For most things is the same really cannot tell the difference.
In certain games tho where things dont run 100% smooth, the bits that arent smooth are not as bad. Obviously could be placebo so I did run frame analysis also which showed higher min's.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***
- Thread starter Jamin280672
- Start date
For most things is the same really cannot tell the difference.
In certain games tho where things dont run 100% smooth, the bits that arent smooth are not as bad. Obviously could be placebo so I did run frame analysis also which showed higher min's.
Associate
- Joined
- 21 Sep 2018
- Posts
- 895
i just chanced upon THG's 2700x review again and it got me thinking...if we're waaaay overhyping ryzen 3000 with respect to performance (i suspect performance/watt is likely to be better than intel though)
now from the tech press rumours, the chip demoed at computex was 8c ryzen 3000 chip, rumoured to be at 4.5ghz?
from THG:
from computex (courtesy: https://www.purepc.pl/procesory/amd_ryzen_3000_navi_i_x570_zmasowany_atak_juz_lipcu):
now if ryzen 2700x at 4.3ghz scores 1930-1935 cb @ 4.3ghz (from THG)
and ryzen 3000 8c scores 2057 (say 2055-2060) cb @ 4.5ghz (from computex)
then 2057/1933*4.3 = 4.58ghz
so according to my man-maths, ipc doesn't appear to have changed that significantly, and definitely not as much as the 10-15% asclaimedrumoured...
(performance/watt is better though, no doubt about that - 2057cb @ 133w vs 1933cb @ 135w)
happy to be pointed in the right direction though, if i'm just making **** up lol
Not sure how Gen 2 is in Cine MT but GEN 1 and 1+ run MT, at stock, closer to the base clock not boost clock. So, that ES could have been running closer to 4GHz. My 2700 at stock runs at 3.3GHz in Cine MT and only gets 1550 pts. Does get 1900 @ 4.1GHz but with 3533MHz CL14 ram.
A better comparison is Cine ST.
Last edited:
Associate
- Joined
- 27 Apr 2007
- Posts
- 966
Your just repeating what was not necessary in the first place as if you are adding to the conversation when you were just criticising randomly for no good reason.Not really an observation, more of a comment.
The context was gotten, i was replying on the simple fact that people are getting hung up on higher numbers.
The point in these chips, why we all do this, is better performance. More fps in games, quicker renders, etc.
Whilst i dont disagree that for marketing high numbers would be nice, at the end of the day, when it comes down to it, if AMD released the next Zen at the same clock speed but it was 20-30% quicker, would we really care?
Whatever way performance comes we'll either laud AMD for improving clocks, laud them for getting more performance per clock, or criticise them for missing out on keeping their momentum going.
People who get hung up on bigger numbers, need to calm down and take a step back, as long as we get the performance, the methods are moot.
Lastly, the attitude given was not constructive to the conversation.
Pot, kettle, black.
Associate
- Joined
- 5 May 2017
- Posts
- 142
Here you can see the difference.
https://youtu.be/RmxkpTtwx1k
Look at 3:03. Ryzen 3000 shows a impressive improvements in comparison with ryzen 2000.
Last edited:
https://www.tomshardware.com/news/userbenchmark-benchmark-change-criticism-amd-intel,40032.html
So UserBenchmark decided to change the way it scores the CPUs because third-generation Ryzen was "scoring too high."
Fortunately, I captured a few screenshots before it was changed, so we can see the difference:
The way they come across in that article, and on their own site, pretty much strips them of any respect that I had for them. Thank the Lord for alternative services, that's all I will say.
So UserBenchmark decided to change the way it scores the CPUs because third-generation Ryzen was "scoring too high."
Fortunately, I captured a few screenshots before it was changed, so we can see the difference:
The way they come across in that article, and on their own site, pretty much strips them of any respect that I had for them. Thank the Lord for alternative services, that's all I will say.
Last edited:
https://www.tomshardware.com/news/userbenchmark-benchmark-change-criticism-amd-intel,40032.html
So UserBenchmark decided to change the way it scores the CPUs because third-generation Ryzen was "scoring too high."
Fortunately, I captured a few screenshots before it was changed, so we can see the difference:
The way they come across in that article, and on their own site, pretty much strips them of any respect that I had for them. Thank the Lord for alternative services, that's all I will say.
Since intel can't stay competitive with innovation/new products/competitive pricing, they try new dirty tricks from the circus in order to pretend they stay competitive. Poor artists...
Since intel can't stay competitive with innovation/new products/competitive pricing, they try new dirty tricks from the circus in order to pretend they stay competitive. Poor artists...
You know what this change results in, right?
Though the i5-8400 will destroy the i3-9350KF in pretty much any game, the new system being heavily weighted towards single- and quad-core performance results in the effective speed of the i3 processor being 14% higher. Absolutely ludicrous.
I never paid attention to these effective speed numbers in the first place, but that will be tenfold now. This reminds me of CPUBoss where it recommended a mobile dual-core Core i7 over a desktop quad-core variant because it was more power efficient. LOL! (Seems to be fixed now, but it used to recommend the i7-4500U because it was more efficient.)
It's websites like these that are the cause for new alternatives.
It used to be helpful for a very quick comparison, but now it's utterly worthless. I suspect they will see a rather large drop in traffic as a result of this nonsensical change. I hope the brown envelope from Intel was worth it!
It used to be helpful for a very quick comparison, but now it's utterly worthless. I suspect they will see a rather large drop in traffic as a result of this nonsensical change. I hope the brown envelope from Intel was worth it!
Kinda funny to see Tom's Hardware going out against the change too.
Also, anyone else notice the sheer upvote/downvote difference between the two screenshots? The "old system" screenshots were taken just over a month ago. Yikes.
ok gamers nexus tested undervolting on ryzen 3000.
When he did a manual large undervolt, the cpu did slow down internally (whilst reporting normal clocks).
However when he did a small negative offset it "improved" his 3900x performance as expected.
https://www.gamersnexus.net/guides/3494-amd-ryzen-3000-undervolting-offset-override
When he did a manual large undervolt, the cpu did slow down internally (whilst reporting normal clocks).
However when he did a small negative offset it "improved" his 3900x performance as expected.
https://www.gamersnexus.net/guides/3494-amd-ryzen-3000-undervolting-offset-override
Caporegime
- Joined
- 8 Jan 2004
- Posts
- 32,560
- Location
- Rutland
I really like GNs work, answering some important questions.
https://www.tomshardware.com/news/userbenchmark-benchmark-change-criticism-amd-intel,40032.html
So UserBenchmark decided to change the way it scores the CPUs because third-generation Ryzen was "scoring too high."
Fortunately, I captured a few screenshots before it was changed, so we can see the difference:
The way they come across in that article, and on their own site, pretty much strips them of any respect that I had for them. Thank the Lord for alternative services, that's all I will say.
I am not the only one to notice this before I get accused of been a intel fanboy, but there seems to be tendancy on OCUK forum, that if some happens that is considered harmful to AMD, then the entity is in intel's pocket, but if its the other way round nothing is said. In short I am saying be rational and try to escape the brand loyalty. Userbench gave an explanation, if you read it, take it in, its rational. Ryzen 3000 chips by the way did go up on the cpu ranking charts from the change, certain intel and amd chips did go down, the idea was to prevent cpu's with many cores but poor per core performance been considered faster than cpus with faster cores but less of them. I think what userbench is doing tho is going to be controversial simply because of the fact every individual has their own opinion on how important multi core performance is, and clearly they never going to satisfy everyone.
So my advice really instead of just saying userbench is now 100% trash and useless, look at the two images you just posted and you will realise the only actual difference of note is the "effective speed", the rest is the same type of data. Me personally I never used that effective speed metric before on userbench anyway, I only ever looked at the individual pieces of data below it so for me it hasnt changed.
I think trying to claim a 20 core 2ghz xeon is faster than a 5ghz 8600k for gaming isnt right, and thats why they changed their effective speed metrics.
It seems to me almost everyone on here just wants every website to be blurting out "AMD is super, intel sucks", then they happy. Also even a OCUK forum staff member (rank of don) has noticed the behaviour in another thread they spoke out as well against the AMD fanboyism.
Where we both agree is it is good there is multiple websites. To me userbench is the absolute best one as it includes user data, includes data for pretty much every cpu on the market, it has single core data, it doesnt consider multicore to be the be all and end all, so I like it. So for me I am glad I have it as an alternative to the likes of hardware unboxed and other content creators who seem to think multi core is all that matters. So if you dont like it dont use it. But I am glad they not sheep falling in line to the rest of the media. Note my post doesnt mean I agree with their effective speed system by the way, I think 2% for all core is too low, and I think quad score shouldnt even be a contributor, but it doesnt matter as I never used their cpu ranking or effective speed scoring anyway.
Except that their explanation is false and lies, at the same time.
And this is stupid. It's better to have 16 Bulldozer-type cores than 4 Zen 2-type cores.
yes when he does stuff like this its quality.
Mate, my very next post is comparing two Intel processors, showing how the new change makes no sense whatsoever. I have no reason to believe Intel is behind this.
What I am seeing, is that two benchmarks have now been modified because the developers deemed Ryzen to be performing better than expected. So, rather than allow the architecture to play on its strengths, they knock it down to artificially gimp the difference.
Soldato
- Joined
- 31 May 2009
- Posts
- 21,456
Effective speed is the only thing they have changed, they have trashed that variable, therefore their leading variable is trash.
Might as well get all the other information elsewhere, where they don't muddy the waters in hilarious fashion.
Might as well get all the other information elsewhere, where they don't muddy the waters in hilarious fashion.
Lies is peculation on your own part formed from where I dont know.
They clearly stated most of their viewers are gamers, and hence they trying to rank based on suitability for gaming, on average a 2ghz xeon isnt going to be better than a 4ghz+ 8600k for gaming.
except the value for the site is the other data not "effective speed" if thats all you was using the site for then you was using it for the wrong reasons. The only time I can see people putting value in that figure is if they just really lazy and not wanting to examine the actual proper data.
Their effective speed was "always" formed of opinion anyway.
Its the same thing with reviewers, I nearly always ignore their opinions but I do take in their data.
What are the reasons you consider it trashed by the way?
is it?
because it gives a lower effective speed o your favourite cpu?
the fact it changed at all so isnt reliable?
you disagree with the weighting of single core?
Also I am curious why you consider the actual real data useless, do you only read sites for their opinions and not data?
To you its their leading variable, to me its irrelevant, its just based of an opinion on how they translate the data into ranking.
Also what alternatives is there to userbench that fulfill these requirements?
Must provide single and multi core performance data.
Floating and integer performance.
Provide data for every cpu on the market.
Provide market share data based on number of samples tested.
Provide basic spec info on the same page for quick glance access.
Last edited:
