Soldato
- Joined
- 30 May 2012
- Posts
- 2,526
- Location
- Bristol
Hi all! Since I recently upgraded to a 3770k (thanks Tonester!) from my 3570k, I thought I might as well take the opportunity to run a few quick benchmarks on them.
Both CPUs were ran at 4.5 GHz, my 7950 was ran at 1000/1400 and the RAM at standard 1600MHz CL9. The rest of my specifications are in my signature below.
All the benchmarks were done on a 1920 x 1080 screen at the highest in game settings possible with Virtu MVP disabled. I only had time to run a few games, but here they are anyway!
Cinebench R11.5
3770k - 9.08
3570k - 7.42
As expected, Cinebench makes good use of hyperthreading allowing the 3770k to achieve a 22.4% higher score.
3DM11 (Performance preset - Physics score)
3770k - 10904
3570k - 8523
3DM11 makes good use of hyperthreading for the physics test, giving the 3770k a 27.9% lead over the 3570k for physics score. The graphics score however is within margin of error.
Heaven 3.0 (All settings maxed at 1920 x 1080 fullscreen)
3770k - 1109
3570k - 1109
Exactly the same score, with the most notable thing being the higher minimum FPS for the 3770k.
Sleeping Dogs
3770k - 44.9 average FPS
3570k - 42.6 average FPS
The 3770k gets a 5.4% higher average FPS, though this is borderline margin of error in terms of significance.
GTA IV
I accidentally didn't print screen this, but the scores are logged anyway:
3770k - 64.48 average FPS
3570k - 64.32 average FPS
These scores are very much within margin of error, with the only notable thing being the 48% CPU usage recorded for the 3570k as opposed to 28% on the 3770k.
BF3 (Going hunting mission, framerates recorded using FRAPS)
3770k - 71.4 average FPS
3570k - 70.7 average FPS
Again, these scores are within margin of error. I ran the benchmark for the first 2 minutes or so while you walk to the plane and then I looked left and right a few times while in the plane before stopping the benchmark. As this wasn't a standardised benchmark, there was more likelihood of human error. The most notable thing was the 55 FPS minimum for the 3570k compared to the 60 FPS minimum on the 3770k, though as this wasn't a standardised benchmark it is more prone to human error.
Metro 2033 (First mission, framerates recorded using FRAPS)
3770k - 46.6 average FPS
3570k - 46.5 average FPS
Margin of error again. I ran this benchmark during the first few minutes of the first mission up until the point where the player has just defeated the first load of monsters. The minimum FPS difference was more notable- 28 FPS for the 3770k and 24 FPS for the 3570k, though this was even more prone to error than the BF3 benchmark so this is of less significance.
Overall breakdown of performance differences (3770k % faster than 3570k):
Cinebench R11.5 - 22.4%
3DM11 Performance Physics score- 27.9%
Heaven 3.0 - 0%
Sleeping Dogs - 5.4%
GTA IV - 0.2% (Margin of error)
BF3 - 1.0% (Margin of error)
Metro 2033 - 0.2% (Margin of error)
So pretty much, the 3770k is roughly 20-30% faster in applications that can utilise 8 threads, but doesn't offer anything in the games I tested. Sleeping Dogs could probably just about be called a win for the 3770k, but as previously said I am close to calling it margin of error. BF3 multiplayer does seem to have a bit higher framerates on larger 64 person maps, however MP is too inconsistent to benchmark to actually show the difference accurately.
I have heard that some of the newer releases such as Crysis 3 do benefit from the 8 threads of the i7, but as I don't have it yet, I wasn't able to test the difference myself. With future game releases the i7 might start to show a difference, but we can't say anything for sure!
I hope this might have been of some use to you!
Both CPUs were ran at 4.5 GHz, my 7950 was ran at 1000/1400 and the RAM at standard 1600MHz CL9. The rest of my specifications are in my signature below.
All the benchmarks were done on a 1920 x 1080 screen at the highest in game settings possible with Virtu MVP disabled. I only had time to run a few games, but here they are anyway!
Cinebench R11.5
3770k - 9.08
3570k - 7.42
As expected, Cinebench makes good use of hyperthreading allowing the 3770k to achieve a 22.4% higher score.
3DM11 (Performance preset - Physics score)
3770k - 10904
3570k - 8523
3DM11 makes good use of hyperthreading for the physics test, giving the 3770k a 27.9% lead over the 3570k for physics score. The graphics score however is within margin of error.
Heaven 3.0 (All settings maxed at 1920 x 1080 fullscreen)
3770k - 1109
3570k - 1109
Exactly the same score, with the most notable thing being the higher minimum FPS for the 3770k.
Sleeping Dogs
3770k - 44.9 average FPS
3570k - 42.6 average FPS
The 3770k gets a 5.4% higher average FPS, though this is borderline margin of error in terms of significance.
GTA IV
I accidentally didn't print screen this, but the scores are logged anyway:
3770k - 64.48 average FPS
3570k - 64.32 average FPS
These scores are very much within margin of error, with the only notable thing being the 48% CPU usage recorded for the 3570k as opposed to 28% on the 3770k.
BF3 (Going hunting mission, framerates recorded using FRAPS)
3770k - 71.4 average FPS
3570k - 70.7 average FPS
Again, these scores are within margin of error. I ran the benchmark for the first 2 minutes or so while you walk to the plane and then I looked left and right a few times while in the plane before stopping the benchmark. As this wasn't a standardised benchmark, there was more likelihood of human error. The most notable thing was the 55 FPS minimum for the 3570k compared to the 60 FPS minimum on the 3770k, though as this wasn't a standardised benchmark it is more prone to human error.
Metro 2033 (First mission, framerates recorded using FRAPS)
3770k - 46.6 average FPS
3570k - 46.5 average FPS
Margin of error again. I ran this benchmark during the first few minutes of the first mission up until the point where the player has just defeated the first load of monsters. The minimum FPS difference was more notable- 28 FPS for the 3770k and 24 FPS for the 3570k, though this was even more prone to error than the BF3 benchmark so this is of less significance.
Overall breakdown of performance differences (3770k % faster than 3570k):
Cinebench R11.5 - 22.4%
3DM11 Performance Physics score- 27.9%
Heaven 3.0 - 0%
Sleeping Dogs - 5.4%
GTA IV - 0.2% (Margin of error)
BF3 - 1.0% (Margin of error)
Metro 2033 - 0.2% (Margin of error)
So pretty much, the 3770k is roughly 20-30% faster in applications that can utilise 8 threads, but doesn't offer anything in the games I tested. Sleeping Dogs could probably just about be called a win for the 3770k, but as previously said I am close to calling it margin of error. BF3 multiplayer does seem to have a bit higher framerates on larger 64 person maps, however MP is too inconsistent to benchmark to actually show the difference accurately.
I have heard that some of the newer releases such as Crysis 3 do benefit from the 8 threads of the i7, but as I don't have it yet, I wasn't able to test the difference myself. With future game releases the i7 might start to show a difference, but we can't say anything for sure!
I hope this might have been of some use to you!
Last edited:
so funny.
