[SOLVED] Bottleneck percentage? i5-4690, gtx1050ti

[M4G1CFOG]

hello im planning to upgrade my celeron g1820, its pretty good it runs gta all high settings except aa and particles. but its running on 210 farenheit which causes it to emergency shutdown, aand needs 10-13 mins to be turned on again. Im planning to upgrade my celeron so that i can play it better, will i5-4590 (lga1150) bottleneck gtx1050ti? because im planning to get an graphics card, thanks if you can recommend a better processor, ill also sell my gta acc just to upgrade, im close to 16yr old so dont go saying " go work " thank you C:

 

[Ziadul87]

hello im planning to upgrade my celeron g1820, its pretty good it runs gta all high settings except aa and particles. but its running on 210 farenheit which causes it to emergency shutdown, aand needs 10-13 mins to be turned on again. Im planning to upgrade my celeron so that i can play it better, will i5-4590 (lga1150) bottleneck gtx1050ti? because im planning to get an graphics card, thanks if you can recommend a better processor, ill also sell my gta acc just to upgrade, im close to 16yr old so dont go saying " go work " thank you C:

Bottlenecking is not measurable and definitely not universal. Quad core i5 is good enough if you're getting at a reasonable price (I'm guessing it's used). And any better CPU wouldn't be of much use if you're pairing it with 1050Ti (I'm guessing you already have the 1050Ti). But do check if you're mobo supports i5 4590

 

[John Chesterfield]

Bottlenecking is not measurable and definitely not universal. Quad core i5 is good enough if you're getting at a reasonable price (I'm guessing it's used). And any better CPU wouldn't be of much use if you're pairing it with 1050Ti (I'm guessing you already have the 1050Ti). But do check if you're mobo supports i5 4590

This. Also, make sure you get at least the stock cooler with the i5 4590, which is a decent enough match with a 1050ti.

 

[Joseph_138]

Bottlenecking is not measurable and definitely not universal. Quad core i5 is good enough if you're getting at a reasonable price (I'm guessing it's used). And any better CPU wouldn't be of much use if you're pairing it with 1050Ti (I'm guessing you already have the 1050Ti). But do check if you're mobo supports i5 4590

Ummm....no. Bottlenecking most certainly is measurable.

View: https://www.youtube.com/watch?v=kLkV5yoCUkk

 

[hotaru.hino]

Ummm....no. Bottlenecking most certainly is measurable.

View: https://www.youtube.com/watch?v=kLkV5yoCUkk

In specific cases yes, but not something like "If you pair of a Core i3-10300 with an RTX 3080, you'll bottleneck the RTX 3080 by 50%"

 

[Ziadul87]

Ummm....no. Bottlenecking most certainly is measurable.

View: https://www.youtube.com/watch?v=kLkV5yoCUkk

Let's make an example to make it clear to you.
Let's say there are two different games that require identical graphics performance. But one of them is 50% less CPU dependent than the other. Let's say RTX2060 can process 120 frames of both games per sec. Then if a processor can handle 90fps on the more CPU dependent one, it will likely be able to do ~180fps on the other. Now let's calculate bottleneck.
Game 1:
CPU: 90 ; GPU: 120; Bottleneck: 25% (due to CPU)
Game 2:
CPU: 180; GPU: 120; Bottleneck: 33.3% (due to GPU)


Every game requires a different amount of "CPU power/GPU power" ratio. That's why even with the same CPU+GPU combo, every game will have different amount of bottlenecking.

[ It's even possible for RTX 3080 to bottleneck a i3 10300 given the right circumstances]

 

[M4G1CFOG]

Bottlenecking is not measurable and definitely not universal. Quad core i5 is good enough if you're getting at a reasonable price (I'm guessing it's used). And any better CPU wouldn't be of much use if you're pairing it with 1050Ti (I'm guessing you already have the 1050Ti). But do check if you're mobo supports i5 4590

thank you, and yes i have lga1150 so it can fit in

 

[Ziadul87]

thank you, and yes i have lga1150 so it can fit in

Ok.

 

[Joseph_138]

Let's make an example to make it clear to you.
Let's say there are two different games that require identical graphics performance. But one of them is 50% less CPU dependent than the other. Let's say RTX2060 can process 120 frames of both games per sec. Then if a processor can handle 90fps on the more CPU dependent one, it will likely be able to do ~180fps on the other. Now let's calculate bottleneck.
Game 1:
CPU: 90 ; GPU: 120; Bottleneck: 25% (due to CPU)
Game 2:
CPU: 180; GPU: 120; Bottleneck: 33.3% (due to GPU)


Every game requires a different amount of "CPU power/GPU power" ratio. That's why even with the same CPU+GPU combo, every game will have different amount of bottlenecking.

[ It's even possible for RTX 3080 to bottleneck a i3 10300 given the right circumstances]

And you just proved my point by showing how bottlenecking can be measured. You just measured what was previously said to be unmeasurable.

 

[Ziadul87]

And you just proved my point by showing how bottlenecking can be measured. You just measured what was previously said to be unmeasurable.

Yes. You can measure bottlenecking the way I did.
Have you studied ideal gas in physics? We calculate all kinds of gas properties by assuming the gas is ideal. But real life is different. Real gases don't follow the rules of "ideal gas".

Similarly, in a very ideal situation with a very specific game scene, you can measure bottlenecking. (Kinda like math problemsets). But real life is not ideal and even in the same game the bottlenecking is not constant throughout different scenes. That's why, for a specific CPU+GPU combo it's impossible to measure bottlenecking and say it's XX%.

Do your homework well -_-

 

[USAFRet]

And you just proved my point by showing how bottlenecking can be measured. You just measured what was previously said to be unmeasurable.

In one specific application, with one particular setting level.

There is no universal "These two parts will bottleneck by XX%"

 

[GeebaGamer]

I have a i5 4670K sat alongside a GTX 1080 - runs pretty well to this day.
The i5 is overclocked but they seem a good balance performance wise.

Maybe consider an overclockable CPU but I cant see a 1050ti being held up by an i5-4690 unless you have a high thread count requirement.

 

[sizzling]

I had a 4670k overclocked to 4.3GHz so would outperform the 4690 slightly. It was paired with a 780 running 1080p. Before I upgraded I had noticed the cpu struggling in some games, here is what I remember.

Project Cars would have poor FPS when raining and the cpu hitting 100% on all cores. From memory I was seeing drops to 30 FPS which for me ruined the experience.

BF1 in large multiplayer maps would run with many dips under 60fps which impacted how much I enjoyed the game. Again the cpu was at 100% across all cores when the drops happened.

The worst though was Gears 5. This game was unplayable with fps <30 and looking like a slide show at times.

You can always reduce gpu workload by reducing game settings and/or resolution. However when your cpu is limiting your FPS there is very little you can do to improve performance. So regardless of gpu choice and depending on the games you play and your expectations for FPS you may or may not be satisfied with the 4690. Ultimately 4 thread CPU’s are a limiting factor in some games regardless of gpu choice and in extreme cases (Gears 5) they can be unplayable.

 

[GeebaGamer]

surprises me - my i5-4670K is at 4.1Ghz and is playing Half Life Alyx on 'High' settings which has a higher stated minimum requirement than my CPU and indeed Gears 5.

 

[Ziadul87]

Well, with my i3 6100 i had no problem with reaching 60fps on offline games(FC5/FH4). Only online i play is CSGO. Haven’t played gears though

 

[sizzling]

surprises me - my i5-4670K is at 4.1Ghz and is playing Half Life Alyx on 'High' settings which has a higher stated minimum requirement than my CPU and indeed Gears 5.

My experience of game minimum requirements is they don’t give you any idea how well a game will play, just that it will run. What one game feels is acceptable minimum performance another may not. BF1 was well documented to have issues in big multiplayer maps with a 4 thread cpu and I did find some information describing the rain performance hit in PC. As for Gears 5 that was a real anomaly. However I had many other games at the time that did not have issues.