Upgrade often, but cheap, or buy big for the long haul?

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

VirtualLarry

No Lifer
Aug 25, 2001
56,453
10,120
126
I find that frequent budget upgrades are good, so long as you don't fall beneath the "too cheap to be useful" threshold.

You mean, like $100 and $125 Foxconn NanoPCs with C-70 1.0Ghz APUs that overheat and corrupt installed SSDs? Or some $86 Ubuntu Linux MeegoPad T02s that throttle down during ordinary web browsing or Skype so much that they are practically unusable? Even though 7" Win8.1 tablets using the same reference platform and CPU, can Skype for as long as the battery lasts, without overheating? Or my Asus N2830 laptops that can't play 1080P YouTube completely smoothly? (CPU usage is like 15-20%, then it spikes up to 95% every few seconds. No apparent reason why.)

My most recent folly, was building some 3.7/3.9Ghz FM2 A4-6300 rigs. 8GB RAM, 120GB SSD, nice case. Then I went looking for benchmarks, and found some at CPU-world, that showed that they were worse in MT performance than a lowly 2.5Ghz E5200 or E3300 C2D.

Now I'm debating getting some A8-7600 Kaveri APUs for $90 and putting them in. At least then the PC could game, somewhat, but then I'd have to put Windows on instead of Linux, raising the cost of the PC by $150.
 

waltchan

Senior member
Feb 27, 2015
846
8
81
I always believe in CPU for long term, meaning buy the best and it will last you a while.
If anyone has a Fry's Electronics nearby, i3-4170 frequent-sale for $69 has the lowest price / highest performance ratio you'll find in any LGA1150 processor.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
Just ask those people who bought a Phenom II X4s in 2011 instead of a 2500K to save a mere $100 at the start. Or AM2 K8s instead of Conroe just because the Intel mobos were more expensive. Who is the fool now?
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
My most recent folly, was building some 3.7/3.9Ghz FM2 A4-6300 rigs. 8GB RAM, 120GB SSD, nice case. Then I went looking for benchmarks, and found some at CPU-world, that showed that they were worse in MT performance than a lowly 2.5Ghz E5200 or E3300 C2D.

Here is the side by side of A4-6300 vs. E5200 at CPU World:

http://www.cpu-world.com/Compare/153/AMD_A4-Series_A4-6300_vs_Intel_Pentium_Dual-Core_E5200.html\

And the specific benchmarks used to make that comparison here:

http://www.cpu-world.com/Compare/15...0_vs_Intel_Pentium_Dual-Core_E5200.html#bench

Interestingly, the passmark scores favor the A6-6300 over the E5200 by almost 50%:

http://www.cpubenchmark.net/cpu.php?cpu=AMD+A4-6300+APU (A4-6300 = 2208 CPU marks)

http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+E5200+@+2.50GHz (E5200= 1494 CPU marks )

With these conflicting results in mind, I do wonder which one would be better at gaming? (When the same dGPU is used)

Therefore, I will investigate this at some point using E6550 vs. A6-5400K (which are fairly close to A4-6300 and E5200) using a game or two....with GT730 GDDR5 as the dGPU for both set-ups.

However, with that mentioned for a Linux gaming build I'll bet there is more room for using better hardware. (But I totally get the idea that Linux gaming builds should be cheap! Otherwise, the user risks excessively limiting the capability of the hardware by not spending money on the OS.)

P.S. Regarding the whole Linux vs. Windows thing, I really hope we see some better hardware from Intel and AMD at the lower price points. This to make the low cost Linux gaming builds more effective. I'm not holding my breath....but a BCLK overclockable 2C/4T Pentium and a Low end Maxwell GPU (maybe $50 sale priced GTX 750...or whatever gets release by then) could probably beat a Xbox One in a number of games.
 

tential

Diamond Member
May 13, 2008
7,355
642
121
You mean, like $100 and $125 Foxconn NanoPCs with C-70 1.0Ghz APUs that overheat and corrupt installed SSDs? Or some $86 Ubuntu Linux MeegoPad T02s that throttle down during ordinary web browsing or Skype so much that they are practically unusable? Even though 7" Win8.1 tablets using the same reference platform and CPU, can Skype for as long as the battery lasts, without overheating? Or my Asus N2830 laptops that can't play 1080P YouTube completely smoothly? (CPU usage is like 15-20%, then it spikes up to 95% every few seconds. No apparent reason why.)

My most recent folly, was building some 3.7/3.9Ghz FM2 A4-6300 rigs. 8GB RAM, 120GB SSD, nice case. Then I went looking for benchmarks, and found some at CPU-world, that showed that they were worse in MT performance than a lowly 2.5Ghz E5200 or E3300 C2D.

Now I'm debating getting some A8-7600 Kaveri APUs for $90 and putting them in. At least then the PC could game, somewhat, but then I'd have to put Windows on instead of Linux, raising the cost of the PC by $150.

you could have bought a 150 i5 at this point and saved yourself the trouble...
But it's funny seeingi you jump through every single low end system out there first.

What you should do is write a blog about your experience/testing. That would get a LOT of traffic and you could probably help people who want low end systems that while they don't meet your standards, will meet theirs.

Lol though VL, I can't believe you're still powering through. Moving your way slowly up the chain it's been an amusing ride.
 

VirtualLarry

No Lifer
Aug 25, 2001
56,453
10,120
126
Interestingly, the passmark scores favor the A6-6300 over the E5200 by almost 50%:

http://www.cpubenchmark.net/cpu.php?cpu=AMD+A4-6300+APU (A4-6300 = 2208 CPU marks)

http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+E5200+@+2.50GHz (E5200= 1494 CPU marks )
Thanks for looking that up. Now I don't feel so bad. In Passmark, A4-6300 is 30% faster in ST, and maybe 50-60% faster in MT.

But I totally get the idea that Linux gaming builds should be cheap!

Yeah. These were intended to be cheap internet builds (with Linux by default, Windows for extra $$$ or BYOW). Not necessarily gaming-oriented. But since they were APUs, I wondered what the gaming capability (under Windows) was. I'm sure it would be a lot better if I dropped in Kaveri.

Edit: With the Apex / Allied 300W case PSU, in Linux Mint 17.2, idle power measured with a KaW is 36W. Opening up Firefox and scrolling, brought it to 71W.
 
Last edited:

cbn

Lifer
Mar 27, 2009
12,968
221
106
Not necessarily gaming-oriented. But since they were APUs, I wondered what the gaming capability (under Windows) was. I'm sure it would be a lot better if I dropped in Kaveri.

My A6-5400K (which has 64 more VLIW steam processors than the A4-6300) would play Team Fortress 2 (in Windows) at over 60 FPS on 1080p mixed settings (with dual channel RAM). With single channel RAM it was slower, but still much faster than the Athlon 5350 at the same game.

So not bad really. It would probably also handle the rest of the Valve games like CS:GO fairly well at lower resolution (in Windows). Linux performance should be lower though, but I will check again soon.

P.S. My Dell Optiplex 755 is already set-up with the E6550 and GT 730 GDDR5 so I will run a quick benchmark in Windows (maybe Dirt 3) before removing the GT730 GDDR5 to do my testing with A6-5400K....which will later be used for a SteamOS install (my first time using/testing SteamOS).
 

escrow4

Diamond Member
Feb 4, 2013
3,339
122
106
You mean, like $100 and $125 Foxconn NanoPCs with C-70 1.0Ghz APUs that overheat and corrupt installed SSDs? Or some $86 Ubuntu Linux MeegoPad T02s that throttle down during ordinary web browsing or Skype so much that they are practically unusable? Even though 7" Win8.1 tablets using the same reference platform and CPU, can Skype for as long as the battery lasts, without overheating? Or my Asus N2830 laptops that can't play 1080P YouTube completely smoothly? (CPU usage is like 15-20%, then it spikes up to 95% every few seconds. No apparent reason why.)

My most recent folly, was building some 3.7/3.9Ghz FM2 A4-6300 rigs. 8GB RAM, 120GB SSD, nice case. Then I went looking for benchmarks, and found some at CPU-world, that showed that they were worse in MT performance than a lowly 2.5Ghz E5200 or E3300 C2D.

Now I'm debating getting some A8-7600 Kaveri APUs for $90 and putting them in. At least then the PC could game, somewhat, but then I'd have to put Windows on instead of Linux, raising the cost of the PC by $150.

For an office box that aims to last you want at least 5000 marks. That will give you speed and breathing room. That is also an i3. APUs make no sense unless you really really want a cheap all rounder as opposed to an i3 and dGPU.
 

VirtualLarry

No Lifer
Aug 25, 2001
56,453
10,120
126
For an office box that aims to last you want at least 5000 marks. That will give you speed and breathing room. That is also an i3. APUs make no sense unless you really really want a cheap all rounder as opposed to an i3 and dGPU.

Have you looked at new dGPU prices lately? Any decent ones are $100+.

I remember people arguing the cost/benefit analysis of AMD APUs, and how it was cheaper to get an Intel CPU + dGPU, or even an IGP-less FM2/FM2+ CPU, and a dGPU.

Well, those cheaper dGPUs have pretty-much dried up, making APUs a significantly cheaper proposition, for a budget gaming rig.

Can't wait to see what HBM will do for APUs. Just another year or two to wait...
 

VirtualLarry

No Lifer
Aug 25, 2001
56,453
10,120
126
I guess I should clarify some of my thoughts from the O.P.

When I was suggesting whether to buy big for the long haul, the question was more along the line of, "even if you don't need the performance today".

Say, if you only needed an i3, should you still invest in, say, a HEDT hex-core, with the thought that if you were to keep it for Y years, that it would only cost you X per year, versus, say, buying a new platform with a matching i3, every two generations?

It's kind of a trade-off, really, because in the former case, you get the CPU power for longer, but miss out on platform features, like storage bandwidth increases (which, depending on your compute requirements, can be just as important as raw CPU grunt). If you buy small or medium on the CPU, but upgrade platforms often (or whenever there is a major change to the storage subsystem), then you get more benefits from a platform change, but never really get to have a high-end compute experience with the CPU.
 

Yuriman

Diamond Member
Jun 25, 2004
5,530
141
106
One or two generations don't usually bring major, earth-shattering platform changes, and an i3 really isn't that much more expensive compared with a Celeron, if you look at total system cost. You're right in identifying other inherent improvements in the platform, which suggests the ideal is somewhere in the middle. Low-end chips don't offer a good value in the long term, but neither do high end ones.
 

DrMrLordX

Lifer
Apr 27, 2000
21,813
11,168
136
You mean, like $100 and $125 Foxconn NanoPCs with C-70 1.0Ghz APUs that overheat and corrupt installed SSDs? Or some $86 Ubuntu Linux MeegoPad T02s that throttle down during ordinary web browsing or Skype so much that they are practically unusable?

Yeah. Or stuff that just has a high failure-rate. Cheap HDDs, cheap PSUs, stuff like that. Penny-pinching in those circumstances can force you to replace hardware frequently, which is a path to disaster. Stuff that lasts can save you money in the long run.

Now I'm debating getting some A8-7600 Kaveri APUs for $90 and putting them in. At least then the PC could game, somewhat, but then I'd have to put Windows on instead of Linux, raising the cost of the PC by $150.

I would look at the 7650k as well. You can get some sweet deals on it. If you are paying $90 for the 7600 (why?) then the 7650k is only a few dollars more:

http://pcpartpicker.com/parts/cpu/#m=6&k=26&sort=a7&page=1

Granted, the k isn't much better of a CPU. It just gives you more flexibility.

Have you looked at new dGPU prices lately? Any decent ones are $100+.

I remember people arguing the cost/benefit analysis of AMD APUs, and how it was cheaper to get an Intel CPU + dGPU, or even an IGP-less FM2/FM2+ CPU, and a dGPU.

Well, those cheaper dGPUs have pretty-much dried up, making APUs a significantly cheaper proposition, for a budget gaming rig.

Can't wait to see what HBM will do for APUs. Just another year or two to wait...

I agree here. APUs can still be a hard(er) sell thanks to the memory requirements, so you might still be able to scrape by with an x4-860 + 8 Gb DDR3-1600 + $100 dGPU and do better than you would with, say, a 7870k + 8 Gb DDR3-2400, but the margin is narrowing. RAM prices are also narrowing down to the point where you can get DDR3-1600 kits for $40 and DDR3-2400 kits for $50 (2x4Gb each). But cheap-ish 250/250x video cards are getting a little more expensive, and they may just go away once retailers clear out inventories:

http://pcpartpicker.com/parts/video-card/#c=150,161&sort=a8&page=1
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
APUs can still be a hard(er) sell thanks to the memory requirements, so you might still be able to scrape by with an x4-860 + 8 Gb DDR3-1600 + $100 dGPU and do better than you would with, say, a 7870k + 8 Gb DDR3-2400, but the margin is narrowing. RAM prices are also narrowing down to the point where you can get DDR3-1600 kits for $40 and DDR3-2400 kits for $50 (2x4Gb each).

I've seen the A8-7650K + A68 (uATX) motherboard on sale at Fry's for $89.99 after $10 rebate with the Athlon x4 860K + A68 (uATX) motherboard (also at Fry's) for $79.99 after $10 rebate.

So $10 difference.

As far as RAM costs go, I usually only see DDR3 1600 on sale (as low as $35 for 2 x 4GB)....but this $40 2 x 4GB DDR3 2133 for $40 looks pretty good --> http://www.newegg.com/Product/Product.aspx?Item=N82E16820178772

With that mentioned, the Athlon x4 860K has higher clocks, more overclocking headroom under the stock HSF and won't throttle under graphics load in Windows.

Still $10 plus $5 premium for 2133 RAM (under ideal sale conditions) is a much narrower margin than I remember in the past.

But cheap-ish 250/250x video cards are getting a little more expensive, and they may just go away once retailers clear out inventories:

http://pcpartpicker.com/parts/video-card/#c=150,161&sort=a8&page=1

During sales, GTX 750 1GB has moved into the $50 (after rebate) range. (And actually this is even cheaper than the old R7 250X (after rebate) sale price).
 
Last edited:

cytg111

Lifer
Mar 17, 2008
23,561
13,122
136
Wondering if anyone has done a study on this yet.

Would it be better to buy an i3-level CPU every few generations, or buy the biggest, fastest chip available, and hold on to them as long as possible?

Has anyone actually run the TCO numbers?

You are adressing a moving target, it is not that simple, sweet spot today is 4 cores where 2 cores / 4 threads cut it yesterday but not today. Not too long ago single threaded performance was undisputed king, today its still king but not undisputed.

If you're going for value at the upper end of the spectrum i'd have to place my bet on i5's .. quads still ruling and their clocks get the job done.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Here is the side by side of A4-6300 vs. E5200 at CPU World:

http://www.cpu-world.com/Compare/153/AMD_A4-Series_A4-6300_vs_Intel_Pentium_Dual-Core_E5200.html\

And the specific benchmarks used to make that comparison here:

http://www.cpu-world.com/Compare/15...0_vs_Intel_Pentium_Dual-Core_E5200.html#bench

Interestingly, the passmark scores favor the A6-6300 over the E5200 by almost 50%:

http://www.cpubenchmark.net/cpu.php?cpu=AMD+A4-6300+APU (A4-6300 = 2208 CPU marks)

http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+E5200+@+2.50GHz (E5200= 1494 CPU marks )

With these conflicting results in mind, I do wonder which one would be better at gaming? (When the same dGPU is used)

Therefore, I will investigate this at some point using E6550 vs. A6-5400K (which are fairly close to A4-6300 and E5200) using a game or two....with GT730 GDDR5 as the dGPU for both set-ups.

EDIT: I have to re-run the A6-5400K results (shown below) because I noticed the CPU multiplier was on "30" on the BIOS. I restored optimized defaults and further testing will be done with the processor at the full 3.6 Ghz base clock and 3.8 Ghz turbo.

Benchmark results (using built in benchmark for both games, average of three runs):


War Thunder @ 1080p High (Eastern Front Benchmark)

A6-5400K, 4GB DDR3 1866 (single channel), GT730 GDDR5 = 58.7 avg. FPS, 32 min FPS

E6550, 4GB DDR2 667 (dual channel), GT730 GDDR5 = 47 avg. FPS, 30.2 min FPS


Dirt 3 @ 1080p ultra low, multisampling off

A6-5400K, 4GB DDR3 1866 (single channel), GT730 GDDR5 = 65.9 avg. FPS, 52.32 min FPS

E6550, 4GB DDR2 667 (dual channel), GT730 GDDR5 = 75.9 avg. FPS, 50.66 min FPS

P.S. Operating System for both set-ups was Windows 8.1 Pro (fresh install) with the latest Nvidia driver.
 
Last edited:

ninaholic37

Golden Member
Apr 13, 2012
1,883
31
91
Where's the "buy something old for $40 and never upgrade until your computer dies" option?
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Here is the side by side of A4-6300 vs. E5200 at CPU World:

http://www.cpu-world.com/Compare/153/AMD_A4-Series_A4-6300_vs_Intel_Pentium_Dual-Core_E5200.html\

And the specific benchmarks used to make that comparison here:

http://www.cpu-world.com/Compare/15...0_vs_Intel_Pentium_Dual-Core_E5200.html#bench

Interestingly, the passmark scores favor the A6-6300 over the E5200 by almost 50%:

http://www.cpubenchmark.net/cpu.php?cpu=AMD+A4-6300+APU (A4-6300 = 2208 CPU marks)

http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+E5200+@+2.50GHz (E5200= 1494 CPU marks )

With these conflicting results in mind, I do wonder which one would be better at gaming? (When the same dGPU is used)

Therefore, I will investigate this at some point using E6550 vs. A6-5400K (which are fairly close to A4-6300 and E5200) using a game or two....with GT730 GDDR5 as the dGPU for both set-ups.

Benchmark results (using built in benchmark for both games, average of three runs):


War Thunder @ 1080p High (Eastern Front Benchmark):

A6-5400K, 4GB DDR3 1866 (single channel), GT730 GDDR5 = 55.9 avg. FPS, 35.7 min FPS

E6550, 4GB DDR2 667 (dual channel), GT730 GDDR5 = 47 avg. FPS, 30.2 min FPS


Dirt 3 @ 1080p ultra low, multisampling off:

A6-5400K, 4GB DDR3 1866 (single channel), GT730 GDDR5 = 73.9 avg. FPS, 58.81 min FPS

E6550, 4GB DDR2 667 (dual channel), GT730 GDDR5 = 75.9 avg. FPS, 50.66 min FPS


A6-5400K is 19% faster (in Avg. FPS) vs. E6550 in War Thunder. E6550 is 3% faster (in Avg. FPS) vs. A6-5400K in Dirt 3.

A6-5400K CPU was faster in min FPS for both benchmarks.


Operating System for both set-ups was Windows 8.1 Pro (fresh install) with the latest Nvidia driver.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
My A6-5400K (which has 64 more VLIW steam processors than the A4-6300) would play Team Fortress 2 (in Windows) at over 60 FPS on 1080p mixed settings (with dual channel RAM). With single channel RAM it was slower, but still much faster than the Athlon 5350 at the same game.

So not bad really. It would probably also handle the rest of the Valve games like CS:GO fairly well at lower resolution (in Windows). Linux performance should be lower though, but I will check again soon.

I just finished informal testing of CS:GO With Linux Mint 17.2 and the proprietary AMD driver. The game appears to average around 45 FPS (estimated) @ 1280 x 1024 low using the A6-5400K with 4GB stick of DDR3 1866 (single channel).

P.S. CS:GO is IMO the most demanding game (on hardware) Valve has released so far. The other Valve games should have equal or better performance.
 
Last edited:

HeXen

Diamond Member
Dec 13, 2009
7,832
38
91
I still tend to upgrade after a couple or so years despite buying a high end CPU or GPU for that matter. Mostly because there is something else I want out of it, such as a move to ITX which may have a different socket...stuff like that.

Heck I'll spend big just to have a computer that looks different or has different features...etc. I get bored with stuff too easy plus I often just get in the mood to want something new to play with.
I'm sure many others here are like that as well.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
I just finished informal testing of CS:GO With Linux Mint 17.2 and the proprietary AMD driver. The game appears to average around 45 FPS (estimated) @ 1280 x 1024 low using the A6-5400K with 4GB stick of DDR3 1866 (single channel).

P.S. CS:GO is IMO the most demanding game (on hardware) Valve has released so far. The other Valve games should have equal or better performance.

Increased resolution on CS:GO from 1280 x 1024 low to 1920 x 1080 low, game play stayed smooth and I would estimate average FPS somewhere between 35 and 40.

I also played Borderlands 2 using the same set-up. Through the first few levels average FPS was over 30 (smooth) at a resolution of 1280 x 1024 low.
 

skipsneeky2

Diamond Member
May 21, 2011
5,035
1
71
Where's the "buy something old for $40 and never upgrade until your computer dies" option?

Just ordered a Q6600 for this old Dell Inspiron 518 which has 6gb of ram and a ssd.Q6600 is replacing a E5200.Outside of gaming,it does as much as any other cpu for me.

Biggest question is what gpu to get for this thing,anything that requires a 6 pin is out of the question on this cheap psu.Anything outside of a gt720 most likely would be a upgrade over a ddr2 9500.
 

richierich1212

Platinum Member
Jul 5, 2002
2,741
360
126
Just ordered a Q6600 for this old Dell Inspiron 518 which has 6gb of ram and a ssd.Q6600 is replacing a E5200.Outside of gaming,it does as much as any other cpu for me.

Biggest question is what gpu to get for this thing,anything that requires a 6 pin is out of the question on this cheap psu.Anything outside of a gt720 most likely would be a upgrade over a ddr2 9500.

Easy, grab a 750ti.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |