Overclock.net › Forums › Industry News › Hardware News › [Legit Reviews] CPU scaling with HD 7950
New Posts  All Forums:Forum Nav:

[Legit Reviews] CPU scaling with HD 7950 - Page 4

post #31 of 45
CPUs are mostly worthless for single GPU gaming at high resolutions + max settings. You'll almost always be GPU bottlenecked, period. People who don't understand this will try to link benchmarks at 800x600, or using SLI/CF setups, or using old/subpar games that aren't hardware taxing. That proves nothing.

Of course some CPUs are faster than others, that's what synthetic benchmarks are for. For single GPU gaming at 1080+, it's largely irrelevant.
Ivypascal
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel 3570K, 4.6 GHz @ 1.29V Gigabyte GA-Z77X-UD5H eVGA GTX1080 ACX SC 16GB G.Skill Ripjaws X DDR3 1600 
Hard DriveOSMonitorKeyboard
Crucial M4SSD2 SATA III (128GB) Windows 10 Pro x64 Viewsonic XG2703-GS & Dell 24' 07WFPHC CM Storm Quick Fire Rapid, MX Red 
PowerCaseMouse
Corsair HX850 Coolermaster HAF 932 Logitech G502 
  hide details  
Reply
Ivypascal
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel 3570K, 4.6 GHz @ 1.29V Gigabyte GA-Z77X-UD5H eVGA GTX1080 ACX SC 16GB G.Skill Ripjaws X DDR3 1600 
Hard DriveOSMonitorKeyboard
Crucial M4SSD2 SATA III (128GB) Windows 10 Pro x64 Viewsonic XG2703-GS & Dell 24' 07WFPHC CM Storm Quick Fire Rapid, MX Red 
PowerCaseMouse
Corsair HX850 Coolermaster HAF 932 Logitech G502 
  hide details  
Reply
post #32 of 45
Quote:
Originally Posted by Alchemist07 View Post


As I suspected, even entry level CPUs ($60 or $300?!) are sufficient for modern games at 60 fps +, what we really need is better graphics.
They should have tested some AMD CPUs though.
Guess I am glad I didn't buy an Intel system though and went with a cheaper llano set up, rather spend that money on a beefier graphics card.
AMD should do better marketing though for their APU/CPUs^^

Lol, what? Please, compare the performance gain of graphics cards, to the performance gain of CPUs, starting with the Nvidia 8000 series and ATi 3000 series; and CPUs starting with Core 2 Duo/Quad.

8000 series to GTX 680 is getting close to an order of magnitude. Take one gen, GTX 295 vs GTX 480, and the 480 beats the 295, which is a dual-GPU. At stock clocks.. Core 2 Quad to i7-2600k is maybe 50%. CPU performance gain is going very sluggish, but without fail people buy the 10% annual performance increase from Intel.
 
Lanbox Lite
(16 items)
 
 
CPUGraphicsRAMHard Drive
i3-2310M Intel HD 3000 8GB DDR3 Samsung 830 64GB SSD 
Hard DriveOSMonitorCase
640GB Hitachi HDD Windows 7 13.3" LCD Magnesium Alloy, 3.2lbs 
CPUMotherboardGraphicsRAM
Phenom II X4 955 BE Gigabyte GA-MA78GPM-UD2H MSI Hawk R5770 3x2GB G.Skill DDR2 800 4-4-4-12 
Hard DriveHard DriveOptical DriveCooling
1TB Samsung Spinpoint F3 250 GB WD Caviar Black Samsung 20X DVD-R/RW Thermaltake MaxOrb 
CoolingOSMonitorPower
Noctua NF-B9-1600 Windows 7 Pro 64-bit BenQ E2420HD, 24" 1920x1080 TT Purepower 500W 
Case
TT Lanbox Lite 
  hide details  
Reply
 
Lanbox Lite
(16 items)
 
 
CPUGraphicsRAMHard Drive
i3-2310M Intel HD 3000 8GB DDR3 Samsung 830 64GB SSD 
Hard DriveOSMonitorCase
640GB Hitachi HDD Windows 7 13.3" LCD Magnesium Alloy, 3.2lbs 
CPUMotherboardGraphicsRAM
Phenom II X4 955 BE Gigabyte GA-MA78GPM-UD2H MSI Hawk R5770 3x2GB G.Skill DDR2 800 4-4-4-12 
Hard DriveHard DriveOptical DriveCooling
1TB Samsung Spinpoint F3 250 GB WD Caviar Black Samsung 20X DVD-R/RW Thermaltake MaxOrb 
CoolingOSMonitorPower
Noctua NF-B9-1600 Windows 7 Pro 64-bit BenQ E2420HD, 24" 1920x1080 TT Purepower 500W 
Case
TT Lanbox Lite 
  hide details  
Reply
post #33 of 45
Quote:
Originally Posted by jrbroad77 View Post

Lol, what? Please, compare the performance gain of graphics cards, to the performance gain of CPUs, starting with the Nvidia 8000 series and ATi 3000 series; and CPUs starting with Core 2 Duo/Quad.
8000 series to GTX 680 is getting close to an order of magnitude. Take one gen, GTX 295 vs GTX 480, and the 480 beats the 295, which is a dual-GPU. At stock clocks.. Core 2 Quad to i7-2600k is maybe 50%. CPU performance gain is going very sluggish, but without fail people buy the 10% annual performance increase from Intel.

Not this guy. Lol

First gpu mx400

Second radeon 9600xt

Third 7300gt

4th 4850 512

5th 48701gig

6th dual 4870s

Final gtx580 sc

Lol
Behind the Curve
(16 items)
 
  
CPUMotherboardGraphicsRAM
i5-4570 GA-Z87-D3H MSI GTX 980ti Gaming 6G Adata green special DDR3 1333 4x 4GB (16GB kit) 
Hard DriveOptical DriveCoolingOS
Seagate Barracuda Samsung DVD+RW CM 212 Windows 8.1 pro 64bit 
MonitorMonitorKeyboardPower
LG 34" 21:9 LG 23" 16:9 DELL!!! Corsair TX750w 
CaseMouseMouse PadAudio
CM 692 with window Razer-LACHESIS Razer-Sphex Pad Logitech 5.1 surround 
  hide details  
Reply
Behind the Curve
(16 items)
 
  
CPUMotherboardGraphicsRAM
i5-4570 GA-Z87-D3H MSI GTX 980ti Gaming 6G Adata green special DDR3 1333 4x 4GB (16GB kit) 
Hard DriveOptical DriveCoolingOS
Seagate Barracuda Samsung DVD+RW CM 212 Windows 8.1 pro 64bit 
MonitorMonitorKeyboardPower
LG 34" 21:9 LG 23" 16:9 DELL!!! Corsair TX750w 
CaseMouseMouse PadAudio
CM 692 with window Razer-LACHESIS Razer-Sphex Pad Logitech 5.1 surround 
  hide details  
Reply
post #34 of 45

 

This article should help some people looking for a good budget build. It really boils down to using a single GPU @1080p or below. 

 

If you're in a bind, buy a cheap CPU and get a better GPU. You won't notice a huge difference in most games with the faster CPU, and if you planned correctly you can upgrade the CPU later on if you feel the need to have something faster.


Quote:
Originally Posted by jrbroad77 View Post


Lol, what? Please, compare the performance gain of graphics cards, to the performance gain of CPUs, starting with the Nvidia 8000 series and ATi 3000 series; and CPUs starting with Core 2 Duo/Quad.
8000 series to GTX 680 is getting close to an order of magnitude. Take one gen, GTX 295 vs GTX 480, and the 480 beats the 295, which is a dual-GPU. At stock clocks.. Core 2 Quad to i7-2600k is maybe 50%. CPU performance gain is going very sluggish, but without fail people buy the 10% annual performance increase from Intel.


You're being quite generous saying 10% a year, haha. They are really pushing maybe 5% every generation. Most of the overall system performance gains aren't really due to the CPU, but the new chipset from Intel with each new CPU. I'm still on Lynnfield, and I have great performance. The only thing I wish I had was better features on my motherboard (USB 3, Pci-E 3, etc). I don't see the need to spend $400 for those features though....

Mini Trendy
(0 items)
  
Reply
Mini Trendy
(0 items)
  
Reply
post #35 of 45
And yet another stupid article where they fail to provide results that matter, minimum frame rate results!

Most CPU's can provide average results that are close to each and the only way to separate them is to compare the minimum frames.
post #36 of 45
Quote:
Originally Posted by Mr Frosty View Post

And yet another stupid article where they fail to provide results that matter, minimum frame rate results!
Most CPU's can provide average results that are close to each and the only way to separate them is to compare the minimum frames.


How do you figure? The average frame rate would be lower if the minimum frame rate is lower. I'd call foul if they posted only the maximum frame rate, because any CPU/GPU can pull off a few seconds of a high FPS and that would misrepresent the overall performance of the given hardware configuration. 

 

 

Mini Trendy
(0 items)
  
Reply
Mini Trendy
(0 items)
  
Reply
post #37 of 45
Quote:
Originally Posted by trendy View Post



How do you figure? The average frame rate would be lower if the minimum frame rate is lower. I'd call foul if they posted only the maximum frame rate, because any CPU/GPU can pull off a few seconds of a high FPS and that would misrepresent the overall performance of the given hardware configuration. 


I'll give you an example...

My old system was a 4.1ghz Phenom 2 x6 running crossfire 5850's clocked at 850Mhz core.

Running Crysis in DX10 very high with 4xAA at 1080p resulted in an average of 55fps and a minimum of 38fps.

Overclocking the 5850's to 1Ghz cores made no difference to the benchmark results at all.

Then I moved to a 2500k, same 5850's, same drivers, same clocks, same game settings only this time average was 60fps and minimum jumped to 48fps.

Overclocking the 5850's to 1Ghz improved things even more.

The 2500k only added 5fps onto the average which could be blown off as error of margin but a 10fps on the minimum? And believe me when I say you could feel the extra smoothness during fire fights on the 2500k.

Minimum frame rate has always been a bigger indication of CPU performance over average and it always will be.
post #38 of 45
Legit Review isn't very legit with bottlenecked reviews masquerading as scaling tests.
The 3930
(26 items)
 
Junior's 3930
(22 items)
 
DATA/HTPC
(20 items)
 
  hide details  
Reply
The 3930
(26 items)
 
Junior's 3930
(22 items)
 
DATA/HTPC
(20 items)
 
  hide details  
Reply
post #39 of 45
Quote:
Originally Posted by Mr Frosty View Post


I'll give you an example...
My old system was a 4.1ghz Phenom 2 x6 running crossfire 5850's clocked at 850Mhz core.
Running Crysis in DX10 very high with 4xAA at 1080p resulted in an average of 55fps and a minimum of 38fps.
Overclocking the 5850's to 1Ghz cores made no difference to the benchmark results at all.
Then I moved to a 2500k, same 5850's, same drivers, same clocks, same game settings only this time average was 60fps and minimum jumped to 48fps.
Overclocking the 5850's to 1Ghz improved things even more.
The 2500k only added 5fps onto the average which could be blown off as error of margin but a 10fps on the minimum? And believe me when I say you could feel the extra smoothness during fire fights on the 2500k.
Minimum frame rate has always been a bigger indication of CPU performance over average and it always will be.


They tested a range of CPUs that are all in the same "family". You switching CPUs and seeing a performance difference makes sense, because the Phenom 2 can't touch a Sandy Bridge in an apples to apples test (I.E. Single Core application tests).

 

The review in the OP shows 6 CPUs being used with identical hardware.

 

The G620 isn't going to run desktop publishing applications faster than the i7 2600k, and there will be a point where you hit a CPU bottleneck in heavily multi-threaded games (look at the MS3 benchmark)... but for $200+ less to get you in the door on a gaming PC... get the G620 and upgrade later.

Mini Trendy
(0 items)
  
Reply
Mini Trendy
(0 items)
  
Reply
post #40 of 45
linux got kernel 3.3 this week and it fix a part of the bufferbloat issue at the web end edge!instead of just counting packet they count bit, so the buffer doesnt get clogged!wish they had thot of that when they made everything !but then like the the year 2000 bug !SORRY WE DIDNT THING ABOUT IT!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Legit Reviews] CPU scaling with HD 7950