AMD Ryzen - Games in 1080p can improve more than 35% !

[PlayStation 2]

I'm going to be completely honest here: it seems really pointless to use F1 2016 as a basis for benchmarking.

[SpaceGhostC2C]

7 minutes ago, Hunter259 said:

It's turned off due to a sleep bug more or less. The bug causes Windows timing to slow down which artificially inflates benchmark scores. HPET also can cause a very small hit to performance in certain cases. AMD just said to do that due to the bug and trying to gain every last drop of performance they can.

The sleep bug is on top of the HPET problem.

 

6 minutes ago, Dan Castellaneta said:

I'm going to be completely honest here: it seems really pointless to use F1 2016 as a basis for benchmarking.

If you ask me, it seems really pointless to use games as a basis for benchmarking  

Of course, it can always help to see a benchmark for a game you intend to play. I still find it strange for gaming performance to influence the purchase decision of anyone looking into the X99 platform or R7s.

[zMeul]

30 minutes ago, SpaceGhostC2C said:

Apparently you haven't been following the topic: they've been recommending this, and explaining the reason, from day release day (essentially, software-based clock management interferes with the CPU's own clock management). That's why using "High performance" removes the need to switch HPET off (or at least in this test).

mno!

HPET and power management (high performance) have absolutely nothing in common - nothing

don't understand what HPET does? do a google search - turning HPET off is a very very stupid idea

[N3v3r3nding_N3wb]

11 hours ago, stconquest said:

So devs are to build games for the future hardware that is not available to them? 

 

Anyways, give Ryzen some time to mature as there might be some nice surprises.

I mean, if it had this problem, it could have a problem with going between, say, a 7700k and a 6900k.

[leadeater]

1 hour ago, Misanthrope said:

If I have only one point of contention it is not a misquote. If it suits this new "rule" that wants to turn LTT Forums into endless, pointless walls of text and scrolling through unreadable replies I'll limit myself to bolding the parts I am replying to, you can assume I concede or have no comment to the parts I do not reply to then.

Well I will apologize for the abruptness of my reply, was just leaving for work. However the very next few words from the quote you took of me had the same view you were pointing out. That is a misquote:

 

misquote

mɪsˈkwəʊt/

verb

 

1. 1.

   quote (a person or a piece of written or spoken text) inaccurately.

   "the government insisted that the official had been misquoted"

   synonyms:

   misreport, misrepresent, misstate, quote incorrectly, take/quote out of context, distort, twist, slant, bias, put a spin on, pervert, falsify, garble, muddle, mistranslate; raremisrender "my original statement has been misquoted"

noun

 

1. 1.

   a passage or remark quoted inaccurately.

 

I have a few suggestions

• If you can't be bothered reading to the end of the sentence don't reply
• The post had a point, the same point you made. Continuing to read for your own benefit would have helped, there isn't any excuse of such laziness
• Not bothering to read ones point then creating an entire reply railing against them is rather disrespectful since your entire point you were trying to make was the the very same one.
• If you are going to try and make a point make sure it has some merit to it and evidence otherwise it just looks foolish.

 

In my view what you did is unacceptable, not from a moderation point of view but just common decency. 

[N3v3r3nding_N3wb]

10 hours ago, zMeul said:

please do

 

I think it's only a 2.89 percent performance increase due to this game optimization, if my math is right. Is that what you get too?

 

Actually, it might be 19.57% minus whatever the performance gain of engaging Windows high performance mode is.

[PCGuy_5960]

Just now, N3v3r3nding_N3wb said:

I think it's only a 2.89 percent performance increase due to this game optimization, if my math is right. Is that what you get too?

If you look closely... Ryzen got a huge increase when they overclocked the infinity fabric by using faster RAM....

[zMeul]

Just now, N3v3r3nding_N3wb said:

I think it's only a 2.89 percent performance increase due to this game optimization, if my math is right. Is that what you get too?

 

Actually, it might be 19.57% minus whatever the performance gain of engaging Windows high performance mode is.

why I would even bother crunching the numbers since the whole idea revolves around OCing the CPU

it's BS PR, and people that don't get it deserve what's coming for them

[Dabombinable]

17 minutes ago, zMeul said:

mno!

HPET and power management (high performance) have absolutely nothing in common - nothing

don't understand what HPET does? do a google search - turning HPET off is a very very stupid idea

I thought that my computer was relying on the RTC:

[PCGuy_5960]

1 minute ago, zMeul said:

why I would even bother crunching the numbers since the whole idea revolves around OCing the CPU

Or overclocking the infinity fabric, in other words the bridge between the 2 quad core modules in the CPU.....

[Dabombinable]

Those going on about HPET, I might disabling this myself and see what happens, as its currently enabled in my rig (4790K+GTX 970+16GB DDR3 1333):

https://forums.anandtech.com/threads/do-you-have-hpet-enabled-or-disabled.2362109/

[N3v3r3nding_N3wb]

3 minutes ago, zMeul said:

why I would even bother crunching the numbers since the whole idea revolves around OCing the CPU

it's BS PR, and people that don't get it deserve what's coming for them

I was mostly referring to this point in the graph, where they show no clock speed changes.  That shows that moving from a reused OS and game install to clean ones and moving to Windows high performance mode increases the performance by 19.57% (18.5 + 1.07 = 19.57).

[zMeul]

9 minutes ago, PCGuy_5960 said:

Or overclocking the infinity fabric, in other words the bridge between the 2 quad core modules in the CPU.....

call it wtf you want and be as vague as possible

when you use higher speed RAM profiles you OC the memory controller(s)

[PCGuy_5960]

Just now, zMeul said:

when you use higher speed RAM profiles you OC the memory controller(s)

And in Ryzen you also OC the Infinity Fabric. Which VASTLY IMPROVES performance as it lowers the latency between cores in different CCXs....

[zMeul]

5 minutes ago, N3v3r3nding_N3wb said:

I was mostly referring to this point in the graph, where they show no clock speed changes.  That shows that moving from a reused OS and game install to clean ones and moving to Windows high performance mode increases the performance by 19.57% (18.5 + 1.07 = 19.57).

games should be tested with a clean install, especially since they tend to leave behind residual config files and sometimes even compiled shaders from different HW

who tested it with old installs should slap themselves in the head

[zMeul]

1 minute ago, PCGuy_5960 said:

And in Ryzen you also OC the Infinity Fabric. Which VASTLY IMPROVES performance as it lowers the latency between cores in different CCXs....

they should call it "the cloud"

wtf infinity fabric does even mean!?!? that's why I asked if anyone has knowledge of a good Zen CPU block diagram

[PCGuy_5960]

Just now, zMeul said:

games should be tested with a clean install, especially since they tend to leave behind residual config files and sometimes even compiled shaders from different HW

And 99% of reviewers used a clean install of Windows. AMD, what was the point of this chart?

[zMeul]

Just now, PCGuy_5960 said:

And 99% of reviewers used a clean install of Windows. AMD, what was the point of this chart?

not clean Windows, clean game installs .........

[PCGuy_5960]

1 minute ago, zMeul said:

not clean Windows, clean game installs .........

And clean installs of Windows....

EDIT: Reviewers probably used clean installs of games as well...

Edited March 14, 2017 by PCGuy_5960

[zMeul]

1 minute ago, PCGuy_5960 said:

And clean installs of Windows....

that's a given

[PCGuy_5960]

8 minutes ago, zMeul said:

they should call it "the cloud"

wtf infinity fabric does even mean!?!? that's why I asked if anyone has knowledge of a good Zen CPU block diagram

If you take a look at the Ryzen die, you will notice that there are 2 quad core diesCCXs. The infinity fabric is what connects the 2...

[Coaxialgamer]

12 hours ago, huilun02 said:

Inb4 someone comes in saying its a shit game and isolated case. 

 

11 hours ago, Droidbot said:

Cherry pick, cherry pick, cherry pick. 

Love AMD's marketing team

 

[N3v3r3nding_N3wb]

2 minutes ago, zMeul said:

they should call it "the cloud"

wtf infinity fabric does even mean!?!? that's why I asked if anyone has knowledge of a good Zen CPU block diagram

http://semiaccurate.com/2017/01/19/amd-infinity-fabric-underpins-everything-will-make/

 

That's the best picture I could find in a few minutes.

[zMeul]

1 minute ago, N3v3r3nding_N3wb said:

http://semiaccurate.com/2017/01/19/amd-infinity-fabric-underpins-everything-will-make/

 

That's the best picture I could find in a few minutes.

as I said in the other thread, that's crap - it doesn't say anything

I'm searching something like this:

Spoiler

^ that's Bulldozer btw

[N3v3r3nding_N3wb]

3 minutes ago, PCGuy_5960 said:

If you take a look at the Ryzen die, you will notice that there are 2 quad core dies. The infinity fabric is what connects the 2...

Well, it's a single die, just with two CCX on it.