Fighting for Workstation Upgrades

[beestee]

Hello,

I could use the groups input on what seems to be a sensitive subject within our company. Any comments and input will be appreciated.

We have a group of six 3D artists and we specialize in commercial architectural visualization. We are trying to get upgrades for our hardware and software, but it is literally as difficult as pulling teeth to get what we need. The man in charge, footing the bill, really likes our output and sees no need to upgrade our hardware. I feel that nothing will be done until our quality and production make an apparent and irrevertible drop.

Our average production per artist over the last week was 7 projects per artist with an average of 45 hours per artist. These projects range widely in difficulty and scope.

Some of his questions in a recent meeting included:

Why are upgrades necessary when current hardware appears to be productive?

Why is hardware the primary topic of productive needs?

What budget and timeline would be proposed?(as opposed to our current non-existent budget)

Would it be fair to expect increased production from new hardware?

If not, how is the purchse of this hardware justified?


Thanks,

beestee

EDIT: Forgot to mention, most of the computers that we are using are at least 3 years old, and were what I would consider a low-budget, piece-mealed workstation. In fact, I own computers myself that perform better than what the majority of our group works on.

[Morbid Angel]

it sounds like you got a hard one... Often business men will not see any need to invest money into already working pipeline. What you have to explain to them is that, while the pipeline is working and the projects are getting done, it is not by ease.
I would imagine with 3 year old equipment the render times are quite long, performance on heavy scenes can be very laggy. What this results to is artist having to waste a lot of time, waiting for computers to do their thing. So, if thats the case, it can be explained to them that by upgrading the hardware/software you are sure to increase the productivity, and quality of the projects. However, this also depends on the kind of projects that you are working on. If its simple stuff, and there is no need to upgrade then...there is no point. Otherwise, I would say that upgrade every 2 years is a must.

[ene.xis]

that s the worst decision ever to make as a company manager!
it s an huge mistake to relly on the aquired output quality.

As a principle u cannot expect to mantain the current product sucess by itself...the product must to be always in an evolutive state, if it stagnates his success will also drop...as so investment and marketing are keys to achieve success througout time!

[unknown]

thats stoopid business sense. I run a company with some other friends.

Im in charge of IT spending so we get the best. makes perfect sense. its also easy to work out. how much time spent waiting for a render, against how much it costs you to sit there, against buying a renderbox. Makes perfect sense. As someone at ILM said (ill be damned if i can find the article but it was amazingly obvious) the perfect situation is an Artist never waiting for a render.

Its much cheaper to buy loads of machines than have loads of artists sitting around twiddling there thumbs. It also makes more sense bulk buying, better discounts, its also better to buy the best, as all mahcines should be replaced every other year. Also buy with a guarenetee and call out, in case anyhting blows, never build your own. If a workstation breaks then it gets replaced.

buy max and subscriotion, this future proofs max, vray gets updated for free. so no cost there (praise be to the vray).

at the end of the day you shouldnt blame your tools, but with the right tools you get maximum benefits. then its down to effciency, and talent. another difficult area. Every good company in the IT sector forks out for top quality workstations. it would be madness not to.

think that covers everything.

[EricCamper]

THis is exactly what we go through every time....BUT I will say the the MAN in charge understands the need for equipment that is of the "age" I guess. Is it FAIR to say that your productivity will INCREASE? YES...But will you put out twice as many jobs....Probably not....I mean I do most of the rendering at night anyways......So the better workstation will only aid me through the day in a way that perhaps I wont crash or run out of memory as much..etcc..etc..
Of course that is the perfect business situation...Have people that look over all those aspects that he brought up and really fine tune it...at least your boss isnt picking the WORKSTATION or Software for you. I knew a firm that is exactly what they did. " Hey guys... I just picked up these great deals over the holiday at Walmart....and look...You get a free Lexmark printer" Oh and here is 3d Home architect...I have heard good things" LOL

Good luck my friend...I would suggest start cutting small wires and soldering....But I am sure you alreadty thought of that...

[beestee]

Thanks everyone, these comments will surely help. Keep 'em coming.

...at least your boss isnt picking the WORKSTATION or Software for you.

We narrowly escaped a mandatory SketchUp switch.
Not that SketchUp is bad or anything, I just feel that it would be a great leap backwards from where we are now.

[glyph]

why not persuade him to upgrade one or two machines now... (presumably he has to clear this with the MD or higher ranked drone so needs ammunition himself)
then after a few months the improved throughput (and/or other variables) will be clear for all to see (or measure in his case)

advantages :
he can spread the cost over the year
he dosent have to commit to upgrading all machines till its proven
machines purchased later might be faster /cheaper

I believe that could be called a 'win-win' (gotta learn to talk his jargon and he'll be putty in yer 'ands)

[trick]

A good professional ALLROUND architectural CG-artist only needs very little time to do lighting and material setup, which are the only things that need the perfect workstation. If you're smart you can do more in POST then spending lots of GHz/Hour on doing the above. I did MODEL very large project in 3DStudio DOS with my pentium 50MHz, including 8K composition in PShop. Lots of people want the perfect output straight out of the rendersoft, but even for animations this surely is not neccessary. Now if each of your crew has different tasks, and one of them indeed only does material and lighting setups, then buy him the fastest one can get...With 40 jobs per week it does not seem your clients are complaining, but you could always convince your staff to put an expensive setup on one of your co-workers to see if it makes any difference in quality and workload...

[chris-x//]

unhappy here too, here they don't wanna upgraded till they can be sure we won't lose any second, but I mean second waiting for the render to come...

I said killing rendetime was impossible, that only we could go and reduce them even with buying damn single quad core... But no, I have to prove we won't wait any single second for the render. So I gave up, if I make them buy new machines and they see me or one of my collegues wait for one second, they will start complaining...

They told me "don't try to fool us !!!" since then I do not take any responsability...

Simple though, we can go faster, for not that much money, but noway they will "tilt"...

Guess we have those 2003 xeon still for a looooooong time. 2 QX6700 would litterally doom the sort of renderfarm we have...

[dlparisi]

Re: Fighting for Workstation Upgrades

Our average production per artist over the last week was 7 projects per artist with an average of 45 hours per artist.

So they're spending like 6.5 hours a week on average per job? That doesn't seem too productive. But at 6 artists x 7 jobs = 42 jobs per week means you guys sound very, very busy (even if you don't get them all done in week). Even if new workstations only increase true productivity by 10-20% that's an extra 4-8 jobs which should equal several thousand dollars worth of work. Your boss should realize that machines are cheap (relatively), it's the staff that's expensive. I imagine your salary + overhead per week is approx. equal to the cost of a new workstation, probably more. I'm not talking dual quad core xeons for everyone but a basic technology upgrade (let's just say a dual core w/ 2gb ram) doesn't seem unreasonable. Retire the other machines to the render farm...

[dlparisi]

A good professional ALLROUND architectural CG-artist only needs very little time to do lighting and material setup, which are the only things that need the perfect workstation. If you're smart you can do more in POST then spending lots of GHz/Hour on doing the above. I did MODEL very large project in 3DStudio DOS with my pentium 50MHz, including 8K composition in PShop. Lots of people want the perfect output straight out of the rendersoft, but even for animations this surely is not neccessary.

So you're still working on this 50mhz pentium in 3D Studio ? Of course not, time is money and although you were able to produce renderings with that setup I'm sure it took longer than you would have liked. Otherwise why have you upgraded and why have you switched to Vray over scanline? Of course he doesn't need the latest and greatest hardware but an upgrade every three years or so is the cost of doing business and staying competitive in this industry.

[trick]

...and although you were able to produce renderings with that setup I'm sure it took longer than you would have liked...

I was illustrating modelling and postproduction on a lower end setup because if architectural VIZ (not animation) is your main work, this is a large part of your job. If time is crucial and you need to deliver on the same day renderspeed WILL be very important and then a faster setup will pay itself back within no time, but otherwise rendering can pass overnight.

[percydaman]

If the guy's happy with your current work and how quick you get it done, then hey, what are ya gonna do. But ask him if he's interested in getting work either 1) done quicker or 2) Better quality work. Ask him that and any answer he gives that isn't a yes, will make him feel like a moron.

[Geertvdp]

Although I agree that for the artist it can't go fast enough, one has to balance the cost and performance of a system with the projected workload and the competition (if any). In your case however, working on 3 year old equipment results in a huge performance gap. My personally rule is to upgrade a personal workstation when you can double your speed (cut rendertimes in half). Practically, that's about every 12 to 18 months...

The main objective of a company is to maximize profit. From there all calculations must start. Several directions you can go though:

-If you have more work coming that what you can do now, then it makes a lot of sense to buy new, faster hardware.

-if you have faster machines, you can take on new types of jobs or deliver even higher quality, maybe accept more animation work instead of doing still images?

-If you don't, maybe the competition is able to deliver faster then you, with your aging equipment? If that's the case you'll have to upgrade to stay competitive.

-If you do not have any more work and competition is nothing to worry about, buying faster machines would actually enable you to reduce your most important cost factor and fire.... 1 or 2 artists ...
Not exactly what you'd like to hear, no?


As suggested, it might be wise to phase in a few new machines instead of investing lots of money. That way you can adapt better to changing situations without going broke...

Currently quad xeon 5310 and 5320's or QX6700 systems are the best price/performance buys you can make for a renderfarm/workstation. (see http://www.chaosgroup.com/forum/phpB...7726&start=125 ).
Personally, I'm against buying the latest fastest thing because it doesn't have the best price/performance. (unless speed is extremely crucial and the client is prepared to pay for the additional cost)

[jujubee]

Simple - just tell the owner he would cut the workload nearly in half.