The Ultimate Do-It-Yourself Linux Box
Never underestimate the importance of a good power supply for your do-it-yourself system, especially if you intend to use two video cards configured in SLI mode. You can experience all kinds of bizarre symptoms of instability if you underpower your system with an inadequate power supply. Don't go for anything less than a 500-Watt power supply if you intend to use two video cards in SLI mode. Go even higher if you intend to add other PCI cards to your system. And, when you shop for power supplies, be careful to look for efficiency ratings. Some power supplies boast good peak output, but the sustained output can still be inadequate.
There's almost no point in choosing a power supply based on its rated mean time between failures (MTBF), that is, how long it should last. We suspect these figures reflect how long the power supply lasts assuming the fan never fails. Unfortunately, power supply fans fail all the time. The power supply overheats, and kablooey, so much for the mean time between failure rating. Your mileage may vary, but we've had the best luck with Enermax power supplies and their fans.
You will need a power supply with a 24-pin power connector to the motherboard for any of the motherboards we tested. You also will need a power supply with connectors for two video cards, so that you can use these motherboards in SLI mode. We chose the Enermax ELT500AWT 500-Watt power supply for our system. It sells for about $100 US, depending on your source. We also used a Silverstone SST-ST65ZF 650-Watt power supply for a similarly configured system. It sells for about $170 US, depending on your source. We've had these power supplies only for a few weeks at the time of this writing, but they both work well so far—knock on wood.
We pulled out all the stops when it came to a case for our Ultimate Do-It-Yourself System. We chose the Silverstone TJ07-S case, which sells for about $365 US, depending on your source. This is quite expensive for a case, but it is worth the investment. First, the thing is huge. It's larger than any tower case we've ever tried. This gives you tons of room to work when you insert cards and cable the system. On the other hand, if you're looking for a case you can place on top of the desk instead of beside it, this is definitely not the case for you.
The hard drives are tucked away in two separate removable compartments, and each compartment is cooled with its own separate 120mm fan. It has two more 120mm fans at the top of the case, and two rear 92mm intake fans. The case is remarkably quiet, considering it has six fans, not including the CPU fan, power supply fans and so forth. All this ventilation keeps everything very cool without having to invest the time and effort in creating a liquid-cooled system.
The case has a flip-down front accessory panel with connectors for audio, USB and FireWire. The panel is flush with the front of the case, so you simply press on the bottom of the panel to open it. Some people might be annoyed that there's no spring loading, no button and no catch for the panel, either in its open or closed state. We have no complaints with it though.
The only way to press the reset switch is to use a wire tool, which you insert into a small hole in the front of the case. Some people will hate this feature, others will appreciate how it protects you from accidentally resetting your system.
You can certainly find adequate cases for far less money, and some of them may even place things like the accessory jacks in more convenient locations. But, we found this to be a superb case primarily because of how easy it is to work inside it (thanks to its gigantic size) and superior ventilation without having to use liquid cooling.
If you're going to go the budget route, there are so many decent cases from which to choose, we're hard pressed to recommend one over another. Each has its advantages and disadvantages. We chose the Thermaltake Tsunami VA3000BWA, which sells for just over $100 US at most outlets. It's not as high class as the Silverstone, and we found it frustrating to get the DVD drive installed, but it's a fair case for the price. It has a flip-up top accessory panel. People who leave things on top of their computer case will find this inconvenient, but the same must be said of the Silverstone case, as it has top-mounted fans you can block by leaving manuals or other paraphernalia on top of the computer.
Yes, you can use two NVIDIA cards in SLI mode on Linux—if you don't mind running the proprietary NVIDIA driver, which taints the Linux kernel. We chose a pair of eVGA GeForce 7900GT cards with 256MB of RAM. These cards are a great compromise between price and performance. The combination of cards totals at about $500 US, which is less than the price of a single NVIDIA 7900 GTX card (about $570 US). eVGA sells basically the same cards with different clock speeds at different prices. For example, budget-minded folks can get a single eVGA GeForce 7900GT Signature 256MB (same basic model as the ones we used in SLI mode) with higher clock speeds for about $360 US.
You obviously have more choices than eVGA when it comes to video card manufacturers. We chose eVGA for our examples simply because the company produces a large selection of prices and configurations of NVIDIA cards, which made it easy to pick cards to fit varying budgets. We've had good success with other brands as well.
If you opt to use the NVIDIA proprietary drivers, you need to add the following line to your xorg.conf file:
Option "SLI" "Auto"
We also recommend that you dig through the NVIDIA HOWTO to learn how to specify whether you're using a digital or analog connector. Some monitors like to guess which interface you're using for five seconds or so, which can cause annoying delays when you start your desktop.
|Designing Electronics with Linux||May 22, 2013|
|Dynamic DNS—an Object Lesson in Problem Solving||May 21, 2013|
|Using Salt Stack and Vagrant for Drupal Development||May 20, 2013|
|Making Linux and Android Get Along (It's Not as Hard as It Sounds)||May 16, 2013|
|Drupal Is a Framework: Why Everyone Needs to Understand This||May 15, 2013|
|Home, My Backup Data Center||May 13, 2013|
- Designing Electronics with Linux
- Making Linux and Android Get Along (It's Not as Hard as It Sounds)
- Dynamic DNS—an Object Lesson in Problem Solving
- New Products
- Using Salt Stack and Vagrant for Drupal Development
- Validate an E-Mail Address with PHP, the Right Way
- Build a Skype Server for Your Home Phone System
- A Topic for Discussion - Open Source Feature-Richness?
- Tech Tip: Really Simple HTTP Server with Python
- Why Python?
- Not free anymore
2 hours 1 min ago
5 hours 49 min ago
- Reply to comment | Linux Journal
5 hours 57 min ago
- Understanding the Linux Kernel
8 hours 11 min ago
10 hours 41 min ago
- Kernel Problem
20 hours 44 min ago
- BASH script to log IPs on public web server
1 day 1 hour ago
1 day 4 hours ago
- Reply to comment | Linux Journal
1 day 5 hours ago
- All the articles you talked
1 day 7 hours ago
Enter to Win an Adafruit Pi Cobbler Breakout Kit for Raspberry Pi
It's Raspberry Pi month at Linux Journal. Each week in May, Adafruit will be giving away a Pi-related prize to a lucky, randomly drawn LJ reader. Winners will be announced weekly.
Fill out the fields below to enter to win this week's prize-- a Pi Cobbler Breakout Kit for Raspberry Pi.
Congratulations to our winners so far:
- 5-8-13, Pi Starter Pack: Jack Davis
- 5-15-13, Pi Model B 512MB RAM: Patrick Dunn
- 5-21-13, Prototyping Pi Plate Kit: Philip Kirby
- Next winner announced on 5-27-13!
Free Webinar: Hadoop
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers
Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.
Some of key questions to be discussed are:
- What is the “typical” Hadoop cluster and what should be installed on the different machine types?
- Why should you consider the typical workload patterns when making your hardware decisions?
- Are all microservers created equal for Hadoop deployments?
- How do I plan for expansion if I require more compute, memory, storage or networking?