Cloud Linux Reviews & Articles

There have been 45 Linux hardware reviews and benchmark articles on Phoronix for cloud. Separately, check out our news section for related product news.

Yesterday AWS announced the public preview of Amazon Linux 2027 as their successor to the current Amazon Linux 2023 in-house distribution. With all the software upgrades and improvements like building packages now with link-time optimizations (LTO), enforcing x86-64-v3 builds, and other changes, I was eager to run some performance benchmarks of Amazon Linux 2023 vs. Amazon Linux 2027 in the EC2 cloud.

After originally announcing Graviton5 last December, recently AWS finally made the M9g and M9gd instances generally available as the first featuring these new in-house ARM server processors for the EC2 cloud. Graviton5 makes use of Arm Neoverse-V3 cores compared to Neoverse-V2 with Graviton4, support up to 192 cores, and feature a higher 3.3GHz clock speed compared to 2.8GHz on the prior-generation Graviton CPUs. Here is an initial look at how the Graviton5 processor performs over Graviton4.

Google Cloud recently launched their N4A series powered by their in-house Axion ARM64 processors. In that launch-day benchmarking last month was looking at how the N4A with Axion compared to their prior-generation ARM64 VMs powered by Ampere Altra. There were dramatic generational gains, but how does the N4A stand up to the AMD EPYC and Intel Xeon instances? Here are some follow-up benchmarks I had done to explore the N4A performance against the Intel Xeon N4 and AMD EPYC N4D series.

AMD Secure Encrypted Virtualization with Secure Nested Paging (SEV-SNP) provides memory encryption and integrity protections that can be especially useful in modern cloud computing. Typically a 2~10% performance overhead is reported when engaging AMD SEV-SNP for these hardware-backed security protections. In this article is an extensive look at the current AMD SEV-SNP performance impact for confidential computing on modern EPYC servers. The current Ubuntu 24.04 LTS was tested as well as an Ubuntu 26.04 development snapshot in evaluating the latest optimizations and what is on the horizon this year for AMD EPYC Linux server performance.

Back in 2024 Google rolled out their Axion in-house ARM processors with the Google Cloud C4A instance type. Today they are expanding their Axion offerings in Google Cloud with the N4A instances now out of preview. The Google Cloud N4A instances are designed for scale-out web servers and microservices, containerized applications, back-end application services, databases, data analytics, and cost-effective development/staging/testing environments.

Last week the Microsoft Azure HBv5 instances reached general availability as powered by the custom EPYC 9V64H CPUs with HBM3 memory. These very interesting EPYC processors for memory bandwidth intensive workloads were announced last year while have finally reached GA with jaw-dropping results for software able to take advantage of the 6.7 TB/s memory bandwidth thanks to the HBM memory. The Azure HBv5 benchmarks last week showed how they compare to prior generation HBv4 instances while this article is taking things further and putting the performance into perspective against the older HBv2 and HBv3 instances.

Google Cloud today is rolling out their N4D compute instances that are optimized for cost/price-performance and geared for general purpose workloads. The N4D instances are powered by 5th Gen AMD EPYC "Turin" processors and offer very nice performance and value over their prior-generation general purpose VMs.

Nearly one year ago Microsoft announced the HBv5 virtual machines powered by a custom-designed AMD 4th Gen EPYC processor with high bandwidth memory (HBM3). Finally today the Azure HBv5 series is reaching general availability for those with memory-intensive HPC applications and other workloads. Microsoft kindly provided Phoronix with HBv5 access in advance to begin testing these new VMs with the AMD EPYC 9V64H CPUs featuring HBM memory, so here are some of the first independent benchmarks of these exciting processors powering Azure's new HPC VM instances.

With Amazon recently launching their M8a AWS instances powered by 5th Gen AMD EPYC "Turin", for their M8 class instance types there now are all the latest-generation CPU options with AMD EPYC Turin (M8a), Intel Xeon 6 Granite Rapids (M8i), and their in-house Graviton4 processors (M8g). After recently looking at the M7a vs. M8a performance with Amazon EC2, many Phoronix readers expressed interest in seeing an M8a vs. M8i vs. M8g performance showdown so here are those benchmarks.

Oracle recently launched their E6 compute shape for Oracle Cloud and powered by AMD EPYC 9005 "Turin" processors. Oracle Cloud Infrastructure also launched their Compute Cloud@Customer X11 and Private Cloud Appliance X11 platforms that are all powered by the E6 compute shape with 5th Gen AMD EPYC. For those curious about the performance and value of the Oracle Cloud E6 shape compared to prior-gen E5 as well as alternatives from Amazon AWS, Microsoft Azure, and Google Cloud, these benchmarks are geared for you.

Last week Amazon/AWS announced the new EC2 M8a instances as their latest-generation, general-purpose compute instances now powered by AMD EPYC 9005 "Turin" processors. Amazon announced the M8a as having up to 30% higher performance and up to 19% better price performance over M7a. With my testing of both at 32 vCPUs, the new AMD EPYC Turin instance provided 1.59x the performance over the prior-generation EPYC Genoa instance!

Last week alongside announcing the AMD Instinct MI350X/MI355X and the ROCm 7.0 software preview, AMD also introduced the AMD Developer Cloud as a new means for developers to easy try out Instinct accelerators with their own software and with the ROCm compute stack already setup. Having tried out prior AMD cloud compute environments, as soon as my email invite for the AMD Developer Cloud arrived I decided to give it a try.

Back in April at Google Cloud Next was the introduction of the new C4D family of VMs powered by AMD EPYC 9005 "Turin" processors. Back on launch day I looked at the C3D vs. C4D performance at some of the smaller, more common VM sizes. In today's article is a look at the top-end performance of the C4D family with 384 vCPUs. For those wondering about the compute potential of the c4d-standard-384, here are some benchmarks of this 192-core / 384-thread EPYC Turin configuration compared to the prior C3D AMD EPYC Genoa based instance that topped out at 360 vCPUs.

As part of the announcements coming out today from Google Cloud Next 2025, the embargo has now lifted on the new Google Cloud C4D VMs. Powered by AMD EPYC 9005 "Turin" processors, the new C4D instances deliver incredibly high performance and can scale up to 384 vCPUs with 3TB of RAM. For web servers, databases, CPU-based machine learning, and other workloads, the new Google C4D instances deliver incredible uplift compared to the prior-gen C3D instances. Here are some of the first public, independent benchmarks of Google's new C4D family.

After talking about AmpereOne for the Oracle Cloud Infrastructure since last year, earlier this month Oracle finally announced general availability on their OCI Ampere A2 instances for tapping into these latest-generation Ampere Computing AArch64 server processors in the public cloud. Here's a brief look at the OCI Ampere A2 performance with AmpereOne compared to their prior A1 instances powered by Ampere Altra.

Scaleway by way of their Scaleway Labs group recently launched the Elastic Metal RV1 (EM-RV1) as the world's first RISC-V servers available in the cloud. These RISC-V cloud servers are built around the T-Head 1520 SoC and are an interesting way to explore the RISC-V architecture and/or otherwise make use of RISC-V for CI/CD deployments or other testing purposes. In this article are some benchmarks showing the RISC-V EM-RV1 performance against Intel and AMD x86_64 Linux.

Some time ago I ran through a number of benchmarks of Google Cloud's C3D VMs powered by AMD EPYC Genoa processors. The AMD EPYC 9004 series showed terrific performance with strong generational improvements over the Intel Xeon Scalable processors. Following that a request came in to examine the PingCAP TiDB database performance given its growing popularity. In this article we'll review those benchmarks showing how GCE C3D delivers strong performance advantages for TiDB.

Back in August Google Cloud announced the C3D instances powered by AMD EPYC 9004 "Genoa" processors while only last week was C3D promoted to general availability. Curious about the performance of C3D after being impressed by AMD EPYC Genoa bare-metal server performance at Phoronix as well as what I've seen with Genoa in the cloud at Microsoft Azure and Amazon EC2 / AWS, here are some benchmarks of the new C3D up against other GCE instances.

While back in November was when AWS originally announced new EC2 instances powered by 4th Gen AMD EPYC "Genoa" processors, only this week did they bring their M7a general purpose instances to a general availability state where anyone can access them. Being very impressed with 4th Gen EPYC bare metal as well as with Azure's HPC cloud, I fired up some benchmarks of the new Genoa-powered EC2 M7a instance compared to the new M7i instances powered by Intel Xeon Scalable "Sapphire Rapids" as well as showing how the competition is to Amazon's in-house Graviton ARM-based server processors.

The past several weeks at Phoronix has been very exciting with benchmarking AMD EPYC Genoa-X processors (along with Bergamo) and the incredible uplift delivered by these latest AMD server processors. But for those not yet having the opportunity to test the new EPYC Genoa-X processors locally, those wishing to evaluate the Genoa-X capabilities in the public cloud prior to making an investment in these high-end server processors with 3D V-Cache, or those simply preferring the ease of cloud infrastructure, Azure's new HBv4 series provide an excellent route for leveraging AMD Genoa-X compute capabilities in the cloud. Here are benchmarks of the new Azure HBv4 powered by EPYC Genoa-X compared to prior Azure HPC VMs. The Azure HBv4 performance is outstanding with incredible generational uplift and leading value among Microsoft's HPC-focused VMs.

While waiting for AMD 4th Gen EPYC "Genoa" instances to become available via the major public cloud providers, I was curious to see how existing AMD EPYC Milan instances compare to Intel's new Sapphire Rapids instances in public preview on Google Cloud. While expecting some friendly competition, at the same vCPU size EPYC Milan was managing to deliver not only better performance-per-dollar but also even better raw performance in numerous workloads against the Google Cloud C3 Sapphire Rapids.

Back in October Google announced their Compute Engine C3 instances in private preview that featured 4th Gen Xeon Scalable "Sapphire Rapids" processors as well as making use of Google's custom Intel Infrastructure Processing Unit (IPU). Since then, back in January, was the big Sapphire Rapids launch with the likes of the Xeon Platinum 8490H being released. Last month meanwhile Google promoted the C3 VMs to public preview state. The Sapphire Rapids C3 VMs remain in "public preview" from Google Cloud during which time there are no charges involved for the CPU costs. For those wondering about the core-for-core performance of Sapphire Rapids in Google Cloud, here are my initial benchmarks of the C3 series.

Last week at Intel's Innovation conference the Intel Developer Cloud "DevCloud" was announced, while on the AMD side there is already something similar: the AMD Cloud Platform. At the tail end of 2021, AMD announced the Accelerator Cloud as a way for trying out the latest EPYC CPUs and Instinct accelerators complete with a pre-configured ROCm compute software stack. The AMD Cloud Platform is a currently parallel effort to the Accelerator Cloud with the former intended more for developers while the latter is more customer-oriented. After trying out the AMD Cloud Platform, it's indeed an easy way to evaluate the latest AMD data center wares while having a easy-to-deploy, pre-configured software environment.

Over the summer Google announced Tau T2A as the first Google Cloud Compute Engine VM to run on Arm. The T2A series is powered by Ampere Altra processors and complement the Tau T2D series powered by AMD EPYC 7003 "Milan" processors that launched last year. I have been trying out the Tau T2A series for the past several weeks and have some initial benchmarks to share today for showing how the Ampere Altra Arm VMs can perform against the existing T2D series.

Today the AMD EPYC 7003 Milan-X processors are officially shipping. See my AMD EPYC 7773X Linux review for more details and plenty of benchmarks. The 768MB of L3 cache per CPU won't be of benefit to all workloads, just as the forthcoming Ryzen 7 5800X3D is focused on gaming. Aside from the dozens of benchmarks covered in my review, if you are still left wondering about whether other workloads stand to benefit from Milan-X, thankfully it's easy to already test drive it in the cloud with Microsoft Azure. Here are some Microsoft Azure HBv3 benchmarks looking at the Milan-X uplift in the cloud.

Last year Amazon launched the EC2 M6a instances powered by AMD EPYC 7003 series while this week they have expanded their range of AMD Zen 3 offerings by launching the EC2 C6a series. The EC2 C6a instances are designed for compute-intensive workloads (hence the "C" series) and AWS is promoting it as offering up to 15% improvement in price-performance over prior-generation C5a instances and up to 10% lower cost than comparable x86-based EC2 instances. I've run some benchmarks of the new EC2 C6a instances looking at how they perform over the prior 2nd Gen EPYC C5a based instances, against the Intel Ice Lake competition over in the M6i stack, and also how the C6a competes with Amazon's own Graviton2-based C6g type.

Announced earlier this year for Google Cloud was a new family of virtual machines called Tau VMs. The initial T2D instances are powered by AMD EPYC 7003 "Milan" processors to deliver leading performance and are also positioned to deliver great value in going up against the likes of Amazon's Graviton2 instances. Tau VM instances are now available as a preview and Google has provided us with gratis access to the new instance types for benchmarking.

Earlier this week Amazon introduced Intel Xeon Scalable 3rd Gen "Ice Lake" powered EC2 cloud instances and marks their first x86-based sixth-generation offerings that follow their "M6g" Graviton2 instances launched last year. Curious about the "M6i" Ice Lake performance with AWS, here are a number of benchmarks looking at the performance and value of the new M6i instances compared to former Intel M5 instances as well as Amazon's own M6g Graviton2 instances.

One of the exciting elements of last month's AMD EPYC 7003 "Milan" series launch was having same-day availability in public clouds. Microsoft as one of AMD's cloud partners worked closely to deliver launch-day availability in their public cloud using EPYC 7003 series processors with the new "HBv3" instances focused on high performance computing (HPC) virtual machines. Here are some benchmarks of the Azure HPv3 instances compared to prior generation Microsoft Azure HPC instances available on-demand in their cloud.

Complementing Amazon's recently launched EPYC 7002 "Rome" CPUs in the EC2 cloud, the "c5a" series has now been extended with the "c5ad" line-up of AMD EPYC Rome processors that now have local NVMe-based solid-state storage directly attached. Initial tests of the Amazon EC2 C5ad instances are promising and indeed offering better value than the comparable Intel Xeon instances.