- Joined
- Feb 22, 2009
- Messages
- 757 (0.14/day)
System Name | Lenovo 17IMH05H |
---|---|
Processor | Core i7 10750H |
Video Card(s) | GTX 1660 Ti |
Audio Device(s) | SSL2 |
Software | Windows 10 Pro 22H2 |
Benchmark Scores | i've got a shitload of them in 15 years of TPU membership |
It's time for another round of my gaming benchmarks series. This time in a relatively short time i had the chance to compare 3 processors, which people probably use in their mainstream gaming computers. There are quite many high end Core i5 processor benchmarks out there, those compare Core i5 "K" old and new generation chips, so that i felt the need to fill the space in the lower price segment, as well as include a skylake Core i3. Also, all of the skylake vs sandy bridge comparisons i've seen include DDR4 vs DDR3, which i feel is a bit unfair for old apple vs new apple comparison...
Now, there is no point arguing if you should upgrade your 4 generations older processor to a new one for anything other than gaming - the fact that each generation, whether it's tick ir tock, adds 7 - 10 % IPC performance improvement for programs is a fact, so it's up to your needs if you need the performance boost. And having in mind the improvements of the I/O of the past and present systems, that generational gap seems even more wide.
But let's just stick with gaming and similar systems.
12 games were tested using 1920x1080 resolution, no AA and custom quality/performance settings. Since this is a CPU, and not a VGA benchmark, the settings were not set to the absolute maximum, but rather to the point of maximum with playable settings, above 60 FPS.
The benchmark rig is
Core i5 2500 3.7/3.6/3.5/3.4 GHz 6 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Core i5 6500 3.6/3.5/3.4/3.3 GHz 6 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Core i3 6320 3.9/3.8 GHz 4 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Asus P8P67/Gigabyte GA-B150M-D2V
Corsair Vengeance 2X4 GB DDR3 1600 MHz CL9
Adata SP900 256 GB MLC SSD/Patriot Pyro 120 GB MLC SSD (Windows directory)
Western Digital SE 2 TB 7200 rpm 64 MB HDD (game directory)
Zotac GeForce GTX970 AMP! Omega Core 4 GB
Windows 7 Pro 64 bit.
NVIDIA Forceware 376.33
---------------------------------------------------------------------------------
VIDEO PRESENTATION
Alien Isolation
Real cores of Core i5 6500 and Core i5 2500 perform better than virtual ones of Core i3 6320, and IPC is not important here.
Arma 3 Apex Edition
While virtual cores of Core i3 6320 can not match real ones of Core i5 6500, the IPC improvements of skylake chips over sandy bridge make the difference in FPS as Core i3 6320 is slightly ahead of Core i5 2500.
Call of Duty Advanced Warfare
The game favors real cores above virtual ones, and IPC improvements also do matter as Core i3 6320 is slightly faster than Core i5 2500.
Company of Heroes 2
Again, IPC matters here, and virtual cores are being utilized almost as good as real ones.
Crysis 3
Core i3 6320 only suffers in minimal FPS when compared to Core i5 6500 and even to Core i5 2500 to a smal extent. IPC influence is relevant here.
Dying Light
IPC is not important in performance, and virtual cores of Core i3 6320 do just as good as real ones.
Doom
It's a copy cat situation of Dying Light.
Far Cry 4
Again, Core i3 6320 suffers in minimal FPS when compared to both Core i5 processors. Can not see the IPC influence here.
Metro Last Light Redux
For some unexplained reason, Core i5 6500 is far inferior to Core i5 2500. I've repeated this test so many times - even disassembled PC per new! Did tests in different days even! I've got no arguments here. The only thing relevant is that Core i3 6320 is slightly slower than Core i5 2500, which is normal.
Rainbow Six Siege
While the difference is not big between Core i5 2500 and Core i5 6500, the fact that Core i5 6500 gets beaten by Core i5 2500, which is not even "K" is bizzare to say the least.. Anyway, real cores are preferred over virtual ones here.
Rise of Tomb Raider
Virtual cores perform as good as real ones and IPC influence does not matter here.
Witcher 3 Wild Hunt
I already knew from my previous Ivy Bridge Core i3 vs Core i5 test that virtual cores perform just as good as real ones in Witcher 3, even in minimal FPS, so why should it be different with skylake? IPC influence is not relevant here.
CONCLUSIONS
1. The IPC improvements from generation to generation do matter in gaming! However, the difference is so small, that it does not affect the gameplay performance and experience!
2. While clearly Core i3 virtual cores tend to lag behind Core i5 real ones in some cases, especially in minimal FPS region, overall the same generation Core i3 delivers the same gaming experience as Core i5 in mid-range systems. This clearly would not be the same in high-end systems with extreme quality settings.
3. If you have a mid range PC (GTX970) for 1080P gaming, there is no way you should replace your sandy bridge Core i5 2500 just for the FPS sake. Unless it's for programs or hardware compatibility, the Core i5 2500 is to stay in 2017.
Now, there is no point arguing if you should upgrade your 4 generations older processor to a new one for anything other than gaming - the fact that each generation, whether it's tick ir tock, adds 7 - 10 % IPC performance improvement for programs is a fact, so it's up to your needs if you need the performance boost. And having in mind the improvements of the I/O of the past and present systems, that generational gap seems even more wide.
But let's just stick with gaming and similar systems.
12 games were tested using 1920x1080 resolution, no AA and custom quality/performance settings. Since this is a CPU, and not a VGA benchmark, the settings were not set to the absolute maximum, but rather to the point of maximum with playable settings, above 60 FPS.
The benchmark rig is
Core i5 2500 3.7/3.6/3.5/3.4 GHz 6 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Core i5 6500 3.6/3.5/3.4/3.3 GHz 6 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Core i3 6320 3.9/3.8 GHz 4 MB cache (C1E disabled, EIST enabled, turbo enabled, TT disabled)
Asus P8P67/Gigabyte GA-B150M-D2V
Corsair Vengeance 2X4 GB DDR3 1600 MHz CL9
Adata SP900 256 GB MLC SSD/Patriot Pyro 120 GB MLC SSD (Windows directory)
Western Digital SE 2 TB 7200 rpm 64 MB HDD (game directory)
Zotac GeForce GTX970 AMP! Omega Core 4 GB
Windows 7 Pro 64 bit.
NVIDIA Forceware 376.33
---------------------------------------------------------------------------------
VIDEO PRESENTATION
Alien Isolation
Real cores of Core i5 6500 and Core i5 2500 perform better than virtual ones of Core i3 6320, and IPC is not important here.
Arma 3 Apex Edition
While virtual cores of Core i3 6320 can not match real ones of Core i5 6500, the IPC improvements of skylake chips over sandy bridge make the difference in FPS as Core i3 6320 is slightly ahead of Core i5 2500.
Call of Duty Advanced Warfare
The game favors real cores above virtual ones, and IPC improvements also do matter as Core i3 6320 is slightly faster than Core i5 2500.
Company of Heroes 2
Again, IPC matters here, and virtual cores are being utilized almost as good as real ones.
Crysis 3
Core i3 6320 only suffers in minimal FPS when compared to Core i5 6500 and even to Core i5 2500 to a smal extent. IPC influence is relevant here.
Dying Light
IPC is not important in performance, and virtual cores of Core i3 6320 do just as good as real ones.
Doom
It's a copy cat situation of Dying Light.
Far Cry 4
Again, Core i3 6320 suffers in minimal FPS when compared to both Core i5 processors. Can not see the IPC influence here.
Metro Last Light Redux
For some unexplained reason, Core i5 6500 is far inferior to Core i5 2500. I've repeated this test so many times - even disassembled PC per new! Did tests in different days even! I've got no arguments here. The only thing relevant is that Core i3 6320 is slightly slower than Core i5 2500, which is normal.
Rainbow Six Siege
While the difference is not big between Core i5 2500 and Core i5 6500, the fact that Core i5 6500 gets beaten by Core i5 2500, which is not even "K" is bizzare to say the least.. Anyway, real cores are preferred over virtual ones here.
Rise of Tomb Raider
Virtual cores perform as good as real ones and IPC influence does not matter here.
Witcher 3 Wild Hunt
I already knew from my previous Ivy Bridge Core i3 vs Core i5 test that virtual cores perform just as good as real ones in Witcher 3, even in minimal FPS, so why should it be different with skylake? IPC influence is not relevant here.
CONCLUSIONS
1. The IPC improvements from generation to generation do matter in gaming! However, the difference is so small, that it does not affect the gameplay performance and experience!
2. While clearly Core i3 virtual cores tend to lag behind Core i5 real ones in some cases, especially in minimal FPS region, overall the same generation Core i3 delivers the same gaming experience as Core i5 in mid-range systems. This clearly would not be the same in high-end systems with extreme quality settings.
3. If you have a mid range PC (GTX970) for 1080P gaming, there is no way you should replace your sandy bridge Core i5 2500 just for the FPS sake. Unless it's for programs or hardware compatibility, the Core i5 2500 is to stay in 2017.
Last edited: