Tag Archives: II

Linux’s Slowdown, Raptor’s Talos II, Clear Linux & Other Dominating Topics This Month


Another month is in the books as we approach the 15th birthday of Phoronix in June. This month on Phoronix there was a lot of interest in the major Linux kernel slowdown that ended up being attributed to “STIBP” but fortunately improvements are on the way, a lot of interest in Raptor Computing Systems’ Talos II libre system and their lower-cost Blackbird board, excitement starting to heat up around Linux 4.21 changes, and the Mesa 18.3 release being on approach while Mesa 19.0 feature development is in full swing.

This month on Phoronix were 24 featured articles/reviews and 293 original Linux/open-source news articles all written by your’s truly. It’s slightly under the average of 10 original news articles per day and one featured article per day due to some of the larger articles this month requiring more testing and my wife being hospitalized with some issues the better part of a week that complicated some benchmarking. In addition, this month also marked the debut of Phoronix Test Suite 8.4 as the latest quarterly release of our flagship open-source, cross-platform benchmarking software.

If you enjoy the punctual and interesting content on Phoronix each and every day, please consider showing your support by joining Phoronix Premium this holiday season or a tip via PayPal if you would like to show your support. At the very least please do not view this site with any ad-blocker as the advertisements are the primary means for allowing this site to continue in its 15th year and for the continued open-source development around the Phoronix Test Suite, Phoromatic, OpenBenchmarking.org, etc.

With that out of the way, here’s a look at the most popular featured articles for November:

Bisected: The Unfortunate Reason Linux 4.20 Is Running Slower
After running a lot of tests and then bisecting the Linux 4.20 kernel merge window, the reason for the significant slowdowns in the Linux 4.20 kernel for many real-world workloads is now known…

The Spectre/Meltdown Performance Impact On Linux 4.20, Decimating Benchmarks With New STIBP Overhead
As outlined yesterday, significant slowdowns with the Linux 4.20 kernel turned out to be due to the addition of the kernel-side bits for STIBP (Single Thread Indirect Branch Predictors) for cross-HyperThread Spectre Variant Two mitigation. This has incurred significant performance penalties with the STIBP support in its current state with Linux 4.20 Git and is enabled by default at least for Intel systems with up-to-date microcode. Here are some follow-up benchmarks looking at the performance hit with the Linux 4.20 development kernel as well as the overall Spectre and Meltdown mitigation impact on this latest version of the Linux kernel.

Raptor Talos II POWER9 Benchmarks Against AMD Threadripper & Intel Core i9
For those curious about the performance of IBM’s POWER9 processors against the likes of today’s AMD Threadripper and Intel Core i9 HEDT processors, here are some interesting benchmarks as we begin looking closer at the POWER9 performance on the fully open-source Raptor Talos II Secure Workstation. This open-source, secure system arrived for Linux testing with dual 22-core POWER9 CPUs to yield 176 total threads of power.

Benchmarking The $25 Raspberry Pi 3 Model A+ Performance
Released last week was the Raspberry Pi 3 Model A+ as their latest ARM SBC coming in at the $25 USD price point and their last board release before doing a redesign. I was able to snag a Raspberry Pi 3 Model A+ for $25 with availability appearing to be better than some of the past Raspberry Pi releases. Here are some initial benchmarks of the RPi 3 Model A Plus compared to a few other ARM boards.

20-Way AMD / NVIDIA Linux Gaming Benchmarks For The 2018 Holidays
If you are hoping to pick-up a new graphics card during the upcoming holiday sales, here is a 20-way NVIDIA GeForce and AMD Radeon Linux gaming benchmark comparison using a wide assortment of GPUs while using the very newest graphics drivers and a variety of OpenGL/Vulkan titles.

16-Way Graphics Card Comparison With Valve’s Steam Play For Windows Games On Linux
While Steam Play is still of beta quality on Linux for running Windows games on Linux using their Wine-based Proton compatibility layer, Steam Play has been fast maturing since it was rolled out to the public in late August. The game list continues growing and with regular updates to Steam Play / Proton / DXVK (Direct3D 10/11 over Vulkan), more games are going online for running on Linux and doing so with decent performance and correct rendering. Given the most recent Steam Play beta update vastly improving the experience in our tests, here are the first of our Steam Play Proton benchmarks with Ubuntu Linux and using sixteen different NVIDIA GeForce / AMD Radeon graphics cards.

A Look At The Open-Source Talos II POWER9 Performance Against x86_64 Server CPUs
In the benchmarks earlier this month looking at the Talos II POWER9 dual 22-core performance its performance was compared to various AMD Threadripper and Intel Core i9 CPUs. They were used as comparison points since all of those CPUs sport four memory channels, including the Sforza POWER9 CPUs, while IBM caters the larger LaGrange/Monza POWER9 modules with eight memory channels as competition to Xeon and EPYC. But for those wondering how the POWER9 Sforza performance compares to Intel Xeon and AMD EPYC processors, here are some benchmarks.

Void Linux, Solus, Manjaro, Antergos, Sabayon & Clear Linux Put To A Performance Battle
Given last week’s new images release of the rolling-release, systemd-free, original-creation Void Linux I decided to take it for a spin with some fresh benchmarking as it had been two years or so since last trying out that Linux distribution with its XBPS packaging system. For seeing how the performance compares, I benchmarked it against some of the other primarily enthusiast/rolling-release/performant Linux distributions including Antergos, Clear Linux, Debian Buster Testing, Fedora Workstation 29, Manjaro 18.0, Sabayon Linux, Solus, and Ubuntu 18.10.

Initial Benchmarks Of OpenBSD 6.4, DragonFlyBSD 5.3, FreeBSD vs. Linux
Given the recent release of OpenBSD 6.4, FreeBSD 12 now being in beta, and DragonFlyBSD 5.3 evolving nicely for what will eventually ship as DragonFlyBSD 5.4, here is the start of some fresh benchmarks between the BSDs and a few Linux distributions to see how the performance compares as we approach the end of 2018.

A Look At The GCC 9 Performance On Intel Skylake Against GCC 8, LLVM Clang 7/8
With GCC 9 embarking upon its third stage of development where the focus ships to working on bug/regression fixes in preparation for releasing the GCC 9.1 stable compiler likely around the end of Q1’2019, here is a fresh look at the GCC 9 performance with its latest development code as of this week compared to GCC 8.2.0 stable while using an Intel Core i9 7980XE test system running Ubuntu Linux. For good measure are also fresh results from LLVM Clang 7.0 stable as well as LLVM Clang 8.0 SVN for the latest development state of that competing C/C++ open-source compiler.

And the most popular news article for the month:

NVIDIA RTX, AMD On Linux & Distro Performance Dominated Linux Discussions In October
During the month of October on Phoronix there were 330 original news stories and 26 featured articles / Linux hardware reviews penned by your’s truly.

Raspberry Pi 3 Model A+ Released For $25 USD
The Raspberry Pi Foundation today unveiled the Raspberry Pi 3 Model A+ ARM SBC that costs just $25 and offers Bluetooth, dual-band 802.11ac WiFi, and a 1.4GHz Cortex-A53 processor.

Apple’s New Hardware With The T2 Security Chip Will Currently Block Linux From Booting
Apple’s MacBook Pro laptops have become increasingly unfriendly with Linux in recent years while their Mac Mini computers have generally continued working out okay with most Linux distributions due to not having to worry about multiple GPUs, keyboards/touchpads, and other Apple hardware that often proves problematic with the Linux kernel. But now with the latest Mac Mini systems employing Apple’s T2 security chip, they took are likely to crush any Linux dreams.

Patches Revived For A Zstd-Compressed Linux Kernel While Dropping LZMA & BZIP2
For more than a year it’s been talked about adding an option to support Zstd-compressed Linux kernel images while it looks like that Facebook-backed high performance compression algorithm for kernel images could soon finally be mainlined.

Linus Torvalds Comments On STIBP & He’s Not Happy – STIBP Default Will End Up Changing
It turns out that Linus Torvalds himself was even taken by surprise with the performance hit we’ve outlined on Linux 4.20 as a result of STIBP “Single Thread Indirect Branch Predictors” introduction as well as back-porting already to stable series for cross-hyperthread Spectre V2 protection. He doesn’t want this enabled in full by default.

An Open Letter To Solus From Its Founder Ikey Doherty
Solus, the promising Linux distribution started back in 2015 by Ikey Doherty that led to the creation of its own “Budgie” desktop, has been without its founder since this summer. While the circumstances under his decision to fade away from the project aren’t clear, he is well and has shared this message to relay with the community.

Some Linux Users Are Reporting Software RAID Issues With ASRock Motherboards
Making the rounds this morning is an ASRock forum post about a motherboard accidentally and repeatedly wiping out Linux Software RAID meta-data. A few Phoronix readers have also reported similar issues such as in the forums and Twitter. This appears to stem from an UEFI issue.

What Do You Want To See Out Of The Redesigned, Next-Gen Raspberry Pi?
With the launch this week of the Raspberry Pi 3 Model A+, they made it clear whatever comes next will be a big re-design compared to all of the Raspberry Pi ARM single board computers up to this point. So what would you hope they incorporate into the next-generation of these low-cost boards?

The Linux 4.20/5.0 Kernel Is The Biggest All Year With 354+ Thousand Lines Of New Code
The Linux kernel will be ending 2018 on a high note with the current merge window for what will be called either Linux 4.20 or Linux 5.0 is the biggest kernel update by lines of code in more than one year.

Debian Packages To Eliminate Vendor-Specific Patches, Affecting Downstreams Like Ubuntu
Debian packages have supported the concept of vendor-specific patches whereby when DPKG unpacks a source package on different operating systems / distributions (such as Debian vs. Ubuntu), different patches could be selectively applied. Ubuntu is one of the main benefactors of this feature while effective immediately these vendor-specific patches to source packages will be treated as a bug and will be unpermitted following the Debian 10 “Buster” release.

Here’s to a very interesting December ahead with a lot of interesting benchmarks and other content planned.

How to Deploy Hyperledger Fabric on Kubernetes – Part II | Linux.com

We recently hosted a webinar about deploying Hyperledger Fabric on Kubernetes. It was taught by Alejandro (Sasha) Vicente Grabovetsky and Nicola Paoli from AID:Tech.

The webinar contained a detailed, step-by-step instruction showing exactly how to deploy Hyperledger Fabric on Kubernetes. For those who prefer reading to watching, we have prepared a condensed transcript with screenshots that will take you through the process that has been adapted to recent updates in the Helm charts for the Orderers and Peers.

Are you ready? Let’s dive in!

What we will build

First, we will deploy a Fabric Certificate Authority (CA) serviced by a PostgreSQL database for managing identities.

Then, we will deploy an ordering service of several Fabric ordering nodes communicating and establishing consensus over an Apache Kafka cluster. The Fabric Ordering service provides consensus for development (solo) and production (Kafka) networks.

Finally, we will deploy several Peers and connect them with a channel. We will bind them to a CouchDB database.

Read more at Hyperledger

A Look At The Open-Source Talos II POWER9 Performance Against x86_64 Server CPUs

In the benchmarks earlier this month looking at the Talos II POWER9 dual 22-core performance its performance was compared to various AMD Threadripper and Intel Core i9 CPUs. They were used as comparison points since all of those CPUs sport four memory channels, including the Sforza POWER9 CPUs, while IBM caters the larger LaGrange/Monza POWER9 modules with eight memory channels as competition to Xeon and EPYC. But for those wondering how the POWER9 Sforza performance compares to Intel Xeon and AMD EPYC processors, here are some benchmarks.

With Raptor Computing Systems having kindly sent over this Talos II Secure Workstation a few weeks ago, it’s been a delight testing this box that is fully open-source down to the motherboard firmware and BMC. More background information on the Talos II can be seen in the earlier article.

Pricing on the Talos II motherboard is $2499 USD and the 22-core POWER9 CPUs cost $2625 a piece, which does make it more expensive than all but one (the dual EPYC 7601) of the Intel/AMD server configurations tested in this article. I tested the available Intel Xeon Scalable platforms I had, which were the Xeon Silver 4108 and Xeon Gold 6138 with the Tyan GT24E-B7106. With the TYAN Transport SX TN70A-B8026 I tested the available EPYC CPUs of the 7251, 7351P, 7401P, 7551, and 7601 SKUs. There was also the Dell PowerEdge R7425 with dual EPYC 7601 CPUs.

The EPYC 7601 CPU carries a retail price of around $4,379 USD while the dual 22-core POWER9 CPUs come in at over $5,200 plus the $2k+ motherboard, but with SMT4 that configuration yields 176 threads. The Xeon Gold 6138 processors are of similar price to the POWER9 22-core CPUs at about $2679 USD. Granted, of the systems tested, only the Talos II was open-source down to the motherboard firmware and BMC.

Ubuntu 18.10 was running on all of these systems with the Linux 4.19.1 kernel and GCC 8.2 code compiler for these basic reference benchmarks between POWER9 and Xeon/EPYC.

Total War: WARHAMMER II Now Available For Linux Gamers, Powered By Vulkan


Feral Interactive just lit up the Linux build of Total War: WARHAMMER II on Steam.

Total War: WARHAMMER II, the latest in the Total War franchise from Creative Assembly, is now natively available for Linux as well as macOS. Pricing is $59.99 USD via the Feral Store or Steam.

The game debuted for Windows in September of 2017. The Linux port is exclusively powered using the Vulkan graphics API. The Linux system requirements put the GTX 680 or R9 285 as a minimum while they recommend a GeForce GTX 970 or better or Radeon RX 480. The Linux Vulkan driver requirements aren’t stiff either at Mesa 18.1.5+ or NVIDIA 396+, but obviously you may find better performance with the newer drivers.

Feral didn’t seed early review access to this game for Linux testing, but will be running benchmarks starting today. I know the game though is automated benchmark friendly, so hopefully by tomorrow morning I will have a large AMD/NVIDIA Linux graphics card comparison for this game. Stay tuned for GPU/driver benchmarks for this latest AAA Linux game release!