Article Thumbnail

Most Reliable PC Hardware of 2012

Written on December 31, 2012 by Matt Bach
Share:

Introduction

At Puget Systems, we record a huge amount of data for each and every system we sell; including benchmarks, BIOS screenshots, thermal images, and system photos. In fact, much of this data is published on our website and can be accessed through our part information pages. Simply view a part information page, scroll down to the Links section, and you can view BIOS Screenshots, Thermal Images and Product Photography from recent systems we built. This is still in early development, but we love being transparent so we are always looking for meaningful ways to share any data we have.

While this data is nice, one of the most important things we track is the failure rates of individual components. Reliability is of our primary values, so this data is invaluable for tracking both individual component, product line, and overall brand failure rates. With 2012 coming to a close, we thought we would run some reports and share what hardware we found to be the most reliable in 2012.

Since we are a custom computer company and do at times special order in components to meet a customer's specific needs, there is one stipulation we are imposing on what hardware we will be allowing into this list. Specifically, we are only considering high volume items from our product line so that we have a large enough sample size to make an informed call on the reliability of the component.

With that said, let's take a look at one of the most at risk components in a computer: the motherboard.

Motherboard

A reliable motherboard is an absolute must in a high quality computer. Not only is a motherboard very difficult to swap out, but the effects of a poor quality motherboard can be far reaching and difficult to troubleshoot. This is complicated by the fact that motherboards are one of the most complex components in a computer. You have SATA, USB, fan, and wireless controllers as well as physical ports, audio chips, and everything else that is needed to inter-connect every component in your system. This is a huge number of small parts that have to work perfectly together, and any one of these could potentially have a problem. If there is a single dead USB port, slight static over the audio, or the voltage levels are measured outside of norm, it does not meet our standards and is considered to have failed. Because of this, motherboards have the overall highest failure rate of any core component with 1 out of every 25 motherboards failing for one reason or another.

At Puget Systems, every motherboard must pass an extensive qualification process but there is no substitute for reviewing hard evidence after offering the product for a period of time. We have to simply keep an eye on our failure reports and quickly move on any trends we may see. From these failure reports, we found three motherboards that had absolutely zero failures in 2012.


Interestingly, two of these boards are from Asus' Workstation class of motherboards. Asus states that their Workstation series "...bring[s] you ultimate reliability and quality through our 24x24 initiative, which means 24-hour non-stop operation and a 24-month life cycle supply guarantee...", which we can completely confirm with our reliability data. In fact, these are the only WS boards that we sold a significant number of in the past year, so the Asus Workstation series has a perfect record for 2012. 

CPU

We could make a list of all the different CPU models that are extremely reliable, but that would be a list of almost every CPU we sold in 2012. So instead, we are simply going to say this: every CPU made in 2012 is highly reliable. For CPUs as a whole, we saw an extremely low .47% failure rate, and that includes failures due to overclocking that are more our fault than the manufacturer's. So just know that when picking a CPU, you don't need to worry about the risk of failure if you are using it in a standard computer system.

 

RAM

At Puget Systems, we almost exclusively use Kingston RAM in our computers. We do so because we know from experience that they are extremely reliable. Any time we tried other prominent brands (when Kingston was either in shortage or did not offer what we wanted), we almost always ended up moving back to Kingston once we were able. This is primarily due to the fact that as a whole, Kingston is as much as three to four times more reliable than other brands. Because of this, we have very low RAM failure rates so this section is going to include quite a bit of different models. 

Desktop DDR3 RAM Failure Rate 
Kingston DDR3-1600 4GB (KVR16N11/4) .25%
Kingston DDR3-1600 8GB (KVR16N11/8) .386%


Standard desktop RAM as a whole has an low failure rate (roughly 1%), but interestingly we did not have any models that had a 0% failure rate. However, out of all the desktop RAM we sold in 2012 there were two models that had a much lower failure rate than the other models. Although this is not as good as a 0% failure rate, a .25% failure rate means that you would have to go through 400 sticks to find just a single bad stick, which is still very impressive.

Server/Workstation DDR3 RAM Failure Rate
Kingston DDR3-1600 4GB ECC (KVR16E11/4) 0%
Kingston DDR3-1600 4GB ECC Reg. (KVR16R11D8/4I) 0%
Kingston DDR3-1600 8GB ECC (KVR16E11/8) 0%
Kingston DDR3-1600 8GB ECC Reg. (KVR16R11D4/8I) 0%
Kingston DDR3-1333 4GB ECC (KVR1333D3E9S/4G) .29%
Kingston DDR3-1333 4GB ECC Reg. (KVR1333D3S4R9S/4GI) 0%
Kingston DDR3-1333 8GB ECC (KVR1333D3E9S/8G) 0%
Kingston DDR3-1333 8GB ECC Reg. (KVR1333D3D4R9SK3/24GI) 0%


ECC RAM is specifically designed to have a low failure rate, so it is no surprise that we had quite a few models that had absolutely no failures in 2012. In fact, the only ECC or ECC Reg. RAM that had a failure was the Kingston DDR3-1333 4GB ECC which had only a single stick fail. This means that overall for 2012, ECC RAM had a failure rate of just .06%

Laptop DDR3 SODIMM RAM  Failure Rate
Kingston SODIMM DDR3-1600 8GB (KVR16S11/8) 0%
Kingston SODIMM DDR3-1333 8GB (KVR1333D3S9/8G) 0%


For laptop SODIMM RAM, there were two models that we found to be highly reliable. Interestingly, both are 8GB models and each has a 0% failure rate.

Hard Drives

Just like RAM, we typically only use a few brands that we historically know to be extremely reliable. With the rising popularity of SSD drives, we decided that this section would be best divided into two sections separating SSD and traditional platter drives.

Solid-State Drives - 0% Failure
Intel 320 80GB (SSDSA2CW080G310)
Intel 320 300GB (SSDSA2CW300G310)
Intel 330 120GB (SSDSC2CT120A3K5)
Intel 335 240GB (SSDSC2CT240A4K5)
Intel 510 120GB (SSDSC2MH120A2K5)


Most of these SSDs are mid-range models, which historically have an overall lower failure rate than higher-end models. Still, we have to give an honorable mention to the Intel 520 120GB, Intel 520 180GB, Intel 520 240GB and the Intel 520 480GB as each had only a single drive fail in 2012. Even discounting those drives, SSDs as a whole are extremely reliable with a low .36% failure rate overall in 2012.

Traditional Platter Drives - 0% Failure
Western Digital Caviar Green 3TB (WD30EZRX)
Western Digital Scorpio Blue 1TB (WD10JPVT)
Western Digital RE 1TB (WD1003FBYX)
Western Digital Caviar Black 2TB (WD2002FAEX)
Western Digital VelociRaptor 1TB (WD1000DHTX)


Traditional platter drives are at a higher risk of failure than SSDs (1.11% versus .36%), but there were still a number of drives with no failures in 2012. Interestingly, of these drives, none were from the same product line. So in this case, it appears that we cannot say that one line is more reliable than another in terms of failure rates (reliability for things like RAID arrays is an entirely different matter). Honorable mention to the Western Digital Caviar Green 2TB and the Western Digital Scorpio Black 750GB as each had one drive fail in 2012.

Video Cards

Video cards are one of the most heat-sensitive components in computers today, partially from the fact that unlike CPUs it is very difficult to upgrade the cooling. Because of this, it is a rare video card that has a 0% failure rate. Most of the cards we sold in 2012 had a failure here and there, but there were three that stood out with absolutely no failures.

EVGA Geforce GTX 580 1536MB
015-P3-1580-TR
MSI Geforce GTX 680 2GB
N680GTX-PM2D2GD5
XFX Radeon HD 7950 3GB
FX-795A-TNFC

These are all fairly high-end video cards which, due to the extra power and heat, are usually at a higher risk of failure. Due to this, the fact that all three of these cards had no problems in 2012 is extra impressive. If you are looking for a more mid-range card, the EVGA Geforce GTX 670 2GB and the EVGA Geforce GTX 650 1GB both only had a single card that failed in 2012. For a workstation, the PNY Quadro 2000 1GB and the PNY Quadro 4000 2GB also had only a single failure in the last year.

If you are trying to choose between an NVIDIA or an AMD video card, both have similar failure rates overall. In 2012, NVIDIA cards had a 5.3% overall failure rate while AMD cards had a 6.14% overall failure rate. 

Power Supplies

A reliable power supply is important not just because it provides power to your entire system, but also because when a power supply fails, it can do so in a way that can damage any component it is hooked up to. So instead of just having to replace the power supply, you may also have to replace the motherboard, CPU, RAM, or more. Because of this, paying a little extra money to have a highly reliable power supply is always a good investment. And if you're looking for the most reliable, there are two power supplies that had no failures at all in 2012:


These are both fairly high wattage power supplies, so we also wanted to give a special mention to the Seasonic X-560 560W power supply. The Seasonic X-560 had a very low .56% failure rate in 2012, which makes it an excellent choice when you don't need a power supply as large as the two above.

Conclusion

So there you have it: the most reliable hardware for 2012. While there is too much data to make many broad generalizations, there are a few simple ones that can be made. First, server hardware such as Asus' Workstation series of motherboards and ECC RAM have very low failure rates. This is great to see as we recently changed our Obsidian workstation (which has a focus on extreme reliability) to utilize the Asus P8C-WS motherboard and ECC RAM. Second, Intel SSDs are very, very reliable. Even the product lines that did not have a 0% failure rate only had a failure here and there which is still great for overall reliability.

If you liked seeing this data, let us know in the comments below. We always want to publish articles that are interesting for both our customers and other readers, and your feedback helps us decide what type of articles to focus on in the future!

Frank

this is awesome!  would love to see more of this. 

In fact, it would be very cool to see the failure rates in the past 3 mos, 6 mos, 1 year...  next to each piece of hardware as a sort of live chart.  You could also include a metric to compare it to the category as whole (for example, the failure rates of all the power supplies you sell, or maybe just other brands, etc.) Then, as people are configuring their system, they can see the reliability information right there.  I know when I have configured Puget systems in the past, one of the highest requirements I have is reliability, and being able to see this as I select between the various power supplies in a Puget machine (for example) would be very valuable.

Posted on 2013-01-01 00:51:55

We might eventually figure out a good way to present component reliability on our configuration pages, but there are a lot of hidden factors that could negatively skew that data. For example, anything that was our fault (oops, I just dropped the chassis kind of thing) we don't want to show up in the reliability data even though we treat it as a failure. Also, we have to be careful about presenting just percentages for low-volume products. If we have only sold a handful of a video card and one failed, it wouldn't be fair to show the failure percentage since it would appear to be very high. In articles like this we can filter those out and present just the data we are confident in, but automating that process is very difficult.

This is certainly something we would love to do, but until we can iron out all the small details we will likely limit this data to these types of articles. Of course, our sales reps can pull this data for themselves at any time, so if you are configuring a system and want the best reliability they can certainly help you pick the best components.

Posted on 2013-01-01 01:15:16
Frank

this is awesome!  would love to see more of this.

When configuring Puget machines, often times reliability is the number one concern.  It would be great if the failure metrics you guys keep could be shown in live form next to each component as you are configuring the system to aid you in your decision.  Also, you could include some sort of baseline failure rate to give the failures in the last 3, 6, and 12 months some relative measure... such as the failures of all the parts in that category, or maybe failure rates for that brand versus others, etc.

Keep it up guys.  Hope you have a great 2013!

Posted on 2013-01-01 00:55:45
Marcus Streips

Very Nicely done.  I have to say that I am very very impressed with all of the QA information and processes described on this web site. It is very telling that after a month of assiduous research, this company's web site is the only one I have found that actually discusses hard numbers rather than resorting to puffery to illustrate its concern for selling and standing behind a quality product.  I can say the same about its shipping policies which state clearly that it does not recommend "going cheap" when selecting a shipping option for a $4000 plus piece of hardware.  They also immediately addressed my emails with thoughtful and actually helpful responses and even talked me out of selecting more expensive OC options which they deem a source for customer dissatisfaction.  In fact, this is the only company I have come across that puts customer satisfaction in front of claiming the best specs or best prices.  Kudos and keep up the great work!

Posted on 2013-01-03 13:39:04
Stefan Wessels

If I lived in the US I would be ordering from you guys! Thanks for a truly useful article (as is most of the hardware articles on your site) - keep it up!

Posted on 2013-08-06 06:24:28
nleksan

Huh... Of the systems I've built and managed over the past few years, Samsung has been far and away the most reliable SSD manufacturer, with Intel tying for 2nd place with Crucial (both sacrificing speed for reliability, something I've never felt to be true with Samsung), and in third place is Plextor and their M5P Xtreme which is built with kinda the opposite mentality of Intel/Crucial, and goes for the extra speed with a slight decrease in reliability.

However, all the above are below 1 percent.
If you want to stay that way, DO NOT buy junk from companies like Corsair who simply relabel someone else's products, or OCZ who can very rarely get it right. Those two companies were the worst, with 8 of the 29 Corsair and 18 of the 76 OCZ drives failing, and that's excluding DOA (which would be 15/29 and 25/76)!!!

Only zero failure HDD so far has been the WD10EZEX, with a total of 55 in customer machines and because I was so impressed with their performance, silence, and infinitesimal heat production, I have now 27 of them spread across my different machines (X79 rig, Z77 benching rig, audio studio rig, HTPC1, HTPC2, and both backup server and main file server). They are superior in every way but warranty to the blacks, but I do still use and recommend the RE4/RE drives for true RAID or mission critical work, and the RE/4 is the only other drive model used in either of my servers (RE4 1+2TB, RE 2+4TB).

WD Green and Red have been below average, Seagate has performed above average as a whole and in fact better than they ever have, Hitachi has great drives so long as you avoid anything budget, and Toshiba appear to be simple rebrands of HGST models.

Posted on 2013-08-13 19:29:41