Rolling Your Own Business Desktops? 643
mike asks: "I'm mulling the logic of my company building its own desktop computers. As the IT Manager (plus sysadmin, janitor...) of a struggling-yet-thankfully-still-alive dotcom, money is really tight. We have around sixty ~400MHz desktops which are increasingly showing their age. Acceptable P4 systems from the big guys run at least $1000. By recycling the OS (Win2k), case, cdrom, floppy, and K/V/M, I figure I can assemble a good AMD system for about $600. That's a 40% savings. Is it worth it? The cost difference could very well determine whether this project proceeds or gets put on the back-burner again."
"Some negatives about rolling my own:
- Management: I won't get the special business features offered by some manufacturers. Dell's OpenImage, for example, looks awfully nice. But how much does that really buy me in a company of 60 machines? I don't use such stuff now; am I missing out on nirvana?
- Time to build: Even though we'd leverage Ghost wherever possible, handmade systems nevertheless take time to build, load, & configure.
- Supporting different platforms: Because money is so tight, I can at best afford a capital replacement rate of 25%-33% (15-20 units) per year. That means I'm committing to the support of 3 or 4 different platforms. Having just one platform is great, but how many companies, even ones that actively strive for it, truly enjoy that luxury? I inherited two platforms (Micron & Gateway); support isn't that bad. With proper planning, I don't see why we can't support four.
- Hardware quality: How much can I trust a popular Athlon chipset in a business environment? I feel silly bringing this up because I have a few Athlon systems at home, each with a different chipset, and they've been nothing but rock solid. But I know the lack of a really good chipset has been a large contributor to why AMD's aren't more prevalent in the business world. (well, that and long term bullying by Intel).
- I don't get a proven, prepackaged system that works right out of the box.
- Cost savings. Plain & simple.
- Increased horsepower per dollar spent.
- By choosing my own equipment (mobo especially), I suffer fewer OEM shortcuts.
- I have to admit that I'd enjoy the pure geek satisfaction of rolling out 'my' creation to the company.
For those that are curious, Ask Slashdot did an article on the AMD issue, here.
Why not just only buy / build new machines for ... (Score:3, Insightful)
I mean really suzy in the phone center has no need for over 400 mhz, I'm striving along just fine on my 667.
Cost savings? (Score:5, Insightful)
-matt
most folks won't need the upgrade (Score:2, Insightful)
Re:Can you actually recycle the OS? (Score:2, Insightful)
Why upgrade? (Score:5, Insightful)
If you must, go out and get some low-end consumer PCs and buy a bunch of spares: it's less work than building your own and still very cheap.
Support issues (Score:2, Insightful)
If you're already supporting the systems, though, as you make it seem... then this may not be an issue for you. Just find out about RMA policies of your vendor beforehand!
Just one tech's opinion (Score:4, Insightful)
Even if they cost a little more, I think you'll find yourself grateful for a warranty to fall back on. Plus, when machines go boom, you aren't instantly blamed. If you roll your own, any system that crashes will be pinned on YOU, and you alone.
I know that's not a situation that I'd like to be in. Would you?
Re:Go for it (Score:5, Insightful)
I've done it... (Score:4, Insightful)
I worked in a 50-user shop, and provided services and equipment to a 200-user shop under contract.
In our case, the only way to get decent specs and meet the client's budget was to roll our own. The other options were too few systems, or systems too cheesy to contemplate. Cheesy as in crap, not as in creamy goodness.
If you go down that path, my suggestion would be to make sure you have confidence in your component choices, and that all your component choices interoperate flawlessly. Any system you have to see again will blow the savings - your first callback or return could be fatal. Make sure you source quality components, and if you're trying to minimize the number of discrete configurations, buy all your components at once.
Spend money on decent cases with good power supplies. Don't yield to the urge to "cheap out" on components that "don't matter" - they all matter. Don't buy cadillac parts, but make sure everything you do buy is good quality, sound, and durable. Keep extra original parts on hand, especially a mobo or two.
Come up with a logo and have the stickers printed - it amazed us how many people would readily accept a brand they'd never heard of, but would never accept an unbranded system.
Your initial problem will be evaluating a number of different hardware options, then settling on those you want to standardize on. Once you get to that point, what do you do with the bastard love children of your prototype period? Don't deploy them to users, you'll water down any faith and confidence your production systems should inspire.
A bit of comparison (Score:2, Insightful)
DIY Rollout:
Pros:
You know exactly what parts are, or are not, in your systems, and can (usually) get spares easily.
OS installation/options/configuration is(are) also a known quantity.
Can be extremely cost effective to roll out.
Cons:
Warrenty is provided by whoever you bought your parts from (new), or long-past dead on recycled parts (in most cases).
Tech support? Look in the mirror!
Large Scale network support? See above.
There are a few good reasons for a DIY rollout, but the long term support may be the price you pay later. If you have confidence in your skills, and have a friend or two that can help you out when "it" hits the fan on the next "I Love You" type virus hits, I'd say, Save The Cash, and Go For It!
If your Boss (the guy signing the cheques) want "Guarantees", you just might have to talk to a Big Name company.
Here's a thought: Try selling off your older componants. The extra revenue, however small, might be enought to help get things rolling.
Good Luck eigther way!
Two words: Dell Refurbished. (Score:3, Insightful)
I got two Celeron 1.1GHz systems and a Pentium IV 1.6GHz for $588 each (shipped!) Here [slashdot.org] is a Slashdot post that details my experiences with them.
There was absolutely no way I could undercut Dell on price by building my own -- especially not when you include the cost of Windows XP (preinstalled), one-year on-site warranty, and the awesome cases that open with the press of a button.
It really doesn't make sense to build PC's yourself anymore when manufacturers are offering PC's like this for bargain-bin prices. Plus, you can always recycle monitors as well -- that's what I did with this set.
Building your own will certainly give you job security (as someone else mentioned), but it will also give you no end of headaches. Why doesn't video card A work with motherboard B? And installing Windows 60 times is enough to make even the bravest person run away in fear. Even with a copy of Ghost in hand, you still have the daunting task of putting everything together (and charging the company for your effort). In the end, it's really not worth it to either you or the company. Besides, do you really want to spend the next two weeks testing out RAM and hard drives by hand? Bleh.
Re:I think time is probably the critical factor... (Score:5, Insightful)
Preach it, friend!
For the last near 20 years I often built systems for friends, family, or businesses who wanted to save money. But these days I can't build them cheaper than Walmart sells them. The only time I build a system now for anyone other than myself is if all they want is some old wreck good enough to get on the internet and I already have the parts laying around.
To the guy who started this discussion: You start out by comparing a pre-built P4 system to a scratch-built Athlon system. You also need to be looking at Duron/Celeron pre-builts. I bet there is only a handfull of people in your company, if ANY, who need the power of a P4 or a top of the line Athlon. A Celeron or Duron would them just fine, and you're not going to build something from scratch with either of those that is cheaper than what Dell or Walmart can sell them to you for.
What's time & effort worth? (Score:3, Insightful)
Put this all on paper, generate some good estimates of costs & time allowences, failure rates & resolution times then present it to the CFO. Even for a company in a cash crunch these are generally compelling arguments that are well understood by the numbers folks.
They they don't bite then ask yourself if you want to hang around babysitting these monstrosities as the rest of the world moves on?
Been there, but with servers. (Score:2, Insightful)
Re:Microsoft allow it? (Score:5, Insightful)
Waaaaaaaaaaait a minute. (Not you, the original questioner).
What's really wrong with these systems in the first place?
He's got an assload of 400 MHz P2s, probably Slot-1-based, and each box has either 64 or 128M of PC100 SDRAM.
Why not buy a bunch of Celeron or P3-800ish chips and FCPGA (new-sk00l slotket), and another 128M of RAM for each of them?
I think you could get a decent CPU and RAM upgrade for less than $100 per box.
Moreover, you wouldn't have to reimage any drives - it'd be a straight hardware swap, with maybe 15 minutes to figure out what voltages the motherboard supported, and to configure the slotket or motherboard correctly. (If you had quality components to begin with, this might even be automatic).
Add onto that maybe 15 minutes per desktop to properly apply thermal transfer paste.
No EULA concerns, no hardware/driver concerns, and it's dirt cheap.
I'd bet that you, plus one or two of the "hardware geeks" (you know who they are :-) in the office could do this overnight for $100 per desktop, plus the price of a case of beer and a couple of large pizzas with all the trimmings.
Re:Go for it (Score:3, Insightful)
It's been done and worked. (Score:1, Insightful)
At a company I worked for our team couldn't weren't allowed to purchase computer systems, but needed new machines. The manager went to a small local computer store and bought 30 cases, 30 motherboards, 30 videocards
The biggest complaint I have with Dell, et al. is the use of "custom" solutions. Old Compaq DeskPro's were the worst things I've dealt with, I absolutely hated having to find the proper EisaConfig disk for the things everytime I added RAM. I _really_ hated the time I nuked the invisible harddrive partition Compaq had put EisaConfig on.
Don't get systems with special cute management "features" they all to often break with a minor OS update, and the vendor rarely if ever keeps the systems upto date.
Re:$600? Surely you can do better than that. (Score:3, Insightful)
What on earth is an office PC doing that needs an AthlonXP 1600+? (OK, the sysadmins need that to play Quake after-hours, but what about the guy who only uses Excel and Powerpoint? :-)
For that matter - sure, the WD 5400RPM 40G drives are down to $52 - but what are office PCs doing that requires 40G?
This may depend on what he's already got -- if these PCs have only 100M of space left on ancient 2G drives, then fine, upgrade the drives to 40G. But if they've already got 6-8G drives (which probably have 4-5G free), and all the "real stuff" is stored on a central server, and all the user machines have several gigs free, isn't that enough space for your employees to store their downloaded MP3z and pr0n? ;-)
What does he need a newer video card for? Are his users likely to run 1600x1280 on their 17" monitors?
For office computing, you can often KIWI - Kill It With Iron. Add more RAM, swap the CPU for a P3 at either 133 MHz FSB or 100 MHz FSB, and see if it still sucks. If it ceases to suck, the problem's solved, probably for less than $100 per desktop.
Re:And don't forget (Score:5, Insightful)
This guy is either going heavy 3D, or something like that...or he has quite a strange concept of "obsolete". Add some RAM to those machine be happy with them.
*sight* People don't know how to take care of computers anymore :-(
Re:$600? Surely you can do better than that. (Score:2, Insightful)
Re:Go for it (Score:3, Insightful)
I don think so mon (Score:2, Insightful)
Add up not just what you make an hour, but what your company pays to keep you around, including overhead. Most places you can double your wage. That's what you 'really' cost. Less? OK. Point still holds. We're buying perfectly servicable machines with giga wazoo drives and gigahz processors sans monitors (which have not broken) for between $700-$800 with a three year warranty. It breaks, they ship us a new one or repair it on site. Period. Your parts are costing $600. You're going to have to make money on a hundred dollar margin. That spread is too thin. You've got to manage the parts, store them, and get everything working quickly. That's a lot of prep time. Even if you managed to break even on paper, couldn't you be doing something more useful for the company? I know it's fun (I used to put together computers with nothing but a swiss army knife at trade shows), but you're supposed to be out there making a million dollars. Put your energy where it can be leveraged. --Just my opinion.
Get a local computer store to do it (Score:2, Insightful)
Also, I'd be surprised if you really need to upgrade all your PC's. 400 mhz is still a pretty good box by most standards. And you needn't upgrade to the highest-end Athlon stuff except for maybe a few of the most power-hungry users (CAD, image rendering, etc). For most typical office applications, 400 mhz PII with enough ram (maybe all you need is more ram?) is quite useable and a 1000 mhz Celeron or Duron is plenty. I'm using a 750 mhz PIII right now and it wouldn't occur to me to spend money upgrading it.
Re:I think time is probably the critical factor... (Score:3, Insightful)
To be honest, I rather like Dell's new keyboards (the black ones). Nice soft feel to them. At any rate, just let your secretaries and transcriptionists keep their old keyboards, and everyone with a wheel or trackball keeps their old mouse, and all is well. Very few other people would even know the difference.
We're talking business systems here. No worries of game playing. At most, we just need RealAudio/ Windows Media sounds, and MP3's.
I don't think business PC's should come with more than a 10GB (or less even) HDD. People shouldn't store things on their hard drive, it makes support a nightmare. Give em server space, and lock it with quotas. Very few office workers need more than 100MB in personal space. After that, it's time to clean it out or achive it.
Do your cube farm cohorts a favor and pull all speakers out of there. Give them $5 mono (so they can talk on the phone, listen to office chatter, etc.) headphones instead.
Good point. You should upgrade the RAM with Crucial memory after receiving the machines. Go with at least 128MB for Win9x, 256MB for Win2K, and 384MB+ for WinXP (if you're actually running that POS for some reason). Get it from Crucial and save the extra $100 bucks the OEM would've charged you.
I've never known Dell to not use a 3Com NIC. True, their integrated ones are a 3C920, but they're pretty good. For around $50, you can go with an industry standard 3Com 3C905C.
Though I historically despise Winmodems, they're more than decent enough for occassional use on a business system. Who doesn't have at least DSL anymore? Only older programs still use modems, and they're relatively few and far between. If someone really uses a modem, leave off the internal Winmodem one and get an external USRobotics. Otherwise, stay away from Dell's USRobotics hardware PCI modem, as it goes on Com5, which causes problems with most older apps.
Reuse the old monitors. Also, keep a few in stock for quick replacements when they do die. Most users will stick with 800x600 on a 17", and 1024x768 on a 19"+ anyways, so the clarity isn't too big an issue. Only your heavy spreadsheet users and graphics pros need a really good monitor.
As for Linux compatibility, that would be nice, but is mostly irrelevant since he's already said that it's a Win2K shop. Keep it in the back of your mind, but, for the most part, you'll only need to replace a very few components if you ever migrate. Minus the OS licensing fees, you'll still save a boatload of money.
Your arguments may be valid for a home system, but a business PC is completely different. Most of the points you brought up are simply not relevant for an office worker's PC.
This sounds like a bad idea.. (Score:3, Insightful)
This would also take an insane amount of time. Sixty machines is a lot of boxen; optimistically it would take you an hour per machine to swap all the hardware around and reformat the drives and install Windows. In other words, you'd be down at least a week.
Order sixty new machines from an OEM and you're down 2 days tops. Plus you get the guarantee that the machines work (out of 60 boxes, you're bound to get some bad hardware) and you get a warranty from a reputable company, not to mention saving yourself a MAJOR headache.
If something goes wrong with one of the machines, you just call the vendor and straighten it out. If you roll you own, you have to spend time doing diagnostics, then tracking down the receipts, RMA from parts warehouses, limbo time for replacement parts.. All this time your company is paying you to not do the job they hired you for.
Sure, they can hire on another guy to help you, but then there goes all that money you saved having to pay his salary. So in the end, your company didn't really save any money, they just have 60 new machines with no comprehensive warranty, poor tech support, and probably a very frazzled and stressed admin. The geek factor sounds fun, but in reality, it would be more practical to order from a vendor.
Re:And don't forget (Score:2, Insightful)
Of course, with Microsoft dominating the market, this trend of making hardware obsolete before its time is sure to continue. I guess that just means people like you and I will be able to find great hardware for next to nothing.
Another upgrade suggestions: memory (Score:2, Insightful)
However, I am wondering about the reason why your colleagues find the limitation of 400MHz speed. Often users mistakenly think their CPU is slow when their disk drive is swapping for more memories. One easy workaround is simply addming more memories. I suspect 400Mhz machines used to be shipped with 128Mb of memories. Upgrading those memories to 256Mb or 386Mb will be one of the most effective performance-per-unit-cost upgrade solution.
Also, think about reinstalling Win2k. Win2k is much more stable and faster than Win ME or 98, but formatting and reinstalling the entire partition often cleans up the trashes built up in the system registry, leading to overall performance enhancements. A user might have a lot of autostarting programs installed hidden.
Investing on the displays and the human interface devices (keyboards and mice) can be another solution. If there are a lot of users complaining about their monitor performance, then try to *sacrifice* performance upgrade to the monitor upgrade. A decent 15" TFT flatpanel monotor can cost $400 and people often get less fatigue when they work with TFT flatpanel than CRT monitors. For me, I'd rather have a flat panel monitor than to have processor & memory upgrade. A combination of memory upgrade and the flat panel monitor purchase can be very appealing under certain circumstatnces if the workers have to look at the monitors for a long time every day. You don't need to squeeze your budget with a lot of possible component combinations in this case.
Good luck,
Use nforce to save time (Score:2, Insightful)
Re:I think time is probably the critical factor... (Score:4, Insightful)
Only on a standalone system. On a LAN you end up with workstations with lots of useless diskspace or worst provide an incentive for (l)users to store work on the local HDD.
Comment removed (Score:3, Insightful)
Re:AMD vs. Intel (Score:3, Insightful)
But you have brought up an interesting point. Know the old adagium 'no one's ever been fired for buying IBM'? Well, it's still around today and it's called 'no one's ever been fired for buying Intel'.
If your AMD craps out, your boss will be all over you for choosing a 'non standard CPU'. If your Intel craps out, it's just bad luck (which it is if your AMD craps out as well, but your boss doesn't know that).
Re:Go for it (Score:4, Insightful)
my $.02
Re:Microsoft allow it? (Score:3, Insightful)
But whats the fine line, you change cpus, or keep same cpu but change motherboards....
Personaly, who cares what the EULA is, just do it , as long as its not running on the old pc too.
No offence, but that's an Amateur's question... (Score:4, Insightful)
Remember the article the other day when we all laughed at the Wilkes Barre IT guy who stopped the IBM maintenance on the AS/400 ? Well, this is the same kind of thinking: penny-wise and pound foolish.
Remember: support contracts are a form of insurance. They insulate you from the risk associated with the issue at hand. When looking at any form of insurance, you have to take into account what the worst-case senario is, and if you can handle it. In your case, the scenario is that you have multiple desktop failures, including critical failures of important machines (ie, severl of your main developers). Do a cost analysis: if I do a roll-my-own machine, what's the cost of it breaking? How much does it cost for that developer to have no (or a seriously inferior) machine for a week or more, vs. the 1 or 2 days a supported machine would be out?
For small companies, (especially those heavy in software development) I can't imagine a situation where the TCO of a fully-supported system is worse than a roll-your-own box. None The downtime and IT personnel time alone will kill that equation. For huge companies, it may pan out, but for a 60-desktop company with 1 IT person? Not a chance.
You need to put this into perspective with Management. Once again, they are looking at only the up-front costs, and none of the hidden costs, which in this case are the majority. Explain to them what the true cost of a desktop is, and how NOT buying a supported machine results in a WORSE return over the next year.
Now, here's a couple of recommendations for getting SUPPORTED desktops into your organization while not breaking the budget and still meeting increased performance needs:
I don't mean to harp on you personnally, but this kind of thing is why IT has a long, long way to go before being really professional. Folks, this isn't a garage. IT folks need to quite thinking like it's an expanded hobby, and also need to remind the Executives of this, too. It's a Profession, not a Trade.
-Erik
Systems/Network Architect and former SysAdmin
Make your users suffer (Score:2, Insightful)
roll your own - false economy (Score:2, Insightful)
Those parts that you are going to re-use are also getting old. Here are things that can go wrong:
- The cases have been sitting under somebody's desk collecting dust bunnies. Those fans will start sounding like lawnmowers soon. Then you will have to replace the fan or the power supply. Parts + your time = You just blew your savings there.
- Those CD ROMS have also been collecting dust and getting beat up by users. They'll start to act funny, have trouble reading disks, etc. Inevitably your boss's boss will have trouble with his. After screwing around with it he will call you to replace it. You just blew your savings and lost a little credibility.
Buy from a good manufacturer with good service. If something goes wrong, you can have them mail you a new part. You don't need to stock extras.
At my last job we made a deal with HP. We got the desktops for only about $50 more than we could build them. They came with 3 year warranties. Since we were buying so many of their machines they took care of us. Any parts we needed, we got. Every year we budgeted for 1/3 of our total machines. Now nobody has a machine older then 3 years old, they don't have any oldball machines sitting around, and OS and software deployment only need to be tested on a few types of machines. Users are happy because they have good machines and good response for the occasional hardware problem. Support folks are happy because they are bothered by bad fans, etc.
roll your own (Score:2, Insightful)
two years later, would i ever do it again? nah. replace what you can replace, it's not worth the headache, and while i don't use openimage i will say that using a single vendor when possible is very nice, i have only 4 images for a company now with ~130 desktops...
if you know what you're doing, it will be a fun project... but i'll tell you building all those machines gets pretty damn tedious.