Associate
- Joined
- 20 May 2019
- Posts
- 505
- Location
- London
First of all let me say that I am glad to see new Ryzens coming out and they indeed look like a solid chips especially 12 and 16 cores looks excellent. If I would build a system now, it would be the one capable of running 16 core chip with placeholder chip for now.
With that being said, I don't understand what's going on with the reviewers. Either they are shilling AMD hard or just are clueless, let me explain.
When principle technologies came out with the Intel review everybody jumped do defend that the AMD wasn't tested as they should be, not that mode, not that cooler etc.
So I don't understand what's going on now with themselves.
All of them as one are saying that new Ryzens are running pretty much as fast as it can straight out of the box. And there's no overclocking headroom. Hmm, alright fair enough. Second as "Hardware Unboxed" demonstrated it doesn't really matter what ram are you using 3200 C14 or 3600 C16 as it will run the same. Which makes me think that even if Ryzen can do 3600 it benefits as much from lower latency so 3200 cl14 will do just fine.
So if you are letting AMD boost up by itself as fast as it will go and add a 3200 cl14 ram then it will be reaching its ceilings.
Fair enough.
THEN WHY THE HELL YOU ARE COMPARING IT TO STOCK INTEL CHIPS ? Especially 9700k running at 4.6 all cores at stock. Seriously guys.
So we have established that AMD are being tested at their peak performance.
Let jump on to the side of Intel.
First of all, there is one group of people on the internet who say : ram speed or latency doesn't matter for Intel. Oh yes it does ! Maybe not in all instances but it definitely does !
Hardware unboxed made a video about this titled : Is Fast RAM A Waste? Unleashing the Core i9-9900K with DDR4-4000 / Link Here :
https://www.youtube.com/watch?v=VElMNPXJtuA
So to test Intel near to it's limits like you are doing with AMD it should at least have 4000 CL 17 ram. That's point number one.
Second point, 9900k and 9700k should be tested at 5Ghz synced on all cores. I am running 5.2 and only reason I am not running 5.3 is because my 240 AIO can't cool it. Alright I am not saying you should test with 5.2 but the golden average 5 Ghz synced on all cores should be a default testing method.
Third point. Uncore or Core cache ratio, I can guarantee you that most of these reviewers are testing intel chips with 4300 mhz cache ratio, for example mine runs at 4900 mhz, again I am not requesting that tests have to be done with 4900 mhz, but 4700 mhz should be doable for all chips.
So there. If you want to truly test Intel chips, Sync all cores at 5 Ghz, set core cache ratio to 4700mhz and use at least 4000 Mhz Cas latency 17 17 17 ram kit.
Now they are on equal grounds. Both maxed out.
So let's do few examples.
But first of all, i know how good AMD is, and if you are building PC now for productivity, get an AMD it's not about that but let me show you something in gaming.
This is the Ryzen 7 and 9 review it shows that intel is 17% faster.
But what if you give Intel 4000 Mhz ram instead of 3200.
It's suddenly not 17% faster, but more like 30 % faster. Know what I mean ?
Let's do another one, this time for Battlefield 5
Here's the Ryzen review:
And here is the DDR 4000 review
Lets put them together.
Here we go again, Intel minimum are Ryzens average.
There was more reviews done but strange brigade is well optimised and didn't show big improvements and Hitman was done on DX 11 not DX 12 so I can't compare those.
Last thing I want to say is: there's another group of people, who say : yeah but I game on 1440p so it doesn't matter. Yea... for now, what do you think will happen when 3080Ti comes out.
That's all I got say.
ninezerofive
P.S. I am very happy with my 9700k, but the future is 12 and 16 core.
With that being said, I don't understand what's going on with the reviewers. Either they are shilling AMD hard or just are clueless, let me explain.
When principle technologies came out with the Intel review everybody jumped do defend that the AMD wasn't tested as they should be, not that mode, not that cooler etc.
So I don't understand what's going on now with themselves.
All of them as one are saying that new Ryzens are running pretty much as fast as it can straight out of the box. And there's no overclocking headroom. Hmm, alright fair enough. Second as "Hardware Unboxed" demonstrated it doesn't really matter what ram are you using 3200 C14 or 3600 C16 as it will run the same. Which makes me think that even if Ryzen can do 3600 it benefits as much from lower latency so 3200 cl14 will do just fine.
So if you are letting AMD boost up by itself as fast as it will go and add a 3200 cl14 ram then it will be reaching its ceilings.
Fair enough.
THEN WHY THE HELL YOU ARE COMPARING IT TO STOCK INTEL CHIPS ? Especially 9700k running at 4.6 all cores at stock. Seriously guys.
So we have established that AMD are being tested at their peak performance.
Let jump on to the side of Intel.
First of all, there is one group of people on the internet who say : ram speed or latency doesn't matter for Intel. Oh yes it does ! Maybe not in all instances but it definitely does !
Hardware unboxed made a video about this titled : Is Fast RAM A Waste? Unleashing the Core i9-9900K with DDR4-4000 / Link Here :
https://www.youtube.com/watch?v=VElMNPXJtuA
So to test Intel near to it's limits like you are doing with AMD it should at least have 4000 CL 17 ram. That's point number one.
Second point, 9900k and 9700k should be tested at 5Ghz synced on all cores. I am running 5.2 and only reason I am not running 5.3 is because my 240 AIO can't cool it. Alright I am not saying you should test with 5.2 but the golden average 5 Ghz synced on all cores should be a default testing method.
Third point. Uncore or Core cache ratio, I can guarantee you that most of these reviewers are testing intel chips with 4300 mhz cache ratio, for example mine runs at 4900 mhz, again I am not requesting that tests have to be done with 4900 mhz, but 4700 mhz should be doable for all chips.
So there. If you want to truly test Intel chips, Sync all cores at 5 Ghz, set core cache ratio to 4700mhz and use at least 4000 Mhz Cas latency 17 17 17 ram kit.
Now they are on equal grounds. Both maxed out.
So let's do few examples.
But first of all, i know how good AMD is, and if you are building PC now for productivity, get an AMD it's not about that but let me show you something in gaming.
This is the Ryzen 7 and 9 review it shows that intel is 17% faster.
But what if you give Intel 4000 Mhz ram instead of 3200.
It's suddenly not 17% faster, but more like 30 % faster. Know what I mean ?
Let's do another one, this time for Battlefield 5
Here's the Ryzen review:
And here is the DDR 4000 review
Lets put them together.
Here we go again, Intel minimum are Ryzens average.
There was more reviews done but strange brigade is well optimised and didn't show big improvements and Hitman was done on DX 11 not DX 12 so I can't compare those.
Last thing I want to say is: there's another group of people, who say : yeah but I game on 1440p so it doesn't matter. Yea... for now, what do you think will happen when 3080Ti comes out.
That's all I got say.
ninezerofive
P.S. I am very happy with my 9700k, but the future is 12 and 16 core.
Last edited:
