen Threadripper 3960X.

A month ago I posted benchmarks looking at the performance of Linux 4.16 through Linux 5.4 kernels using an Intel Core i9 workstation. Stemming from that was a request for an AMD EPYC kernel comparison, so I carried out said tests. Due to the Rome support being newer, this round of testing is looking at the EPYC 7642 performance on Linux 5.0 to Linux 5.4.

The AMD machine check exception (MCE) code fix for Linux has landed ahead of this weekend's anticipated 5.5-rc3 release. This AMD MCE fix allows for the AMD Ryzen Threadripper 3900 series processors introduced last month to boot the Linux kernel without hangs or other workarounds.

New AMD Family 17h (Zen) CPU family microcode was merged today into linux-firmware.git.

The AMD Ryzen Threadripper 3960X/3970X are incredibly fast and trounce the competition, but as noted on launch-day most (all?) Linux distributions have a boot issue with them over a machine check exception. There is an easy workaround to let these core-happy CPUs boot and run Linux while the proper fix was queued last week in ras/core in what looks like it will wait until Linux 5.6 for merging.

Released on Friday was a new version of AMD's GPU Performance API "GPUPerfAPI" project under the GPUOpen umbrella. This is the AMD library used by CodeXL, Radeon Compute Profiler, and others for tapping GPU performance counters and to help in analyzing performance/execution characteristics for Radeon hardware. But this new GPUPerfAPI 3.5 release comes with a rather surprising change.

AMD today announced the Ryzen V1000 and Ryzen R1000 series embedded processors intended for mini PCs and other small form factor PCs.

With the IOMMU updates for the Linux 5.5 kernel there is a major rework to the AMD IOMMU driver to make use of more common DMA IOMMU code for implementing the DMA API but with an admitted risk of potential new regressions.

With the AMD Ryzen Threadripper 3970X benchmarks on Windows 10 and Linux, Ubuntu 19.10 and other common distributions were just ~2% faster than the Microsoft OS and Clear Linux was just ~10% faster, based on 80+ benchmarks carried out. Those margins are much closer than we have seen with past iterations of Threadripper, but is that due to the Zen 2 microarchitecture and the improved topology of the new Threadripper CPUs or due to Microsoft's scheduler changes and other software improvements made in Windows 10 November 2019 Update? Here are some benchmarks.

A few weeks back AMD published a TEE "Trusted Execution Environment" driver for APUs on Linux for utilizing the controversial AMD Secure Processor.

As outlined in our AMD Ryzen Threadripper 3960X / 3970X Linux review, these new Zen 2 Threadripper processors are phenomenal processors that offer significant uplift over earlier Threadripper CPUs and easily dominate over Intel's Core i9 HEDT competition. But there is one big issue right now with the Linux support: on Ubuntu and the like, it doesn't boot without a workaround. Here's that workaround for easy future reference.

Announced earlier this month, the Athlon 3000G is shipping today a week ahead of the Ryzen Threadripper 3000 series. The Athlon 3000G is AMD's new sub-$50 processor for lightweight desktop purposes.

When writing about the Intel Jump Conditional Code (JCC) Erratum and how Intel is working to mitigate the performance hit of the CPU microcode update with patches to the GNU Assembler, there was some concern expressed by readers that it might hurt AMD performance. That does not appear to be the case.

While the processors are not shipping until later this month, AMD is today making known their next batch of CPUs being released.

GNU Binutils with its "Gas" assembler has now added the rest of the instructions supported by the AMD Zen 2 microarchitecture that previously were unsupported by this piece of the GNU toolchain.

While we have seen a lot of open-source AMD Linux graphics driver patches for Renoir and that initial support within the 5.4 kernel, support for this 2020 APU platform is still maturing. The newest work on the Linux upbringing for Renoir is enabling the "DMCUB" support.

Complementing the big AMDGPU feature pull request from two weeks ago, on Friday AMD sent out a second batch of features targeting the upcoming Linux 5.5 kernel merge window.

Just earlier this month NVIDIA announced their funding of the Blender Foundation at the flagship "patron" level and now AMD has followed them in backing this foundation for assisting the development of this leading 3D creation software.

Updated AMD Family 17h (Zen / Zen 2) CPU microcode has quietly landed within the linux-firmware Git tree.

Back in February for LLVM Clang 9.0 was the initial AMD Zen 2 "znver2" enablement, but like the GCC support at the time it was the very basics. With time GCC picked up Zen 2 scheduler improvements and other work while sadly in the case of LLVM the improvements are still pending.

RDPRU is one of the new instruction set extensions of AMD "Zen 2" CPUs that is for reading a processor register that is typically limited to privilege level zero. RDPRU allows for reading select registers at any privilege level. With Linux 5.5, the RDPRU presence will be advertised by the CPU features.

AMD today lifted the lid on the Radeon RX 5500 series as their first Navi 14 based graphics card. This is the soft launch with no units shipping yet but expected to starting in November.

Just one week after the release of Radeon Open eCosystem 2.8, AMD has now released ROCm 2.9 as the newest feature release for this open-source GPU Linux compute stack for Radeon hardware.

AMD provided an update on their Linux FreeSync/Adaptive-Sync support at this week's X.Org Developers Conference event in Montreal. There's good news both for HDMI and Wayland Linux users with Radeon graphics.

While the Linux 5.4 cycle just officially began last week and its feature merge window not even over until this weekend, given there are AMD EPYC load balancing improvements and many other kernel improvements in general, I was eager to fire up the in-development kernel on the EPYC 7002 "Rome" series to see how the performance is looking.

While the Ryzen 9 3950X and 3rd Gen Ryzen Threadripper processors were reportedly on track for launching in October with updates as of a few weeks ago, today AMD announced a slight delay in launching these new processors.

While just yesterday the big DRM feature pull was sent in for Linux 5.4, AMD has also volleyed out their initial batch of fixes for this next version of the kernel.

From Rome, Italy this afternoon AMD not only announced more than 100 world records have been broken with their new EPYC "Rome" processors, but there is also a new SKU! Meet the EPYC 7H12.

One of the early pull requests for the just-opened Linux 5.4 kernel merge window is the hardware monitoring "hwmon" subsystem changes. This time around the notable change is CPU temperature reporting for Ryzen 3000 series processors.

In addition to AMD's open-source Linux driver developers being busy in recent weeks bringing up the Renoir APU support, today we've seen the first baby steps towards bringing up "Dali" as another upcoming AMD APU.

Continuing on from last week's testing that found the AMD Ryzen 9 3900X + ASUS CROSSHAIR VIII HERO WiFi consuming more power on Linux compared to Windows 10, here are some additional metrics after spending a good deal of time over the weekend on further tests.

There has been a lot of talk recently of AMD Ryzen 3000 series processors reportedly not hitting their boost clock frequencies, whether stock coolers are adequate for hitting the boost frequencies, and other concerns around the boost behavior on these new Zen 2 processors. AMD issued a statement today they are rolling out a new BIOS/firmware update to help with boost clock frequency optimizations.

Adding to the growing list of features for Linux 5.4 with its cycle officially kicking off in mid-September is a kernel scheduler optimization designed to improve load balancing on AMD EPYC servers.

In addition to the new hardware support and other features queued already in DRM-Next for the upcoming Linux 5.4 merge window, on Friday AMD sent in a final pull request to DRM-Next of new material ahead of this upcoming kernel cycle.

Back on the AMD EPYC 7002 "Rome" launch day I wrote about how AMD is working to return to open-source BIOS / Coreboot support and now there's further confirmation of their work in that direction.

Frequently brought up following our various Ryzen 3000 "Zen 2" benchmarks like the Ryzen 9 3900X vs. Core i9 9900K gaming benchmarks is how the Ryzen 9 3900X is pulling considerably more power than the similarly equipped Intel Core i9 system and those numbers are higher than what is often cited by Windows reviewers as the difference. I've begun investigating that power difference and indeed quite quickly could see Linux power usage being higher than Windows 10.

AMD developers have sent out their latest open-source Linux patches doing their kernel driver share for enabling High-bandwidth Digital Content Protection (HDCP) support for version 1.4 and newer.

Two weeks have passed since AMDVLK 2019.Q3.4 while out this morning is the next iteration of this open-source AMD Radeon Vulkan driver derived from the company's official cross-platform driver code-base.

Just prior to yesterday's Linux 5.3-rc6 kernel release, a change was pulled into the code-base that disables the advertising of RdRand support on older AMD CPUs/APUs.

While AMD's next-gen Renoir APUs are Vega-based and not Navi, beyond the initial Linux driver enablement seen over the past few weeks coming out a few days ago were a set of patches just getting the power management in order.

Among many different Linux/open-source benchmarks being worked on for the AMD EPYC "Rome" processors now that our initial launch benchmarks are out of the way are Linux distribution comparisons, checking out the BSD compatibility, and more. Some tests I wrapped up this weekend were seeing how recent Linux kernel releases perform on the AMD EPYC 7742 64-core / 128-thread processors.

Not directly related to the recent AMD Zen 2 BIOS update needed to fix an RdRand problem (though somewhat related in that the original systemd bug report for faulty AMD RdRand stems from these earlier CPUs), but AMD has now decided to no longer advertise RdRand support for Family 15h (Bulldozer) and Family 16h (Jaguar) processors under Linux.

One month ago we were told that AMD released a BIOS fix to their motherboard partners for addressing the systemd boot issue with Ryzen 3000 series processors that stems from an RdRand instruction issue. Finally over the past week we've seen motherboard vendors pushing out BIOS updates for the prominent motherboards and indeed this takes care of the issue.

AMD is striking well over the past month with their Linux hardware bring-up. In the past month we've seen the Navi 10/12/14 support get in order for Linux as well as support for the future Vega-based Arcturus GPU and now we see the initial enablement patches for their next-generation APUs, Renoir.

If you didn't have a chance since last night to check out our benchmarks of the AMD EPYC 7742 and EPYC 7502 Linux performance, I certainly encourage you to do so. Even if you aren't a server enthusiast, it's incredible to see the engineering achievement of AMD with Zen 2 and how the race is certainly back on in the CPU space. If you are short on time, here's the quick summary of our initial AMD EPYC 7002 benchmark results.