Another Ryzen benchmark

[NumLock21]

Summit ridge is zen?

Ryzen is summit ridge

Zen don't exist?

What?

 

Summit ridge is code name

Zen is architecture

Ryzen is product name

What? x2

 

 

[Space Reptile]

2 minutes ago, NumLock21 said:

What? x2

[Techstorm970]

4 hours ago, drok said:

looks like that firestrike pic came from the future

 

1 hour ago, Zandvliet said:

Haha indeed, 2017-02-25  

BS benchmark confirmed!

[WereCat]

1 hour ago, AresKrieger said:

Why can't anyone just wait a week, I mean seriously this is getting absurd at this point.

Weak people can't wait another week.

[Guest Darth Revan]

 

[Orangeator]

3 hours ago, PianoPlayer88Key said:

Just out of curiosity, I did a few Cinebench single-threaded runs on my i7-6700K, plus one all-threads run.  This was right after restarting Windows 10.  Speeds were adjusted by reducing the maximum processor state percentage in Windows power options, and using CPU-Z to verify the correct speed.  (Nitpick: bus speed was usually 99.7x-99.9x and not 100, resulting in the actual clock speeds being a little less than indicated below.  I'm almost tempted to see about bumping the BCLK to like 101 MHz, if it'll let me.  The chipset is Z170, and the CPU is a K, so I'd think it should.  I just want to be careful, as it IS a laptop.)

 

• First run - 1 thread, 3.4 GHz (1700X stock speed) = 143
• Second run - 1 thread, 3.8 GHz (1700X default boost) = 162
• Third run - 1 thread, 4.0 GHz (What my laptop's 6700K chose to run) = 173
• Fourth Run - 8 threads, 4.0 GHz = 860

 

For obvious reasons I didn't do an 8c/16t run.

 

 

I'd like to know exactly what speed the R7-1700X was running for the single-threaded CineBench test.  For example, If it was 3.4 GHz, it would beat my Skylake CPU by 7.7% in IPC, if I'm calculating right.  If 3.8 GHz, my Skylake would beat it by 5.2%.  If it was boosting to, say, 4.0 GHz, then my 6700K beats it by 12.3%.

 

My Haswell i7-4790K desktop isn't powered up right now.  Linux is hibernated, so I don't really want to boot Windows on it yet, until I finish what I was working on.  Pulling stuff off old hard drives is taking longer than I expected, stretching into weeks or a month or more - not because the drives are dying yet, but because I have so many other things going on, some local storms prompting me to power off the PC for safety, etc.

So an 8 core 16 thread CPU is matching the 6700k (4 core 8 thread) in single-threaded performance? That doesn't sound right at all... If so then, damn AMD is stepping up their game. 

[Space Reptile]

1 minute ago, Tedny said:

only 3,8 Ghz ? 

ghz mean nothing , look at 3ghz pentium 4 vs 1.6ghz athlon 

[porina]

Time to over-analyse those Cinebench scores. I got some Intel numbers in past testing, and as it isn't much affected by ram speeds, it gives an indicator of the CPU itself.

Divide the multi-core by 8, compare it to single, you get an estimate of 25% gain due to SMT, assuming the core clock doesn't change between runs. I got 28% doing similar on Haswell and Skylake previously, so it is in the right ball park.

The next part is more dangerous, as it will depend on the actual CPU clock when it was running. Was it 3.5 GHz as shown in the Cinebench screenshot? Was it 3.9 from the Firestrike shot? Those screenshots do appear to be from the same original source, and I don't know what turbo/XFR or whatever might be going on. Assuming for now it was fixed at 3.5 GHz, that would put the Cinebench IPC pretty much on par with Skylake. If it was at 3.9 GHz to do so, that would put it closer to Haswell IPC. Either way, nothing to complain about!

[patrick3027]

Oh boy, another Ryzen thread... 

[Brehohn]

Because a blurred out picture DEFINITELY looks real. I hate it when people do that.

[Bananasplit_00]

5 hours ago, suicidalfranco said:

every LTT thread about nvidia, amd and intel

someone always ends up getting rekt to shit with facts in them. its the best

[PCGuy_5960]

4 hours ago, Darth Revan said:

 

Damn, you beat me to it!

[AnonymousGuy]

Please DICE, make a separate forum section for AMD posts.

[Coaxialgamer]

5 hours ago, techstorm970 said:

 

BS benchmark confirmed!

Not necessarily . System time might just be wrong . Or firestrike might be reporting it wrong . Idk .

A simple date discrepancy isn't enough to stop the hype train !

 

 

[Coaxialgamer]

On 21/02/2017 at 7:50 PM, patrick3027 said:

Oh boy, another Ryzen thread... 

oh god .

 

edit :

 

Quote

 

Is this what you want from me ?

 

 

[Techstorm970]

4 minutes ago, Coaxialgamer said:

LOL

[Hunter259]

8 hours ago, Morgan MLGman said:

It was probably boosting cause the 1700X has the XFR feature which overclocks the CPU beyond its Turbo frequency if the thermals and voltages are still low-enough, works similar to the GPU Boost 3.0 on Pascal.

 

But it seems like it's a bit over Haswell in terms of IPC meaning it's probably like Broadwell, so that's a good news if confirmed. This is all you need for gaming tbh in terms of IPC.

Lord even 3.5 Ghz Sandy Bridge is pretty much fine. Anything more from AMD would be fantastic.

[Notional]

I too like Firefly!

 

https://videocardz.com/66216/firefly-was-a-great-show