Zen = Sandybridge IPC

Rroff · 18 May 2019 at 17:26

Only thing I can add is that I've benched a 2600 (3000MHz CL15) clock for clock against my 4820K and there is pretty much no difference framerate wise between the two in most games* - any difference is lost in 1-2 FPS. My 4820K will clock higher than typically a 2600(X) will which gives an edge in some games when you push for the max but then there are a small number of games that benefit from the additional cores.

* That I tried.

Pants · 18 May 2019 at 17:29

All other IPC tests are biased, but this is the truth™.

- Intel Fanboys

Locky · 18 May 2019 at 17:36

At the end of the day folks, AMD just isn't as good as Intel on the IPC or clock speeds - it's a budget chip.

~>Dg<~ · 18 May 2019 at 17:36

fanboy because it shows intel is faster in games ?

humbug · 18 May 2019 at 17:38

Pants said:
All other IPC tests are biased, but this is the truth™.

- Intel Fanboys

If Zen gets +15% IPC, i doubt it.... but assuming that, his next video will be about how AMD have finally caught up with Sandy Bridge IPC.

Something is just not right with this, its so far away from reality and so far away from what everyone else has resulted before him that i struggle to believe he believes his own charts, that he didn't look at his own charts and think "hang on, something is very seriously wrong here"

Locky · 18 May 2019 at 17:51

Why are the Ryzen users always last to load games in that we play regular? I have noticed it with a few lads I play with.

Rroff · 18 May 2019 at 17:58

Locky said:
Why are the Ryzen users always last to load games in that we play regular? I have noticed it with a few lads I play with.

Despite some claims otherwise - storage IOPS for some configurations on Ryzen is quite a bit behind an equivalent Intel platform - which can include the kind of patterns of data loading seen in some games. (It isn't always immediately apparent to the same degree in benchmarks as they typically focus on one scenario at a time while the loading of data for games can sprawl over many different scenarios with a mix of small and large datafiles, mix of streaming and more traditional access patterns, mix of sequential and random loading, etc.).

EDIT: I think this is improved a bit in the 2000 series BTW though don't have enough data to be sure - in my albeit fairly limited experience I noticed that the 2000 series systems seem to boot Windows noticeably faster than the older series which probably comes from improvements to storage access but might also be from driver improvements, etc.

humbug · 18 May 2019 at 18:01

Locky said:
Why are the Ryzen users always last to load games in that we play regular? I have noticed it with a few lads I play with.

The ASMedia Chipset that AMD use on 300 and 400 series Boards is not as good as Intel's, that much is true, its been tested.

AMD have developed their own Chipset for the 500 series boards.

Pants · 18 May 2019 at 18:06

~>Dg<~ said:
fanboy because it shows intel is faster in games ?

No, fanboy because you claim this to be "the reality" only because it shows what you want to see, rather than it being an objectively good test that is in consensus with other testers.

~>Dg<~ · 18 May 2019 at 18:07

Pants said:
No, fanboy because you claim this to be "the reality" only because it shows what you want to see, rather than it being an objectively good test that is in consensus with other testers.

think about what you said. then think what you put. you are doing the exact same. just in reverse. the only difference is you have proof showing you that intel is faster .

humbug · 18 May 2019 at 18:24

I'm not getting into his game results as i cannot verify them D.G, i do know Coffeelake is faster as a gaming CPU, what reviews i know of show Zen+ is faster in games than Sandy Bridge, i don't care about that so much.

what i do care about is what i can easily debunk, and i have, in that his Zen results are way under what they should be, that doesn't give me much confidence for the rest of his video, nor should it you.

Orange Nexus · 18 May 2019 at 18:26

Locky said:
At the end of the day folks, AMD just isn't as good as Intel on the IPC or clock speeds - it's a budget chip.

^this. My 1600 @ 3.9 v i7 2700K @ 4.5. No difference in gaming between the two. Except one is offline now cos of MDS

https://i.imgur.com/ghLLxVn.png

humbug · 18 May 2019 at 18:33

Orange Nexus said:
^this. My 1600 @ 3.9 v i7 2700K @ 4.5. No difference in gaming between the two. Except one is offline now cos of MDS

https://i.imgur.com/ghLLxVn.png

Is that R15 score with your sandy at 4.5Ghz? just asking because this would be the first time i have seen a comparison with Sandy outside his testing and there is a 600Mhz difference between the two.

Orange Nexus · 18 May 2019 at 18:37

humbug said:
Is that R15 score with your sandy at 4.5Ghz? just asking because this would be the first time i have seen a comparison with Sandy outside his testing and there is a 600Mhz difference between the two.

Yes. It is still running. But so are my Phenoms. Man, they last forever. Had to resurrect the Phenom to replace the Sandy. My system is use for work. But both are going come R3.

humbug · 18 May 2019 at 18:40

Orange Nexus said:
Yes. It is still running. But so are my Phenoms. Man, they last forever. Had to resurrect the Phenom to replace the Sandy. My system is use for work. But both are going come R3.

Ok, given the score and clock difference Zen has about 18% higher IPC than Sandy Bridge, not knocking it, Sandy was a great chip but yet again it flies in the face of Chris's result.

Pants · 18 May 2019 at 19:02

~>Dg<~ said:
think about what you said. then think what you put. you are doing the exact same. just in reverse. the only difference is you have proof showing you that intel is faster .

I'm doing nothing of the sort. I haven't claimed anything yet. My point is it's only proof when it's showing what you want, otherwise it's bias.

Why is this one proof and the trust rather than another one. It's because it shows what you want. Admit it.

~>Dg<~ · 18 May 2019 at 19:06

admit what ? its not showing what i want only the truth. everyone knows intel have faster gaming chips. this is just another benchmark which shows the reality. some just dont like its shows how a accient sandybridge cpus basically have the same gaming performance of ryzen. can i ask what cpu you use ? honestly ? i have ryzen and intel. you ?

Pants · 18 May 2019 at 19:25

~>Dg<~ said:
admit what ? its not showing what i want only the truth. everyone knows intel have faster gaming chips. this is just another benchmark which shows the reality. some just dont like its shows how a accient sandybridge cpus basically have the same gaming performance of ryzen. can i ask what cpu you use ? honestly ? i have ryzen and intel. you ?

Ryzen (1700) and Intel (still have an old 2500K but sold my 4770K, haswell macbook pro). If I get time later I'll fire up Cinebench at 3ghz and see if his setup looks OK or not. The 2500K is sitting in a box in the attic so can't easily spin that up unfortunately.

Pants · 18 May 2019 at 19:27

dual post

PieEater · 18 May 2019 at 20:39

Just run Cinebench on my 2700K @ 4.7GHz can someone run it on their 2700 / 2700X for comparison. Ram is DDR3 1600 at XMP settings.