Ryzen "2" ?

[humbug]

Agreed, now look at the single. Lets try with SMT/HT off

So Coffeelake has 1.2% higher IPC, well what do you know i was wrong, by 1.2%, is that your Mhz scaling that's made that 1.2% difference in what i said? which to be clear was "About the same IPC"

Sure, you win.

At least there is no more ambiguity about this.

 

[gavinh87]

So Coffeelake has 1.2% higher IPC, well what do you know i was wrong, by 1.2%, is that your Mhz scaling that's made that 1.2% difference in what i said? which to be clear was "About the same IPC"

Sure, you win.

So we all ignore the single core difference despite being clocked the same........

You need to put a disclaimer with that claim. Something like this. *In games and programs where ALL threads are used*
We've all seen that SMT is better than HT.

Now.... that heaven benchmark?

 

[humbug]

So we all ignore the single core difference despite being clocked the same........

You need to put a disclaimer with that claim. Something like this. *In games and programs where ALL threads are used*
We've all seen that SMT is better than HT.

Now.... that heaven benchmark?

I got what i wanted, maybe later, or maybe i don't see the need for it.

 

[DragonQ]

SMT/HT aren't used until all physical cores are used.

A cursory glance at Task Manager suggests the opposite. Linux's thread management might be more sensible though.

 

[gavinh87]

I got what i wanted, maybe later, or maybe i don't see the need for it.

Just see it as decency, given how I did similar for you

 

[DragonQ]

It's a loaded question I'll admit. But that doesn't make it invalid. You can't pretend those scenarios don't exist, it's a statistical fallacy to suggest every CPU intensive application is now capable of using more than 6 threads (Which is what you're doing by dismissing my question)

I'm not dismissing anything though, SMT's a good thing, I've said nothing contrary to that. But there's no "use all threads" button in software.

As far as Ryzen goes I put my money where my mouth was over half a year ago. But I'm not going to pretend it's better than it is.

Are we seriously having an argument about what would win in a 6 core race between a 6c/6t chip and a 6c/12t chip? What waste of time. We all know Intel has higher single threaded IPC in most scenarios. In fact let's call it IPCPT from now on.

 

[DragonQ]

We've all seen that SMT is better than HT.

HT is SMT.

 

[gavinh87]

HT is SMT.

AMD's implementation of it is better. Happy now?

 

[DragonQ]

AMD's implementation of it is better. Happy now?

Let's face it, all of this IPC and thread talk is (sometimes) fun and interesting but in the end all that people care about is (a) how does it perform in applications X and Y at stock, and optionally (b) how does it perform in applications X and Y with a reasonable overclock? The end user doesn't care if an application uses 1, 4, or 50 threads. It's all application specific; a game that uses 16 threads might still be faster on a 4c/4t Intel chip at 5 GHz if one thread is the bottleneck.

 

[humbug]

Just see it as decency, given how I did similar for you

Well alright, if we are going to be more civil with eachother i'll do one later, when i have the time, i think it will be around 160, not that it matters, i have always agreed that without SMT Intel is quicker.

 

[gavinh87]

Well alright, if we are going to be more civil with eachother i'll do one later, when i have the time, i think it will be around 160, not that it matters, i have always agreed that without SMT Intel is quicker.

We talking about the same thing? Heaven?
Yes we can agree to do that.

 

[gavinh87]

Let's face it, all of this IPC and thread talk is (sometimes) fun and interesting but in the end all that people care about is (a) how does it perform in applications X and Y at stock, and optionally (b) how does it perform in applications X and Y with a reasonable overclock? The end user doesn't care if an application uses 1, 4, or 50 threads. It's all application specific; a game that uses 16 threads might still be faster on a 4c/4t Intel chip at 5 GHz if one thread is the bottleneck.

No point in a forum if you can't discuss the boring stuff that normal people don't care about!

 

[humbug]

We talking about the same thing? Heaven?
Yes we can agree to do that.

Heaven, never run it on this chip but ok that too then.

 

[gavinh87]

Heaven, never run it on this chip but ok that too then.

Sweet, cheers!

 

[Martini1991]

Are we seriously having an argument about what would win in a 6 core race between a 6c/6t chip and a 6c/12t chip? What waste of time. We all know Intel has higher single threaded IPC in most scenarios. In fact let's call it IPCPT from now on.

This doesn't make any sense.
Firstly that's not what was said.
Secondly, I wish people would stop saying single threaded. It's not accurate. An application can use 4 threads would show the same result for a CPU with higher performing cores, despite it being multithreaded.

 

[Martini1991]

A cursory glance at Task Manager suggests the opposite. Linux's thread management might be more sensible though.

Using higher performing threads first is the most logical thing when it comes to performance? So using physical cores before moving onto the logical cores.

If an application utilused 6 threads but was using 3 physical and 3 logical it would perform worse than a cpu with 6 cores with no logical cores

 

[gavinh87]

Old school single threaded engine results

 

[humbug]

I just ran this, stock 1070, performance is right where i was expecting it, why did i run this?

 

[gavinh87]

I just ran this, stock 1070, performance is right where i was expecting it, why did i run this?

Thats not 720P All you are testing there is the GPU.

 

[Doobedoo]

So the 8700k is 8% faster than the 1600 in single core at the same speed and 1.2% higher in multi core at the same speed.

That's a lot closer than some would have you believe and quite remarkable for a chip that cost's half as much.