Desktop refresh best practice
-
@Carnival-Boy said:
@coliver said:
I know this may be an unpopular opinion... but why are you boosting the RAM to 8GB other then it is fairly inexpensive to do so? Do you have applications that will take advantage of this extra RAM? Do you often see issues with lack of RAM?
Not sure. I guess I'd have to test and find out. We do have quite a few users who use Autodesk Inventor View to view engineering drawings and that is fairly memory hungry, but apart from that, most users are just using Microsoft Office and IE. RAM is super cheap though, especially non-HP RAM.
Has any determination been made as to what is causing the old machines to run slowly?
-
No.
-
@Carnival-Boy said:
@coliver said:
I know this may be an unpopular opinion... but why are you boosting the RAM to 8GB other then it is fairly inexpensive to do so? Do you have applications that will take advantage of this extra RAM? Do you often see issues with lack of RAM?
Not sure. I guess I'd have to test and find out. We do have quite a few users who use Autodesk Inventor View to view engineering drawings and that is fairly memory hungry, but apart from that, most users are just using Microsoft Office and IE. RAM is super cheap though, especially non-HP RAM.
Ah, yes that is a hungry app. Although I have a 2GB machine (long story) running it in our shop right now without too much issue. I think you would be very surprised at the performance increases in just upgrading the HDD to an SSD.
@art_of_shred has a good point. If you can figure out what is causing that issue you may be able to get even more life out of the machine by updating that.
-
While we are a very long time away from when Windows 9x was living, I wonder if a simple OS refresh would help, not solve the issue for the moment.
It seemed to me at least with Windows 9x that it's disk use exploded over time, and that once a year I did a refresh on it to eliminate that bloat. This sped the machine back up to a nice performance level.
fast forward nearly 20 years
That isn't to say that this is the fix. It is at best a band-aid for a larger problem. Available funding is always a issue. I don't see how spending $200 in upgrades this year, and $500 in replacement next year is practical. The total cost of that system becomes $700 plus it's original purchase price. If you spent $700 and get 5 years out of them, why spend $700 for just two years - is that a good ROI?
And don't forget to factor your time to upgrade and refresh each one of those stations. How much of your time would be spent on the in place upgrade (End User Desk) or Bench side and the round trip from the desk to bench.
-
Not sure I follow your maths there. But let's say a new PC cost $500. If I'm planning on retiring PCs after 5 years, that's a capitalised cost of $100 per year. So if I can spend an extra $200 and get an extra 2 years then I have broken even: my PC costs remain $100 per year.
-
@Carnival-Boy said:
Not sure I follow your maths there. But let's say a new PC cost $500. If I'm planning on retiring PCs after 5 years, that's a capitalized cost of $100 per year. So if I can spend an extra $200 and get an extra 2 years then I have broken even: my PC costs remain $100 per year.
That is true - But what about your time? Yes the physical cost is still just $100 a year, but what about YOUR cost? Time spent on installing those upgrades?
Original Cost: $500
Device Life cycle: 5 years
Upgrade cost: $200
Upgrade time: 1.5 hours
Upgrade time Cost: 1.5 x $100hr
Additional Cost: $150
Total Upgrade: $350
Two year Life Cycle cost: $175/yrOf course - in the Non Profit arena I have been in the last nine years, that two year extension normally works out to be nearly four years.
-
@g.jacobse said:
...that two year extension normally works out to be nearly four years.
This is where I was going - two years was a bare expected minimum for additional life. I would fully expect most of those machines to give another four years or more.
-
@Dashrender said:
I would fully expect most of those machines to give another four years or more.
Agreed.
-
I would just replace them. The ram is going to be a bottle neck and maybe even the CPU as its fairly low end. I'd just buy standard machines all with 8GB of ram, and a core i5 for future proof.
One you count their age, the cost of upgrades and your cost for doing the upgrades it is not worth it, especially on these very low end systems.
-
Also as far as Windows 10, most people on spiceworks work in the SMB market and are very scared of technology. There's no reason to just stay on windows 7 for another year unless you have to. We plan on having all 8,000 of ours on windows 10 in 6 months.
-
@Carnival-Boy said:
Firstly, as far as the OS is concerned, I'm considering rolling out Windows 10 company wide fairly soon. I note on Spiceworks that the consensus seems to be to stick with Windows 7 for at least another year. I can't quite understand this as 7 will be three versions behind and is over 5 years old. Also, I'm not sure if HP will continue shipping Windows 7 after July?
I don't agree with the SW consensus at all and often point to their reaction in this matter as a point of why most IT people completely misunderstand software maturity. Windows 10 is essentially a "patch" to Windows 7, they are the same OS (Vista is the base of the family of which they are all members.) Windows 10, even as an upgrade patch, is heavily tested already (we moved to it six months ago!!) Windows 7 isn't just old in general (really, really old by IT standards, finding Windows 7 in business use today should be just slightly surprising, seven years old is ancient in IT terms, few OSes have ever been considered even slightly acceptable at that age, XP being one of them) but it is a very immature member of the family. Windows 10 is the third version update to the Windows Vista family post Windows 7.
Other than issues like you mentioned with compatibility for XP mode, I can't fathom any logic of not going to Windows 10. Both previous versions of Windows in between were both faster and more stable than Windows 7. And, of course, new security measures being included and, not to mention, new features that may or may not be useful to your users.
Windows 10 is expected to be even yet more stable than its predecessors, faster and lighter and have even more features. A great way to get more life out of your older systems.
-
Well one reason to not upgrade is the manpower it takes to do the actual upgrade. But beyond application compatibility and the upgrade manpower (and associated costs) I agree, I see no reason not to upgrade.
Though I suppose the usual end user training might be another reason not to. Give users a chance to get it at home first.. learn on their own dime/time. Then after they are more familiar with it, then roll it out to the office environment.
That said I'm with @thecreativeone91 I hope management signs off on us upgrading before the end of the year.
-
@Carnival-Boy said:
They were fairly low-spec at the time - I believe most are HP Pro 3130's with Pentium G6950 2.8 GHz CPU and 3 GB RAM. All are Windows 7. They're pretty slow now. Some of them are performing like dogs for whatever reason.
Are they slower than they used to be? Or are they just starting to show up as being slower because people are doing heavier things?
Moving to Windows 10 and doing a fresh install should both help, a little.
Small things, like adding new RAM could go a really long way for cheap. What we did was go to SSDs on even older machines and the leap in performance was unreal. For under $150 per machine, maybe under $100, you might be able to up the memory and move to SSD, do the re-image as Windows 10 and see the machines easily double, or more, in end user delivered efficiency. And get many years of use yet out of the existing investment. Maybe three to four additional years.
Testing this on one or two machines would be cheap and easy.
-
@Carnival-Boy said:
Secondly, as far as the hardware is concerned, I could upgrade the existing PCs with 8G RAM and SSDs for around $220. That, combined with re-installing the OS and moving to Windows 10 should result in drastically improved performance and they could be good for another few years. They'll still be using the same old Pentium processors though.
I see you already thought of all of that
Generally the CPU is a tiny factor in overall performance for non-gaming systems. If they are manipulating massive spreadsheets or doing something really intensive there can be exceptions, but mostly CPUs are one of the least important factors in performance for desktops today so that the CPUs are old might matter almost not at all.
-
@Carnival-Boy said:
Or I could buy new PCs. I like the look of the sexy new HP mini PCs for around $400 plus another $50 or so to increase the RAM from the measly 4GB that HP still ship as standard. I'm guessing they should perform better than our existing PCs.
What would the specs of the "new" machines be? $450 is more than double the cost of the upgrade, but with additional support might be worth it. But you would be losing the tried and true reliability of the models you have today, so not a complete win. And double the money is still, double the money.
-
@Carnival-Boy said:
Also, I'm interviewing at the moment for a new IT person and one of the interview questions is to find out what he or she would do, since this will be one of their first projects. But I don't know what is a good or bad answer to the question
- so I'm looking for feedback.Keep in mind that this is primarily a tech-driven business decision. There is more business in this decision than there is tech. So if you are looking to hire and IT decision maker, business skills outweight tech ones. But the best pure tech will often (as SW shows consistently) have no concept of how IT fits into the business context and will easily be lost and have no idea what decisions are wise ones or that favour the business.
So remember that IT as two big aspects - the IT people who make decisions and have to understand the business inside and out to do so, and the "button pushers" who can be extremely technical and incredibly good at what they do but are head down in technology and need the decision makers overseeing them or else they have no context in which to put their work and no ability to make reasonable recommendations, let alone decisions.
A good tech should be able to come up with some options. But understanding pricing, support value, financial factors, opex, capex, etc. are skills that the pure techs rarely have. So just be aware that this question is asking for a fair but of business understanding, not just technical understanding.
-
I received an email from Tiger Direct this morning for $69 256 GB SSD (OCZ). You should be able to pickup a 128 for $50 or less and I would think the RAM would be around $100, so well below the $220 you mentioned.
-
@Carnival-Boy said:
tl;dr: When are you rolling out Windows 10? Why wouldn't you roll it out asap*? Is it better to replace PCs or upgrade them with new components?
I tend to be the opposite of most technical people here, partially because I was raised on the business side of the house, but I almost always lean towards rapid software updates and slow hardware updates. Software is where you get your security, stability and features and the costs of upgrading are generally extremely low. Hardware gets you most of your speed, but speed is rarely a primary issue today and is mostly not a business driver. Hardware is what costs the most. So saving money on hardware and spending it on software often gets you, by far, the most bang for your buck not just in good performing, easy to use end user systems but also in security, manageability and other soft factors that are extremely important to a business.
-
@Dashrender said:
I received an email from Tiger Direct this morning for $69 256 GB SSD (OCZ). You should be able to pickup a 128 for $50 or less and I would think the RAM would be around $100, so well below the $220 you mentioned.
Sadly, UK prices are closer to Canadian than to American.
-
@scottalanmiller said:
@Dashrender said:
I received an email from Tiger Direct this morning for $69 256 GB SSD (OCZ). You should be able to pickup a 128 for $50 or less and I would think the RAM would be around $100, so well below the $220 you mentioned.
Sadly, UK prices are closer to Canadian than to American.
Aww.. I didn't put that together - you're right of course. Makes me wonder where he's getting a $400 machine then? and what kind of specs it has? Heck Chromebooks sometimes cost this much.
