Language Selection

English French German Italian Portuguese Spanish

System76 Thelio Mega is a quad-GPU Linux desktop powered by Ryzen Threadripper

Filed under
Linux

System76 began its life as a Linux computer seller only. Essentially, the company would sell re-branded laptops with Ubuntu pre-installed. To provide a class-leading experience, however, System76 also provided top-notch customer service, helping Linux beginners get started with a little hand-holding when needed. This focus on service continues today, and it is largely responsible for the company's success and longevity.

Seeking to better control its own destiny, the company branched out from only being a computer-seller and transformed into a maker too. It's handcrafted Thelio desktops are powerful works of art, comprised of wood, metal, and good ol' fashioned American elbow grease. Yes, these Thelio machines are made in the USA -- Colorado, specifically.

System76 has even created its own operating system -- the Ubuntu-based Pop!_OS, which has been very well received by the Linux community. This Linux distro will work on most computers -- not just Sytem76 machines. If you want vanilla Ubuntu, don't worry -- the company continues to offer that OS as an option when buying one of its computers.

Read more

System76 Launches The Thelio Mega With Threadripper + Four GPUs

  • System76 Launches The Thelio Mega With Threadripper + Four GPUs

    The System76 Thelio Major with AMD Ryzen Threadripper 3900 series is already a beast, but now this Linux PC vendor has managed to outdo themselves once again with the Thelio Mega.

    The System76 Thelio Mega is what the Colorado company calls "the world's smallest quad-GPU deep learning system". The Thelio Mega pairs the latest-generation AMD Ryzen Threadripper platform with up to four NVIDIA Quadro RTX GPUs to make for an incredibly powerful system.

  • System76 unleash a quad-GPU Linux monster with the 'Thelio Mega'

    What has four GPUs and could probably heat up your entire house? System76 have announced the Thelio Mega and it's absolutely ridiculous.

    System76 say that the Thelio Mega is the world's smallest quad-GPU workstation primarily made for deep learning and scientific computing. I'm sure that won't stop someone trying to play games with it though as this might be (probably is) the most powerful Linux desktop you can buy and it's gorgeous too.

Thelio Mega: The World’s Smallest Quad-GPU Deep Learning System

  • Thelio Mega: The World’s Smallest Quad-GPU Deep Learning System

    Most quad-GPU workstations on the market right now use a generic chassis. It works if you want something just for storing components, but what you get is a machine that runs hot and slows down your system. That’s where Thelio Mega comes in.

    We’ve engineered Thelio Mega to ensure your top-line components perform to their fullest potential. Its thermals are actually two separate systems, as we found it more effective to divide and conquer. Unique airflows keep heat generated by the CPU and GPU from mixing, exhaust air effectively, and prevent throttling.
    Heat builds up quickly when you have up to 4 NVIDIA Quadro RTX GPUs stacked atop one another. Rather than use a single vent to cool the entire system, Thelio Mega uses intake fans on the bottom and side panels to blow cool air directly onto your GPUs. Everything down to the GPU brace has been tested for maximum thermal efficiency. The result is consistent access to every last core of performance in your system.

RTX 2080Ti vs RTX 3090 Machine Learning Benchmarks

  • RTX 2080Ti vs RTX 3090 Machine Learning Benchmarks

    NVIDIA’s 2nd generation RTX architecture brings more performance for faster Machine Learning training. We tested four Geforce RTX 2080Ti GPUs against three Geforce RTX 3090 GPUs and found that three RTX 3090s performed similar or better than four RTX 2080Ti’s for most tests with the same batch size. The RTX 3090s offer faster training with larger batch sizes as well, thanks to the additional memory available in the RTX 3090. Three RTX 3090s were used, rather than four, due to their increased power requirements.

    The tests were conducted on the new Thelio Mega workstation from System76. Thelio Mega was engineered specifically for graphics compute intensive workloads.

System76 Launches Quad-GPU Linux Workstation

  • System76 Launches Quad-GPU Linux Workstation

    Denver-based System76 has launched the powerful new Thelio Mega, calling it the world’s smallest quad-GPU Linux workstation.

    The Thelio Mega workstation, which is priced starting at US $7,499, incorporates advanced technologies and high-performance components that make it ideal for deep learning and scientific computing.

Open-source Linux PC Stuffs 4 Quadros Into Mid-Tower Case

  • Open-source Linux PC Stuffs 4 Quadros Into Mid-Tower Case | Tom's Hardware

    Are you looking for a new PC build project that could cost as much as a new car? Do you have four RTX 3090s, or even Nvidia Quadros, kicking around with a Threadripper CPU? Or are you just able to spend up to $45,000 on someone who does building a deep-learning PC for you? Then System 76’s new Thelio Mega, which packs corporate power into a form factor that would be at home in your well, home, is for you.

    System 76 is a bit of an oddity in the extreme high-end PC world. The company is a staunch advocate of open-source and the right to repair, meaning that all of their computers are actually free for you to build yourself, if you can understand their freely downloadable schematics and have all the right parts and tools. They also all pack Linux as their OS, because of course that’s how System 76 rolls. But if you don’t have the confidence or knowhow to build one of their systems from just its schematics, or just can’t get your hand on parts like RTX 3000 series cards, you can also pay the company to build your new system for you.

Behind the Scenes of Thelio Mega Engineering

  • Behind the Scenes of Thelio Mega Engineering

    In 2017, we announced that we were going to bring the design and manufacturing of our products in-house. The driving purpose was to leverage our understanding of our users’ needs in order to engineer better products for them. In 2018, we shipped the first fruits of our labor and, over the course of the last two years, shipped hundreds of updates to the Thelio line as we continuously integrated improvements into our manufacturing. Establishing our factory in Colorado made this possible, but we were just getting started.

    Early this year, we set off to engineer our workstation version of a Le Mans Hypercar. It started with a challenge: Engineer a quad-GPU workstation that doesn’t thermal throttle any of the GPUs. Three GPUs is pretty easy. Stack the forth one in there and it’s a completely different animal. Months of work and thousands of engineering hours later we accomplished our goal. Every detail was scrutinized. Every part is of the highest quality. And new factory capabilities, like milling, enabled us to introduce unique solutions to design challenges. The result is Thelio Mega. A compact, high-performance quad-GPU system that’s quiet enough to sit on your desk.

    We started with simple fan placement experiments to determine the best location for intake and exhaust fans and sizes. Computer fluid dynamics simulations assisted to dial in air flow ducts followed by hundreds of fan placement iterations and thermal tests.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

More in Tux Machines

Assign Actions To Touchpad Gestures On Linux With Touchegg

The application runs in the background, transforming the multi-touch gestures you make on your touchpad into various desktop actions. For example, you can minimize a window by swiping down using 3 fingers, pinch in using 2 fingers to zoom in, etc. This is a demo video recorded by the Touchegg developer (image above credits also go to the dev). Read more

Meet DevTerm: An Open Source Portable Linux Terminal For Developers

You may be familiar with Clockwork company, which earlier launched an open-source Linux-powered portable game console called GameShell for gamers. Now, they’re back with another new portable and modular device called DevTerm for developers, which you can easily carry along wherever you go. Read more

Android Leftovers

today's leftovers

  • GPUOpen Software Updated For The Radeon RX 6000 Series - Phoronix

    AMD has updated their collection of software offered under their "GPUOpen" umbrella for Radeon RX 6000 series / RDNA 2 compatibility. The Radeon GPU Profiler, Radeon Memory Visualizer, and other software packages offered via GPUOpen have been updated with "Big Navi" RDNA2 support.

  • OctopusWAF: A Customizable Open-Source WAF for High Performance Applications

    Mainstream web application firewalls (WAFs) can be very difficult to understand, with thousands of lines of code and obscure plugins. This complexity makes it challenging for developers to modify code to block specific anomalies and secure their applications. But OctopusWAF is different - the open-source WAF is customizable, user-friendly and optimized for a large number of parallel connections - making it ideal for high performance Asynchronous JavaScript and XML (AJAX) applications.

  • ZLUDA: Drop-In Open-Source CUDA Support For Intel Xe / UHD Graphics

    An interesting solution built off Intel's oneAPI Level Zero is the open-source "ZLUDA" that is providing a "Level Zero CUDA" implementation for being able to run programs geared for NVIDIA CUDA atop Intel UHD / Xe Graphics hardware. ZLUDA is a project independent of NVIDIA and Intel but one of the most interesting external projects we have seen so far targeting Intel's Level Zero interface. ZLUDA allows for unmodified CUDA applications to run on Intel GPUs with "near native" performance through this alternative libcuda running with Skylake / Gen9 graphics and newer.

  • Portwell and Congatec spin Elkhart Lake modules in multiple form factors

    Portwell unveiled a “PQ7-M109” Qseven module with Intel’s Atom x-6000. Congatec recently announced x6000 modules in Qseven (Conga-QA7), SMARC, (Conga-SA7), Mini Type 10 (Conga-MA7), and Compact Type 6 (Conga-TCA7) form factors. Portwell has announced the PQ7-M109, its first product based on Intel’s 10nm fabricated Elkhart Lake family of low-power system-on-chips, which includes several Atom x-6000, Celeron, and Pentium models. In September, in reporting on Congatec’s Elkhart Lake based Conga-PA7 Pico-ITX SBC, we promised to cover Congatec’s four Elkhart Lake compute modules in a separate report. Well, better late than ever: We briefly summarize Congatec’s Conga-QA7 (Qseven), Conga-SA7 (SMARC), and Conga-MA7 (COM Express Mini Type 10) and Conga-TCA7 (Compact Type-6) modules farther below.

  • Kubernetes and SUSE Enterprise Storage 7 - SUSE Communities

    Rook is a CNCF – the Cloud Native Compute Foundation (CNCF) hosts Kubernetes and related open source projects – graduated project which automates the installation, deployment and upgrade of Ceph. It takes care to launch and configure all Ceph components correctly, setup Ceph on storage devices and allows Kubernetes applications to use Ceph as storage – for block, file, and object storage. Deployment with Rook is like many other Kubernetes installation, you install Rook using a helm chart that you can configure, and then Kubernetes will do all the necessary steps to setup Ceph. You can also connect to the Ceph dashboard and see how your applications use storage. Once Rook is up, your containerized applications can use Ceph as persistent storage using the usual Kubernetes APIs like PersistentVolumeClaims (PVCs). Running Ceph with Rook on Kubernetes means that you have a smaller footprint overall instead of setting up a separate Ceph cluster and a Kubernetes cluster. Kubernetes will run applications and storage together in the same infrastructure. This is not advised for very large storage installations but a great option for a Kubernetes cluster that needs a smaller storage configuration. Depending on your use-cases and requirements, you can use dedicated storage nodes in your single cluster – and have dedicated application nodes – or use all your nodes for storage and applications.

  • Digest of YaST Development Sprint 113 | YaST

    Time flies and it has been already two weeks since our previous development report. On these special days, we keep being the YaST + Cockpit Team and we have news on both fronts. So let’s do a quick recap. Cockpit Modules Our Cockpit module to manage wicked keeps improving. Apart from several small enhancements, the module has now better error reporting and correctly manages those asynchronous operations that wicked takes some time to perform. In addition, we have improved the integration with a default Cockpit installation, ensuring the new module replaces the default network one (which relies on Network Manager) if both are installed. In the following days we will release RPM packages and a separate blog post to definitely present Cockpit Wicked to the world. On the other hand, we also have news about our Cockpit module to manage transactional updates. We are creating some early functional prototypes of the user interface to be used as a base for future development and discussions. You can check the details and several screenshots at the following pull requests: request#3, request#5.

  • Stantinko Botnet Now Targeting Linux Servers to Hide Behind Proxies [Ed: They say almost nothing about the fact that you actually need to sabotage your GNU/Linux setup and have malware installed on it for this to become a risk. Microsoft propaganda at ZDNet set off this "Linux" FUD.]

    According to a new analysis published by Intezer today and shared with The Hacker News, the trojan masquerades as HTTPd, a commonly used program on Linux servers, and is a new version of the malware belonging to a threat actor tracked as Stantinko.