Puget Systems Pitches Virtues of Prebuilt Versus DIY Based on Component Failure Rates

24

Comments

+ Add a Comment
avatar

lordfirefox

I've never bought pre-built before and I sure as hell am not going to start.
Thanks but no thanks Puget, you can keep your pre-built garbage for yourself as far as I am concerned. No Sale.

avatar

aferrara50

Even if this is true that's what warranties are for.

avatar

NavarWynn

Their conclusion that their systems are more reliable/ fail less than DIY systems is completely unsupported by their posted data. In fact, the failure rate they attribute to DIY/ home built seems to actually be THEIR initial build failure rates.

I see ZERO data collected from ACTUAL DIY/homebuild Builds. If they want their conclusions to be taken seriously, they'd need to actually provide more numbers, and their specific methodology.

In a nutshell, their claim to your money is that a) they do labor, b) they weed out bad hardware, c) when there is a failure, they deal with warranty/ returns-exchange.

Dont get me wrong, there certainly IS value in those things, but ALL the things they do can EASILY be done by a DIYer, and if done so, a DIYer should expect EXACTLY the same failure rates.

As an aside, ALL this info is garnished from initial, and in warranty failures. Even though components often don't have a long lifespan in an enthusiast's rig, I think every one of us wants reliability data from BEYOND the warranty period/condition, afterall, in-warranty failures are covered by the warranty.

To me, the only nugget of wisdom in the whole post is the actual 'qualification process' - i.e. a checklist you can (ie. SHOULD) go through after you've assembled your build. A checklist isn't rocket science (and is one of the reasons so few planes crash), but a comprehensive checklist is ignored by 99% of DIYers. Maybe MPC could do an article on JUST the 'post-assembly' checklist?

avatar

John Pombrio

Hmm, looking at the blog itself, the overall impression is "low end stuff". What that shows is that most of Puget Systems major sales are for lower end systems. At those price points, I would expect a lot more DOAs as the stuff would have a lot less stress testing on them. No wonder they are having such high failure rates! The mobos are nothing like anything I would buy and the memory chips don't even have a heatsinks on them. This blog has really nothing to do with the enthusiast market who consistently buy higher end products and run them more than stock speeds.

avatar

Paul_Lilly

Give it another look, John. Puget listed what it found to be the most reliable hardware of 2013, based on failure rates in-house (bad parts it caught before shipping to the end user) and out in the field. The fact that so many lower-cost products made the list suggests that entry-level parts might be more reliable than you think, not the other way around.

avatar

John Pombrio

Paul, its the mobo and graphics card failure rates that I find suspect. A ~10% failure rate for an Intel based mobo/Intel CPU (4.71% PLUS 4.23%= WHAT exactly?) seems awfully high. I worked on HP/Agilent board test systems and actually saw how PC mobos are tested before leaving the factory (only Intel ones in Puerto Rico tho). Low cost boards usually went through board test but did not have a burn in just boxed and shipped while high end boards are burned in a temperature rack for a day or two then retested. NO WAY is a high end mobo going to have that high a failure rate unless Puget is testing way outside of specs or buying low end mobos.
As for graphics cards, I cannot comment on how they are tested before shipping, but again, 4-10% seems awfully high, especially for EVGA, ASUS, and other high end value added manufacturers.
Now if Puget is buying a lot of boards and graphics cards from low end bulk manufacturers who are skipping the burn in, then those rates make perfect sense.

avatar

Daniel J. Hill

I've only lost one motherboard and that was an ASUS M2N SLI Deluxe. Never went back to ASUS after that, since MSi and Gigabyte have never let me down. I also lost an lost an EVGA 8800GT in SLI setup after 6 years of hard use. The other EVGA 8800GT is still going strong. Never had any RAM problems with GSkill or A-DATA.

avatar

fung0

Ah, the seductive quality of anecdotal evidence. Personally, I've used ASUS motherboards exclusively for over a decade, built a couple of dozen systems and never had the slightest problem... The oldest ones were still running when I mothballed them or gave them away.

avatar

maleficarus™

Interesting but not surprising really. I have used Kingston for two DIY builds myself and both times problem free! Currently using Corsair at the moment and no issues either.

Like another poster already posted I too have never had a hardware failure in any of the 15 or so PC I have built over the years with the exception of mice and keyboards. I only build using ASUS and Intel. I also only use ASUS for my GPU and sound. PSU's I stick with Antec and as matter of fact running a EA500 right now that I have been using since 2009. I have read a lot of people will not touch Seagate for hard drives but I have never had a problem with them so I stick with them as well. Currently I am using a 1TB Seagate Hybrid drive and it runs fantastic!

avatar

fung0

I have to question Puget's failure stats. They say that if you build an Intel/Nvidia system yourself, you have "about a 1 in 7 chance of their being some sort of hardware problem." That's WAY higher than anything I've experienced... unless you count me forgetting to hook up the HD indicator LED.

avatar

Ruins

I ordered a pre-built from Velocity Micro many years ago. Unfortunately, I never got the chance to fully evaluate the machines reliability. The day it arrived, I was home and outside working in the front yard. The UPS driver proceeded to drag the box down the each step of his van, then he picked it up and promptly dropped it from waist high to the concrete driveway.

Once in the house I unpacked the computer and heard some ominous loose parts inside the case. I still plugged the thing in and of course no results on the monitor. Since this was my first computer I did not know anything about the innards or even how to open the case.

So, on to the phone to tech support at Velocity Micro. They talked me through how to open the case and take a look. Once opened, the tall and heavy heat sink was completely off with the CPU still attached bent pins and all. Several of the motherboard capacitors were bent, and one of the RAM modules was at almost a horizontal position to the motherboard. I asked what to do. They say to take pictures and they would send me new parts.

While waiting for the parts to arrive, I researched how to build your own computer, which is exactly what I did when the parts arrived. Although, I was terrified with my first build, the pleasure and feeling of accomplishment once finished was fantastic. I now have eight more builds under my belt. Not one has had a component failure. With all the builds I research the components thoroughly it the hopes that no failures will occur.

I am sure that there are several boutique computer builders that can produce a reliable machine with quality parts, however due to a careless UPS driver, I will never experience that. For me it is home-built.

avatar

vig1lant3

I've been using Kingston RAM for almost 20 years. I think my first set was PC66?. Paired with an AMD K6 (350Mhz!!!)...blistering speed!!

Anyway, I've used everything from A-Data to Team over the years. There are a couple of brands that have never let me down, and Kingston is one of them. Couple that with the fact that Kingston has been around at least twice as long as a lot of other contemporary brands, and it really doesn't surprise me that Puget rated it so high.

What does surprise me is the fact that they came up with a reason that pre-built is better than DIY. Sounds a little self serving to me...I'm suspicious...

avatar

jtrpop

Quality parts = quality products. I'm a little surprised by the RAM. I know Kingston is good, but I know there are other quality brands such as Crucial, Corsair, Transcend and Patriot to name a few. I see a few commenters mentioning GSkill, but I have had way too many incompatibility problems with them, even when their website says it is compatible, that I just cannot buy or recommend them. If I want something to work or a server is involved, I go Crucial.

avatar

Paul_Lilly

I was surprised by the RAM claims as well, which is why I reached out to them for some followup info. While a little dated at this point, Puget posted an article in 2009 that details its experience with RAM brands -- you can check it out here.

avatar

Gman3968

I have built many over the past 15 years. I have never had a failure due to hardware. I have always have success with AMD procs and videocards. I like Gskill RAM and corsair PSUs. I will not pay for you put my machine together. It's a waste of time and money to pay for so called expertise.

avatar

Rogue74

I've built about 4 PCs always using top shelf parts. I've never had any major problems. I had one BFG Tech (RIP...) 1000w PSU die, but that was after around four years. Seems to me like a shameless promotion for this company, which I've never heard of.

avatar

trog69

When I bought this rig from IBuyPower PC, I went to PartsPicker and looked at what each component would cost, and I would have saved about $150, and after their 5% off for deferred shipment, it was only about $50 difference. For someone else to do all the work, and testing, I'll gladly pay $50. Plus, they have a great forum and really fast support response, and a lot of people have said that they have a no-hassle replacement policy.

The only component I've had go bad on me was the EVGA GTX680 FTW card that went from 30c idle to about 50c, and then croaked during Skyrim. The replacement card has been a champ ever since, and I hope it works as well until Summer, when I plan to upgrade to either the GTX 780 ti or it's Radeon equivalent. I'm no fanboi, and I like both AMD and nVidia. My old rig's HD5770 card has been a stout little journeyman for many years, and still plays Skyrim with all the eye-candy at ultra @1080p. With no heat to speak of, it's everything you could want in an inexpensive card. No idea how it's younger brothers compare, but I'm gld I never sold that one.

avatar

OpTicaL

Obvious sales pitch is obvious. I exclusively use G.Skill RAM and they have yet to fail me and MSI's FMS2-A75MA-E35 is one solid microATX mobo *knock on wood*.

avatar

Paul_Lilly

It's not like Puget Systems is trying to hide the fact that it has a vested interest in people buying prebuilt machines versus the DIY route. The level of transparency is actually one of Puget's strong qualities -- there are several articles on the site that run the gamut from testing procedures to blog posts on a variety of topics.

Likewise, I don't have a vested interest in Puget Systems, but I really like that Bach and company share this kind of data and explain why they choose the components they do. There's not a lot of system builders that share this type of info or write up articles talking about thermal imaging.

avatar

Perrorist

Comparing one's own experience with that of a company that builds thousands of PCs means nothing. I'll go with the larger sample size.

avatar

RICHARD RIDDICK

Building it yourself is much better, you pay about the same as for a pre-built system but you get much better components.

avatar

John Pombrio

Interesting post Paul, thanks!
Hmm, notice that they don't say THEY have a very low failure rate, they only point out that their systems they SHIP have a very low failure rate. They also lean towards certain manufacturers doe to past experience.
I have had good luck with DOAs over the past few years in building my own stuff ( zero in the past 3 years). I too prefer only certain manufacturers (ASUS for Mobo, Corsair for cases, memory, and power supplies, Intel CPUs, EVGA NVidia Graphics, Hitachi hard drives (now Seagate Barracudas). I also have not destroyed or broken a part during a build in years and years, mainly becuase of all the stuff I DID break in the past 20 years of rolling my own computers.
Frankly, if I buy quality components and am careful in the build, I expect a boot on the first try and I usually get it. But again, I have had a LOT of experience over the years.

avatar

pastorbob

John => I noticed that too. It would be interesting to see their specs on component failure rate during burn-in. But then again, as you point out, buying quality name-brand components lowers the possibility of failure considerably. Over my thirty years of DIY experience I have only had one case of component failure in the first six months. And that was the one time I went with cheap no-name memory. Unfortunately it was in a system I built for my son. And the problem was intermittent BSD's. But I made good on it and swapped out his memory for Kingston and he used the system for over five years without issues after that. Learned my lesson and always use quality components now.

avatar

EKRboi

While I did find the numbers interesting, its hard to not look at this like a sales pitch. I'm not going to pay extra to a boutique builder to assemble hardware and stress test it for me. Sure it blows to put the rig together and find out you had a bad/weak egg when it won't boot or a component dies during some stress tests, but that's just part of the adventure, inevitably I'm going to run stress tests and benchmarks on new hardware.

On a side note, looking at their #s I see why I have had very little failure rate over the years.