Jump to content

Will a 3960X bottleneck a next gen GPU?

LapX
Go to solution Solved by panther420,
12 minutes ago, Enderman said:

What do you not understand about "it's the same architecture"?

It is the same CPU as the 2600K with 2 cores 4 threads more and a higher TDP.

 

Maybe you weren't aware of this, but extreme edition CPUs have a model number that is 1000 above the consumer generation, despite having the exact same architecture.

A 3960X is not from the 3770K generation, it is from the 2600K generation.

 

This means that the single core performance is the same, and since very few games can use more than 8 threads, the gaming performance is practically identical.

Same with certain rendering programs which don't use many threads, such as adobe.

 

I guess you just weren't around back when these CPUs were just released?

Ok, I found some benchmarks on the 3DMark database.

https://www.3dmark.com/fs/13263681

This is a person with a stock 1080ti and stock 3960x running firestrike 1.1. They get a score of about 22000.

https://www.3dmark.com/fs/12155032

This is a person with a stock 1080ti and stock 7700k running firestrike 1.1. They get a score of about 23000.

 

Need I say much more?

12 minutes ago, panther420 said:

I have to say, I'm a little offended you think I'm that stupid.

 

Maybe you weren't aware of this, but you can't estimate performance of  two CPUs with a different core count and different platform entirely but the same architecture, ESPECIALLY if you are going to include a multithreaded program like adobe premiere. Adobe Premiere (cs6 and later) will make proper use of about 32 threads, and not much beyond that. And no, single core performance is not the same. You CANNOT simply make an assumption that the 3960X will have the same performance in games at 2600k. It might, and at this point I'm less concerned about the fact that the 3960X is not as good as I thought, but more of the fact that your logic is completely flawed. 

I don't think you understand how single thread performance works...

If it's the same generation and same clock speed, then the single core performance is almost identical.

Image result for 3960x 2600k game benchmarks

Yeah the two extra cores make a bit of a difference in certain games, but the majority of games do not use more than 8 threads so there is no improvement there.

 

Why do you think that nobody recommends the 6950X or other i9 extreme edition CPUs for gaming?

 

I think you might want to do some research on how single thread performance affects gaming.

 

1 minute ago, panther420 said:

Ok, I found some benchmarks on the 3DMark database.

https://www.3dmark.com/fs/13263681

This is a person with a stock 1080ti and stock 3960x running firestrike 1.1. They get a score of about 22000.

https://www.3dmark.com/fs/12155032

This is a person with a stock 1080ti and stock 7700k running firestrike 1.1. They get a score of about 23000.

 

Need I say much more?

Artificial benchmarks, unlike most games, can often take advantage of more than 8 cores.

Also, firestrike is a GPU heavy benchmark, so of course the two systems will perform similarly with the same GPU.

Again, go read what I said in my first post about the type of bottlenecking.

Not only that, but the 8700k has better single core AND multithreaded performance with 4 extra threads than the 7700k.

NEW PC build: Blank Heaven   minimalist white and black PC     Old S340 build log "White Heaven"        The "LIGHTCANON" flashlight build log        Project AntiRoll (prototype)        Custom speaker project

Spoiler

Ryzen 3950X | AMD Vega Frontier Edition | ASUS X570 Pro WS | Corsair Vengeance LPX 64GB | NZXT H500 | Seasonic Prime Fanless TX-700 | Custom loop | Coolermaster SK630 White | Logitech MX Master 2S | Samsung 980 Pro 1TB + 970 Pro 512GB | Samsung 58" 4k TV | Scarlett 2i4 | 2x AT2020

 

Link to comment
Share on other sites

Link to post
Share on other sites

13 minutes ago, Enderman said:

I don't think you understand how single thread performance works...

If it's the same generation and same clock speed, then the single core performance is almost identical.

Image result for 3960x 2600k game benchmarks

Yeah the two extra cores make a bit of a difference in certain games, but the majority of games do not use more than 8 threads so there is no improvement there.

 

Why do you think that nobody recommends the 6950X or other i9 extreme edition CPUs for gaming?

 

I think you might want to do some research on how single thread performance affects gaming.

 

Artificial benchmarks, unlike most games, can often take advantage of more than 8 cores.

Also, firestrike is a GPU heavy benchmark, so of course the two systems will perform similarly with the same GPU.

Again, go read what I said in my first post about the type of bottlenecking.

Not only that, but the 8700k has better single core AND multithreaded performance with 4 extra threads than the 7700k.

The points remains though that the 3960X will not bottleneck a 1080ti, not really even in the sense that you said. OP should not upgrade his X79 system, he should really keep it for a little longer, it's stupid to upgrade to a 7700k or 8700k right now.

 

EDIT: You also mentioned in your original post with benchmarks a multithreaded program, adobe premiere. You're now shifting the topic about single threaded. Including benchmarks in adobe premiere for the 2600k only and not the 3960x doesn't make any sense, and as you can see in the Cinebench benchmarks you just posted, there is a multithreaded difference between the 2600k and 3960x.

Spoiler

My main desktop, "Rufus":

Spoiler

PC Specs:

CPU: AMD Ryzen 5 1600

CPU Cooler: Cooler Master MasterLiquid Lite 120

RAM: 2x8gb Corsair Vengence DDR4 Red LED @ 3066mt/s

Motherboard: MSI B350 Gaming Pro Carbon

GPU: XFX RX 580 GTR XXX White 

Storage: Mushkin ECO3 256GB SATA3 SSD + Some hitachi thing

PSU: Seasonic Focus Plus Gold 650W

Case: Corsair Crystal 460X

OS: Windows 10 x64 Pro Version 1607

Retro machine:

Spoiler

PC Specs:

CPU: Intel Core 2 Quad Q9550

CPU Cooler: Stock heatsink

RAM: GSkill 4gb DDR2 1066mt/s

Motherboard: Asus P5n-e SLI

GPU: 8800 GTS 640mb, I swap between that and my 8800 GTS 512mb

Storage: Seagate 320gb right from 2006

PSU: Ultra 600W 

Case: Deepcool Tesseract SW

OS: Windows XP SP3 32-bit, Linux Mint 18.2 Cinnamon 64-bit, Manjaro Deepin x64 (sorta)

Mac Pro Early 2008: Dual Xeon X5482s w/ 32GB RAM & HD 5770 running macOS High Sierra

More PC's

 

Link to comment
Share on other sites

Link to post
Share on other sites

14 minutes ago, panther420 said:

Ok, I found some benchmarks on the 3DMark database.

https://www.3dmark.com/fs/13263681

This is a person with a stock 1080ti and stock 3960x running firestrike 1.1. They get a score of about 22000.

https://www.3dmark.com/fs/12155032

This is a person with a stock 1080ti and stock 7700k running firestrike 1.1. They get a score of about 23000.

 

Need I say much more?

Wow really ? It's hard to find benchmarks with my CPU !

CPU : Intel Core i7 3960X, Mobo : X79-UD3, Memory : 4x4GB Vengeance Black Memory 1600MHz, GPU : Asus GTX 970 Strix, Case : Switch 810 Matte Black, Storage : 256GB Samsung 830 SSD, Seagate Barracuda 1TB, PSU : Thermaltake 750W 80+ Bronze, Displays : 3x Asus 1080p Screens, Cooling : Corsair H100i, Keyboard : Logitech G710+, Mouse : Madcat Cyborg R.A.T.7, Sound : Sennheiser HD598, V-Moda Crossfade LP, Logitech Z-5500, HMD : Oculus Rift CV1

Link to comment
Share on other sites

Link to post
Share on other sites

3 minutes ago, panther420 said:

The points remains though that the 3960X will not bottleneck a 1080ti, not really even in the sense that you said. OP should not upgrade his X79 system, he should really keep it for a little longer, it's stupid to upgrade to a 7700k or 8700k right now.

Wow, it's almost as if you didn't even read my first post.

Maybe let's try again:

 

Well, "bottlenecking" as in you could be getting more performance with a better CPU.

Not "bottlenecking" as in you won't see a performance difference with a better GPU.

 

Most people don't understand that both the CPU and the GPU contribute to the final performance.

 

 

Your firestrike benchmark pretty much proved what I am saying, although the difference in real games will be more because they rely more on single thread performance and aren't just a GPU test.

NEW PC build: Blank Heaven   minimalist white and black PC     Old S340 build log "White Heaven"        The "LIGHTCANON" flashlight build log        Project AntiRoll (prototype)        Custom speaker project

Spoiler

Ryzen 3950X | AMD Vega Frontier Edition | ASUS X570 Pro WS | Corsair Vengeance LPX 64GB | NZXT H500 | Seasonic Prime Fanless TX-700 | Custom loop | Coolermaster SK630 White | Logitech MX Master 2S | Samsung 980 Pro 1TB + 970 Pro 512GB | Samsung 58" 4k TV | Scarlett 2i4 | 2x AT2020

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, Enderman said:

Wow, it's almost as if you didn't even read my first post.

Maybe let's try again:

 

Well, "bottlenecking" as in you could be getting more performance with a better CPU.

Not "bottlenecking" as in you won't see a performance difference with a better GPU.

 

Most people don't understand that both the CPU and the GPU contribute to the final performance.

 

 

Your firestrike benchmark pretty much proved what I am saying, although the difference in real games will be more because they rely more on single thread performance and aren't just a GPU test.

Fair enough. I still think that @LapX should not upgrade his system right now. He will still get a very good gaming experience with the 3960X and 1080ti, especially at higher resolutions.

 

Have a nice day.

Spoiler

My main desktop, "Rufus":

Spoiler

PC Specs:

CPU: AMD Ryzen 5 1600

CPU Cooler: Cooler Master MasterLiquid Lite 120

RAM: 2x8gb Corsair Vengence DDR4 Red LED @ 3066mt/s

Motherboard: MSI B350 Gaming Pro Carbon

GPU: XFX RX 580 GTR XXX White 

Storage: Mushkin ECO3 256GB SATA3 SSD + Some hitachi thing

PSU: Seasonic Focus Plus Gold 650W

Case: Corsair Crystal 460X

OS: Windows 10 x64 Pro Version 1607

Retro machine:

Spoiler

PC Specs:

CPU: Intel Core 2 Quad Q9550

CPU Cooler: Stock heatsink

RAM: GSkill 4gb DDR2 1066mt/s

Motherboard: Asus P5n-e SLI

GPU: 8800 GTS 640mb, I swap between that and my 8800 GTS 512mb

Storage: Seagate 320gb right from 2006

PSU: Ultra 600W 

Case: Deepcool Tesseract SW

OS: Windows XP SP3 32-bit, Linux Mint 18.2 Cinnamon 64-bit, Manjaro Deepin x64 (sorta)

Mac Pro Early 2008: Dual Xeon X5482s w/ 32GB RAM & HD 5770 running macOS High Sierra

More PC's

 

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×