Boxx Technologies 3DBOXX 4860 Workstation

  • Thread starter Thread starter Guest
  • Start date Start date
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
[citation][nom]mayankleoboy1[/nom]though if i were to take each component separately and build our own system, it would be cheaper.[/citation]
You forget, the cost of a commercial PC includes service, support, and licensed certifications.
 
I wouldn't purchase this workstation. First you can build a better base machine for the fraction of the cost. Second you can purchase on your own the software you require or pirate. Third there is a flaw, yes there is always the temptation of mounting the hard drives in that manor but isn't recommended due to the uneven wear on the spindle that can lead to early failure.
 
For this price, I'd go with a workstation from a major player (ex. HP or similar). You could easily build a dual socket workstation with similar (or better) overall performance; remember that many apps that require this level of hardware are optimized for Xeon instruction sets and 8+ threads. Additionally, you're software vendors would actually support their products on a system running within spec. Simply put, this is a toy not an enterprise class product
 
[citation][nom]utengineer[/nom]You forget, the cost of a commercial PC includes service, support, and licensed certifications.[/citation]
Licensed certifications may be confidence inspiring to some, however, I think they are a waste of money. It is just a different form of branding that can be marketed at what is usually an expensive premium. Think THX certification. It was expensive in consumer audio and video, however, in my opinion, it has had it's 15-minutes of fame.
 
I found the conclusions about value entirely reasonable. You guys being critical aren't the market for this box, or the services behind it. Someone whose job isn't messing around with workstation hardware and getting specific software configured and working on it may be very happy to pay someone to do those things.
 
any decent modder can build this at less then half the cost and tweak it further. Not any value to this but for business people who don't know the difference between socket 1366 and widget 1998
 
All you complaining about prices talk your broke a## home, its a great build and did you see they make their own chasis how cool is that.
 
I found the conclusions about value entirely reasonable. You guys being critical aren't the market for this box, or the services behind it. Someone whose job isn't messing around with workstation hardware and getting specific software configured and working on it may be very happy to pay someone to do those things.
Click to expand...

I agree with this, and it also points out something else. For all of us who build custom systems for ourselves and our friends / family, this type of product provides a data point for the value of that work. your time is worth money - don't ever forget that. So if you built a custom workstation w/ comparable parts and OCed to 4.2 GHz and supported it for 3 years, then you are providing roughly $4k of value to the recipient.
 
I think overclocking is almost essential in a workstation. Performance is everything when you have expensive employees waiting on the computer. We run AutoCAD and Revit in my little company, and our manager enforces a very tight budget on us. I built our last 3 CAD workstations myself. I used i5-750 processors, 8GB ram, WD velociraptor drives, FireGL video cards, Antec cases, Win7-64bit, and a good Xigmatek CPU cooler. I also added 2 more fans to the 2 that the Antec cases already had. When I built these over a year ago I wasn't the best overclocker but they would run at 3.9 and pass my stress tests. We run them 24/7 and I set them at 3.5GHz for that. We don't have any temperature problems, the CPU's run very cool. For our software the video card isn't that important, the lowest FireGL will run AutoCAD just as fast as the most expensive one. The reason to even buy the FireGL is to get better drivers that will work with the 64bit OS. I've tried with other brands of cards and they weren't reliable. The only problems we have had with any of these machines has been with the cheaper video card when the FireGL that we wanted to buy was not available. I believe we spent about $1200 each on these machines, I spent about 2 hours each at home building and testing them.

At the time that I built these, the i5-750 system offered a $300 savings over an i7-920 system. Today it might be a different story, and with the i7-SB we might run them at 4.0GHz.
 
[citation][nom]malnute[/nom]...its a great build and did you see they make their own chasis how cool is that.[/citation]

Thanks for pointing that out. The case looks like crap.
For $8000.00 case looks cheap and dull.
 
I've got an older Boxx machine... their systems are top shelf. Sure they're expensive (I paid 10k for mine) but they're not intended for consumer use but for high end business workstation use.

I'd buy one again, quality, fit and finish, component selection... all as good as you can get.
 
interesting article, even more interesting is i have worked extensively with at least 3 of the named blue chip companies mentioned in the intro and i have yet to come across one of these beast, and i have worked in departments which are given the high performance gear. i would be interested in how the companies planned to utilize such raw power

in reality a 100% boost in performance does not equate to a 100% boost in productivity, you are still limited by the user and his work flow, a faster machine does not mean you can reduce head count, dude A is still going have to finish his job (all be it slightly faster) and dude B still has to finish his job, dude A is not going be doing his and dude B's job just cause he got a faster machine

where this kind of power becomes useful is if you got big numbers to crunch (3d rendering, FEA) and even then you'll be better served by a high density setup like a blade server. The only real use for such a machine would be for bespoke jobs whereby a person might need to spot check some work his been doing, but that's not a daily occurrence and you dont give everyone a machine like this just in the off chance they need to spot check their work once a while, in reality this would be a single shared machine that would not be creating production data, in which case an overclocked machine makes sense
 
Sure you can build it, but this is for professionals who don't have time to build systems and would rather buy the product and the support the comes with it. Boxx is a very good company.

On a sidenote I like their case/chasis. Good performance in the end.
 
Their chassis is all metal, layout is great, capacity is great. It includes things you don't get in low end cases like dust filters on the fans and room for redundant PS setups.

Mine has been running 24/7 nearly since I got it (however old Opteron 250's are), maybe 6 or 7 years. It still boots faster than non-SSD systems today (15k SCSI RAID, stupid fast in its day and still very respectable).
 
Ok, my comment got lost on log in and I could not recover it (my fault), but I am going to write again what I think was the essence of my comment.

If you are responsible for a relatively small DCC studio think on these systems as resources and not as complete solutions for every man. It doesn't make sense for every business model to invest on the same systems and in the same way. You may need a specific mighty overclocked machine in relative or even absolute isolation for specific qualified software and for specific problems on your pipeline. If you are dealing with great complexity in a model, set up, animation or compositing you may need all the power you can get and not lose on reliability 24/7. Besides if you don't have the time to build and test a system and you have the resources expend the money on solid machines like the ones Boxx provide. Think on these powerful systems as maybe one for every 5 or 10. Or even in one or 2 for a specific project.

Sometimes you really depend on systems like these for heavy work you can't complete on less powerful systems no matter how much you try or how many hours you dedicate, but for early stages on your pipeline even modest desktops will be enough. And how many extreme workstations and how much performance do you need on each case depends entirely on your experience, budget and specific project. Don't think that an entry level work, even for a very experienced man requires something close to a overclocked dual 6 core Xeon. It wont make them substantially more productive.

I builded a 2600K/Asus/Quadro/Dell Pro monitor in January and the system is very fast and running fine. But to be completely honest is not as reliable as I wanted. And blame me for that, but if I had the resources I would order a Boxx system with only the specific software installed I need and have more peace of mind to be more creative.

Anyway thanks at Tom's for articles like tis one and I am going to read it carefully a second time.
 
What people are also forgetting is the value of the time in building your own box. Did ya read the part where they paying $1500 a week....that's cheap. $50 an hour is an average salary and that means a billing rate of $150 an hour. I build my boxes for my business because I enjoy it, not because it's cost effective. Parts research, ordering, unpacking, building, testing, loading OS and software, overclocking, tweaking for a person who only pays attention to this every 4-5 years is gonna eat much of those self build savings in lost billings.

That being said, they are being a bit greedy with what they are charging.
 
Still can call it their own design which most builders can't there has to be some value in that or did you want yours in pink d.kuhn
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom