Would an i7 4770k with 2 cores hyper threading disabled perform as same as a Pentium G3258

[JollyGamerHD]

I was arguing with someone on youtube. They made a video claiming that his i7 with 2 cores and hyper threading disabled is the same as a Pentium G3258. He thought that because it runs on the same architecture and has the same amount of cores it would perform exactly the same. He paired his CPU with a GTX 780 and obviously the "pentium" was bottlenecking the GPU. Now anyway he tested games like Crysis 3, Battlefield 3 and other games. My concern is that the people who tested with an actual G3258 and for example a 750ti got far better results, in his video, it was stuttering all the time and was getting FPS mostly below 30fps and it was an unplayable experience, the opposite of what you'd achieve on an actual Pentium. Can somebody please explain why exactly the i7 is far worse when practically everything is disabled compared to the Pentium G3258 i just want to know for the sake of the guy who made this video, as he won't accept that the performance is not the same. 

 

Thanks in advance

 

This is his video: 

 

[SLAYR]

If they are clocked the same then they should perform exactly the same, but most people oc the hell out of the g3258.

 

And the g3258 will most likely oc higher then the i7.

[Electronics Wizardy]

If there clock the same, they should preform simmilar. The i7 does have more cache and features than a pentuium(8mb vs 3mb)

[JollyGamerHD]

Just now, SLAYR said:

If they are clocked the same then they should perform exactly the same, but most people oc the hell out of the g3258.

 

And the g3258 will most likely oc higher then the i7.

Yeah he thinks that they would as well, In theory anyway they kind of should. But watching the video its clear that the performance on the i7 is considerably worse. Its funny though, ive linked this guy videos proving it to him and he still won't believe me.

[i_build_nanosuits]

1 minute ago, JollyGamerHD said:

Yeah he thinks that they would as well, In theory anyway they kind of should. But watching the video its clear that the performance on the i7 is considerably worse. Its funny though, ive linked this guy videos proving it to him and he still won't believe me.

i think the videos of people using 750ti are probably running lower settings...the 780 being a powerful GPU is you max out the settings obviously it will stutter and run out of processing heeadroom...it's a 60$ CPU what else do you expect?

 

linus emulated CPU's quite often and they proved to perform on par to the real thing:

 

[Carlos1010]

@SLAYR  @i_build_nanosuits So if I want a CPU to last me years to come should I get a Pentium G3258 or a i7 6700k?

[SLAYR]

1 minute ago, JollyGamerHD said:

Yeah he thinks that they would as well, In theory anyway they kind of should. But watching the video its clear that the performance on the i7 is considerably worse. Its funny though, ive linked this guy videos proving it to him and he still won't believe me.

Very hard to prove unless you own both chips and do everything with the same system, os, and drive as things like malware could also have an effect, or what recording programs.

[i_build_nanosuits]

1 minute ago, Carlos1010 said:

@SLAYR  @i_build_nanosuits So if I want a CPU to last me years to come should I get a Pentium G3258 or a i7 6700k?

humm...i don't know!

[Sauron]

The i7 would perform even slightly better because of the extra cache (clock for clock). If he gets worse performance he has a different problem.

[Curufinwe_wins]

1 minute ago, JollyGamerHD said:

Yeah he thinks that they would as well, In theory anyway they kind of should. But watching the video its clear that the performance on the i7 is considerably worse. Its funny though, ive linked this guy videos proving it to him and he still won't believe me.

The other thing you need to remember is if he is recording this footage on the same pc it will massively hurt the capabilities of a dual core no HT. 

 

So his experience shown here may not be representative of his experience when not streaming.

 

100% a 4770k with two cores disabled no HT will preform IDENTICAL to a G3258 at the same clock speeds.

[i_build_nanosuits]

 

Just now, Sauron said:

The i7 would perform even slightly better because of the extra cache (clock for clock). If he gets worse performance he has a different problem.

exactly

[JollyGamerHD]

1 minute ago, i_build_nanosuits said:

i think the videos of people using 750ti are probably running lower settings...the 780 being a powerful GPU is you max out the settings obviously it will stutter and run out of processing heeadroom...it's a 60$ CPU what else do you expect?

 

I'm not expecting anything but he isnt actually using the pentium, the video that i watched with the 750ti was running on the very high preset at 1080p at well over 30fps at all times in the welcome to the jungle mission which tends to be very demanding. In his video it performs awfully on the i7 and actually well on the pentium. And because you've got the 780 that means bottlenecking, not that it'll perform worse than a 750ti.

[SLAYR]

3 minutes ago, Carlos1010 said:

@SLAYR  @i_build_nanosuits So if I want a CPU to last me years to come should I get a Pentium G3258 or a i7 6700k?

If this is a serious question then make your own thread in the cpu, motherboards, and ram section.

[JollyGamerHD]

2 minutes ago, i_build_nanosuits said:

 

exactly

Yeah but im trying to see whether it would be better or not as we already have results that it performs a lot worse, im just trying to see why. The weird thing is that this guy wont accept that it performs worse. lol

[i_build_nanosuits]

1 minute ago, JollyGamerHD said:

Yeah but im trying to see whether it would be better or not as we already have results that it performs a lot worse, im just trying to see why. The weird thing is that this guy wont accept that it performs worse. lol

the i7 with only 2 cores active and no hyper-threading if clocked at the same clockrate as the pentium would perform ever so slightly better because it has more cache.

turning an i7 to a dual core has no negative impact on it...

[JollyGamerHD]

7 minutes ago, i_build_nanosuits said:

i think the videos of people using 750ti are probably running lower settings...the 780 being a powerful GPU is you max out the settings obviously it will stutter and run out of processing heeadroom...it's a 60$ CPU what else do you expect?

 

linus emulated CPU's quite often and they proved to perform on par to the real thing:

 

yeah well in this case it really didnt, i know for a fact that this guy's test is a failure but im just trying to see why, i think it may be optimisation. The i7 isnt meant to run as he ran it but the pentium is. And the pentium i think tends to have a better single and dual core performance as it only has two cores and is designed to run like that.

[i_build_nanosuits]

3 minutes ago, JollyGamerHD said:

yeah well in this case it really didnt, i know for a fact that this guy's test is a failure but im just trying to see why, i think it may be optimisation. The i7 isnt meant to run as he ran it but the pentium is. And the pentium i think tends to have a better single and dual core performance as it only has two cores and is designed to run like that.

no you can check the cinebench R15 thread on the CPU sub section of the forum...an haswell CPU core is an haswell CPU core and they perform identical if you run them at the same clockspeed...the i7 only has more of them and it has hyper-threading. if you disable HT and lock it to only 2 cpu cores you in fact get exactly the same performance as the pentium running at the same clockspeed both in single-threaded and dual-threaded loads...they are exactly the same...if you want something with faster processing cores then you have to look at architectural improvements done by intel in the form of broadwell or skylake skus.

[JollyGamerHD]

8 minutes ago, SLAYR said:

Very hard to prove unless you own both chips and do everything with the same system, os, and drive as things like malware could also have an effect, or what recording programs.

Nothing would have that big of an effect unless you're playing another game in the background. The bottom line is that the pentium performed way better than the i7 in that state. And yes you would have to own both, but as he doesn't and thinks he knows everything, he thinks that they perform the same even though they don't.

[JollyGamerHD]

13 minutes ago, i_build_nanosuits said:

i think the videos of people using 750ti are probably running lower settings...the 780 being a powerful GPU is you max out the settings obviously it will stutter and run out of processing heeadroom...it's a 60$ CPU what else do you expect?

 

linus emulated CPU's quite often and they proved to perform on par to the real thing:

 

No the 750ti ran on high settings as well and had no problems, also games like that aren't known to be CPU bound at all, unless its a really really bad cpu.

[JollyGamerHD]

12 minutes ago, Curufinwe_wins said:

The other thing you need to remember is if he is recording this footage on the same pc it will massively hurt the capabilities of a dual core no HT. 

 

So his experience shown here may not be representative of his experience when not streaming.

 

100% a 4770k with two cores disabled no HT will preform IDENTICAL to a G3258 at the same clock speeds.

??? the other people playing on the G3258 also recorded it? whats your point? And No he got the same experience either way as he was recording with nvidia shadowplay which only has like a 5% decrease in performance. And Even though they SHOULD perform the same(in theory), just look at his video and compare it with another person's video with and actual g3258, it is like running on a single core processor that is broken vs a g3258.

[i_build_nanosuits]

 

5 minutes ago, JollyGamerHD said:

No the 750ti ran on high settings as well and had no problems, also games like that aren't known to be CPU bound at all, unless its a really really bad cpu.

i think you should take the time to watch this video...digital foundry is a very reputable source...they found that locking the framerate to 30FPS with the G3258 and pairing it with an appropriate GPu and playing on reasonable settings help solving a lot of this issue you might encounter with a low power chip like those dual core...they arent bad and for 60$ they perform well when paired with appropriate graphics cards at appropriate settings...crysis 3 is still considered as a state of the art game and it's very demanding both on the CPU and GPU...that level is called ''welcome to the jungle'' and for a good reason is that there is a lot of grass animations and trees and plants and what not and everything is moving....and a lot of this has physics that is processed on the CPU side..i think there is a reason why this scene was selected and the other games as well.

 

 

[JollyGamerHD]

1 minute ago, i_build_nanosuits said:

 

i think you would should take the time to watch this video...digital foundry is a very reputable source...they found that locking the framerate to 30FPS with the G3258 and pairing it with an appropriate GPu and playing on reasonable settings help solving a lot of this issue you might encounter with a low power chip like those dual core...they arent bad and for 60$ they perform well when paired with appropriate graphics cards at appropriate settings...crysis 3 is still considered as a state of the art game and it's very demanding both on the CPU and GPU...that level is called ''welcome to the jungle'' and for a good reason is that there is a lot of grass animations and trees and plants and what not and everything is moving....and a lot of this has physics that is processed on the CPU side..i think there is a reason why this scene was selected and the other games as well.

 

 

yes im aware of some stuttering issues but they can be fixed, my concern is that on his i7 that was supposed to be like the same but with a gtx 780 got fps in the 20s and was unplayable due to the horrific stuttering. This is still completely different and is a lot better, because its unplayable on the i7 anything you do in that state. The funny thing is that he won't believe that it runs better on other videos.

[JollyGamerHD]

14 minutes ago, i_build_nanosuits said:

no you can check the cinebench R15 thread on the CPU sub section of the forum...an haswell CPU core is an haswell CPU core and they perform identical if you run them at the same clockspeed...the i7 only has more of them and it has hyper-threading. if you disable HT and lock it to only 2 cpu cores you in fact get exactly the same performance as the pentium running at the same clockspeed both in single-threaded and dual-threaded loads...they are exactly the same...if you want something with faster processing cores then you have to look at architectural improvements done by intel in the form of broadwell or skylake skus.

You're missing the point, i know that they're technically supposed to perform the same but if you actually watched his video and compared it to the same game running on an actual pentium you'd know that it performs a lot better on the pentium. Either he's done something wrong or i don't know. Anyway i won't looking for stuff that i know but why this gap in performance is occurring.

[i_build_nanosuits]

Just now, JollyGamerHD said:

yes im aware of some stuttering issues but they can be fixed, my concern is that on his i7 that was supposed to be like the same but with a gtx 780 got fps in the 20s and was unplayable due to the horrific stuttering. This is still completely different and is a lot better, because its unplayable on the i7 anything you do in that state. The funny thing is that he won't believe that it runs better on other videos.

like i said...there is probably a REASON why they run better in your other videos...not that the simulated pentium behaves any different.

they probably don't run ULTRA settings across the board..(do they actually show the settings in the videos you've linked to him?! cause just droping grass quality and lighting quality to medium will have a tremendous benefit to the CPU in that game) and if you run all medium or high settings...you'd be fine for the most part...ULTRA settings are where the real demanding stuff comes into play...running ULTRA vs HIGH is like a 25% performance penalty...and this would be enough to push that little dual-core under the water like you see in the video...it's running demanding stuff...probably way out of it's league with a GTX 780 like that...and this video seem to have been created to prove just that...that a GPU of that caliber is in a different league for a 60$ intel CPU...even if you can overclock it, even if it's single-threaded performance would match that of the more expensive core i7.

[JollyGamerHD]

1 minute ago, i_build_nanosuits said:

like i said...there is probably a REASON why they run better in your other videos...not that the simulated pentium behaves any different.

they probably don't run ULTRA settings across the board..(do they actually show the settings in the videos you've linked to him?! cause just droping grass quality and lighting quality to medium will have a tremendous benefit to the CPU in that game) and if you run all medium or high settings...you'd be fine for the most part...ULTRA settings are where the real demanding stuff comes into play...running ULTRA vs HIGH is like a 25% performance penalty...and this would be enough to push that little dual-core under the water like you see in the video...it's running demanding stuff...probably way out of it's league with a GTX 780 like that...and this video seem to have been created to prove just that...that a GPU of that caliber is in a different league for a 60$ intel CPU...even if you can overclock it, even if it's single-threaded performance would match that of the more expensive core i7.

I'll link you the video, and its not just crysis 3. Watch his video and see how badly the other games perform that are in no way cpu intensive and ill link you the crysis 3 footage from another guy.