Bottleneck Calculator Legit?

[_Hustler_One_]

Hello good people..

 

Have someone used this as reference?

https://thebottlenecker.com/calculator

 

It says that the minimum NVIDIA GPU which doesn't bottleneck the Intel i7-8700K is GTX 1070 Ti, the regular 1070 will still bottleneck the 8700K at 13%, and also any higher than 10% is considered as bottlenecking..

 

Is there any other kind of bottleneck calculator that more reliable?

 

My workstation runs 8700K+GTX1060 6Gb, and the other one runs 8700K+1070Ti.. I'm living on the edge.. the very edge..

[Origami Cactus]

Bottleneck calculators are stupid. The word bottleneck is just used to fearmonger.

Just monitor gpu and cpu usage while doing your workloads, and draw your conclusion from that. Cpu and gpu usage differ per workload, so that is why you cant use a random online calculator.

Most probably both your workstations are fine, especially if the workload is cpu based.

 

[Enderman]

That's BS.

Both the CPU and GPU contribute to performance.

There's also a ton of variables that will change how much performance impact one component has over the other such as resolution, framerate, textures, what games or programs, etc...

 

There is no "X bottlenecks Y" clear answer.

The 8700K is the second-best gaming CPU that exists and will go great with all those GPUs you mentioned above and more.

The more powerful the GPU, the higher performance you will get and the more money you will have to spend.

[CarlBar]

22 minutes ago, _Hustler_One_ said:

Hello good people..

 

Have someone used this as reference?

https://thebottlenecker.com/calculator

 

It says that the minimum NVIDIA GPU which doesn't bottleneck the Intel i7-8700K is GTX 1070 Ti, the regular 1070 will still bottleneck the 8700K at 13%, and also any higher than 10% is considered as bottlenecking..

 

Is there any other kind of bottleneck calculator that more reliable?

 

My workstation rus 8700K+GTX1060 6Gb, and the other one runs 8700K+1070Ti.. I'm living on the edge.. the very edge..

 

I use it for basic sanity checking, (and it came recommended from several people), i wouldn't take it as gospel truth and AFAIK it's gaming focused so for GPU utilizing compute workloads the results would be very different. 

 

Overall though the result sounds very reasonable for gaming if you pay attention to benchmarking.as you need much more powerful hardware on the GPU side to start bottlenecking from CPU, given the amount of step down those cards represent it's not unreasonable to think you'd be getting into the zone where the GPU may become an issue for gaming.

[Enderman]

6 minutes ago, CarlBar said:

 

I use it for basic sanity checking, (and it came recommended from several people), i wouldn't take it as gospel truth and AFAIK it's gaming focused so for GPU utilizing compute workloads the results would be very different. 

 

Overall though the result sounds very reasonable for gaming if you pay attention to benchmarking.as you need much more powerful hardware on the GPU side to start bottlenecking from CPU, given the amount of step down those cards represent it's not unreasonable to think you'd be getting into the zone where the GPU may become an issue for gaming.

No it's not reasonable at all because a GPU doesn't bottleneck a CPU.

Even with something like a GTX1050ti you will get higher performance in a game using an 8700k than an 8600K.

This is because BOTH the CPU and GPU contribute to performance, not the lowest common denominator.

The amount of contribution from each also depends on what game, what resolution, etc etc etc as I said before.

 

So no, there is no correct answer to "does this bottleneck"

The only correct answer is how much performance difference there is between CPU X and Y when using this GPU, running this game, at these settings.

Or, how much performance difference there is between GPU X and Y when using this CPU, with this game, at these quality settings.

Any "bottlenecking calculator" you look at is just BS numbers and a placebo effect to go along with it.

[_Hustler_One_]

Great! Thanks for confirmations

[CarlBar]

34 minutes ago, Enderman said:

No it's not reasonable at all because a GPU doesn't bottleneck a CPU.

Even with something like a GTX1050ti you will get higher performance in a game using an 8700k than an 8600K.

This is because BOTH the CPU and GPU contribute to performance, not the lowest common denominator.

The amount of contribution from each also depends on what game, what resolution, etc etc etc as I said before.

 

So no, there is no correct answer to "does this bottleneck"

The only correct answer is how much performance difference there is between CPU X and Y when using this GPU, running this game, at these settings.

Or, how much performance difference there is between GPU X and Y when using this CPU, with this game, at these quality settings.

Any "bottlenecking calculator" you look at is just BS numbers and a placebo effect to go along with it.

 

You do understand what bottlenecking is right? Oh no of course you don't. That was blindingly obvious from your very first post in this thread and you just confirmed it more. Bottlenecking isn't about getting zero performance improvement from going to a new part, it's about getting less than the full potential. Yes at very severe bottlenecking you'll run into the situation where you get zero performance improvement. There's a whole slew of videos on the internet that show the effect off. But long before you stop getting any gains you'll get to the point where your gains aren't really very good given the difference in the grunt underlying the new component. Most reviewers are agreeing the 2080Ti without hefty CPU OC'ing is running face first into CPU limits as the performance jump is nowhere near what the raw hardware capabilities should allow for.

 

 

[Enderman]

8 hours ago, CarlBar said:

Bottlenecking isn't about getting zero performance improvement from going to a new part, it's about getting less than the full potential.

With that logic then everything is always bottlenecked.

No matter what you do there is a bottleneck because you're not getting infinite performance (which is technically also true)

"The 8700K would be a bottleneck then because the 9900K would get better performance.

The 9900K is a bottleneck too because the 10900K will get better performance.

Etc etc etc."

You never achieve "full potential"

 

This is what real bottlenecking actually is:

As you can see, the performance is limited by the GPU only, regardless of the CPU.

This is very rare, as I said earlier, it almost always depends on both the CPU and GPU, for example see below:

As you can see for the two GPUs above the 1050ti the performance improves as you improve the CPU, but ALSO as you improve the GPU.

The total performance is a combination of both.

 

Please do some research on what bottlenecking is.

https://en.wikipedia.org/wiki/Bottleneck_(engineering)

By definition it is when a single component is limiting the performance of a system.

If you have XXX going through a bottleneck that only lets X out, giving it XXXXXX will also only let X out, not 2X

[CarlBar]

Sorry but it's you who need to do the research there are countless good videos out there by many tech channels that have gone over it in great detail. You can and will see examples like the top image you posted, but that is not the only form of bottlenecking.

 

And your not wrong, every system is bottlenecked to some degree, (and the bottleneck calculator acknowledges this), but no one talking about bottlenecking in a system building or optimizing case cares about that. What they care is losing real world noticeable amounts. A few percent is going to be obscured by program optimization and background clutter quite easily. A 10% bottleneck or more in any one area however is going to result in somthing noticeable underperforming.in all but the most extreme use case scenarios. It may still offer a performance improvement over the next step down, but if it should be giving you a 50% boost and all your getting is a 20% boost you've got a bottleneck. And that sort of bottleneck is a lot easier to encounter than a hard "cannot go any further" bottleneck like in your top imagine.

 

There's more than one type of bottleneck and your focused on the hard type as the only type when other less extreme types exist.

[Dime456]

On 11/10/2018 at 11:25 AM, Enderman said:

That's BS.

Both the CPU and GPU contribute to performance.

There's also a ton of variables that will change how much performance impact one component has over the other such as resolution, framerate, textures, what games or programs, etc...

 

There is no "X bottlenecks Y" clear answer.

The 8700K is the second-best gaming CPU that exists and will go great with all those GPUs you mentioned above and more.

The more powerful the GPU, the higher performance you will get and the more money you will have to spend.

Do you know then would the i5 6600 bottleneck the new rtx2060?

[Dime456]

.

[Enderman]

4 minutes ago, Dime456 said:

Do you know then would the i5 6600 bottleneck the new rtx2060?

Well you will definitely get less performance than a 9900K.

How much less depends on the game, resolution, framerate, quality settings, textures, background tasks, etc.

[The Torrent]

On 11/10/2018 at 10:16 AM, _Hustler_One_ said:

Hello good people..

 

Have someone used this as reference?

https://thebottlenecker.com/calculator

 

It says that the minimum NVIDIA GPU which doesn't bottleneck the Intel i7-8700K is GTX 1070 Ti, the regular 1070 will still bottleneck the 8700K at 13%, and also any higher than 10% is considered as bottlenecking..

 

Is there any other kind of bottleneck calculator that more reliable?

 

My workstation runs 8700K+GTX1060 6Gb, and the other one runs 8700K+1070Ti.. I'm living on the edge.. the very edge..

Way I see them, bottlenecks, cpu at least, are a good think. Give you room for upgradability in the future. Just use whichever graphics card suits your needs and have a cpu good enough (if u can’t be bothered to upgrade cpu in future) to allow a better gpu.

[johmei]

On 11/10/2018 at 5:50 AM, Enderman said:

Even with something like a GTX1050ti you will get higher performance in a game using an 8700k than an 8600K.

(Sorry for necromancing the thread.  It isn't THAT old and I can't stand misinformation)

 

Dude... No.  I don't care what GPU you pair it with, the CPU doesn't contribute enough to most games that you'll see any kind of consistent performance increase in an 8700k over an 8600k.  That is simply not true at all.  There is a limit to this, of course, and it does somewhat depend on the game, but what you said is simply not true no matter how you look at it.

Edited October 11, 2019 by johmei

[DeltaPlays]

If I play Asphalt 9: Legends on a A6-7400K with GTX 1050, will I get lower performance than full? Coz I am not that rich so If I save money to buy something I want full performance. Or should I upgrade my CPU to A10-7890K?

[dragonheals12]

Personally, I like GPU check, because it can help give you a good idea as to what's going on, what to expect, and how much  performance you'll get. It's gpucheck.com, and I find it reasonably reliable. Generally, you don't need to worry too much about a bottleneck unless you're going hard core price to performance. Also, more GPU is generally good, and even older CPUs keep up pretty well with high end GPUs. From what I've been seeing, it looks like there has been a lot more advancement in GPU tech than CPU tech. 

[hotplug7191]

i guess bottleneck checker just to tell you the possibility of bottleneck in percentages. if gpu is bottleneck-ing cpu around 10% to 30% then you could use higher refresh rate monitor or run it on higher resolution than you used to be. it should minimize the cpu bottleneck

[trevb0t]

That's the kind of bottleneck that you wouldn't mind having.

If your CPU wasn't powerful enough for the GPU, that's where the issues may start to manifest.

This just indicates you've got a CPU that will hand a GPU upgrade 3 years down the line.

 

I'm not sure I agree that the 8700K is the "second best gaming CPU", but it's a great unit.

You could stick an RX 570 in there if you wanted, and not notice weird issues in gaming. Your RX 570 would just be using it's full power trying to keep up (unsuccessfully.) 

[trevb0t]

On 10/11/2019 at 7:50 AM, johmei said:

(Sorry for necromancing the thread.  It isn't THAT old and I can't stand misinformation)

 

Dude... No.  I don't care what GPU you pair it with, the CPU doesn't contribute enough to most games that you'll see any kind of consistent performance increase in an 8700k over an 8600k.  That is simply not true at all.  There is a limit to this, of course, and it does somewhat depend on the game, but what you said is simply not true no matter how you look at it.

That's not an accurate statement. The CPU plays a large part in game processing, regardless of which GPU is in the system.

A Ryzen 5 2600 and an i7-9700K will deliver vastly different performance when paired with the same compatible GPU.

[Jerikokai]

this is all fascinating to me... at my local mom and pop computer store they told me upgrading my nvidia gtx 960 to an invidia rtx 2070 super would be fine for my amd a10-7850k cpu. however, when i checked the bottleneck calculator it tells me i am SEVERELY bottlenecking and would need to update which is an additional 500-1000 dollars more (motherboard/processor) depending on how crazy i want to get paying for it.

 

i notice when testing the card that im hovering around 55 frames on 1080p. but some of these youtube videos show this card pushing 100-125 frames. my card is a super OC'd out of the box. i even tried playing an older game last night (F.E.A.R) and it froze during the games start up. 

 

maybe it is the cpu?

[boggy77]

11 hours ago, Jerikokai said:

this is all fascinating to me... at my local mom and pop computer store they told me upgrading my nvidia gtx 960 to an invidia rtx 2070 super would be fine for my amd a10-7850k cpu. however, when i checked the bottleneck calculator it tells me i am SEVERELY bottlenecking and would need to update which is an additional 500-1000 dollars more (motherboard/processor) depending on how crazy i want to get paying for it.

 

i notice when testing the card that im hovering around 55 frames on 1080p. but some of these youtube videos show this card pushing 100-125 frames. my card is a super OC'd out of the box. i even tried playing an older game last night (F.E.A.R) and it froze during the games start up. 

 

maybe it is the cpu?

Monitor cpu and gpu usage during gaming. The one that hits 100% more often is the one that's best to upgrade first.

[Jerikokai]

ah ok i can do that. thanks!

[Radium_Angel]

6 hours ago, boggy77 said:

Monitor cpu and gpu usage during gaming.

Have you a suggestion for a software package that can overlay the game so you can see what's going on?

[boggy77]

16 minutes ago, Radium_Angel said:

Have you a suggestion for a software package that can overlay the game so you can see what's going on?

Msi afterburner

[anuith]

Just FYI from my own experience

 

I used to run a system of i5-6400 with GTX1060 for about 2 years and I thought it was fine. (At the time I've never had any knowledge about bottlenecking)

 

Until I try to play a game and cannot run it 1080p60 when all the YouTubers can with the same GPU. So I did some research until I found this particular calculator website.

 

It shows my CPU is 27% weak for this GPU.

 

So I upgrade to i5-9500F and it can handle the game really fine even though not all-Ultra. But A LOT better with 60fps almost all the time.

 

I think this is one kind of bottlenecking. And that "Does CPU X do better than CPU Y with GPU A?" should be called bottlenecking in some degree too.