Yes but not by that much, the IPC difference between Broadwell-E and Ryzen is know to be between 0 and 12% to Ryzen, not 30% and the 7700K is not 180% faster than both.
Something is not right with ATOS.
-
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" thread (inc AM4/APU discussion) ***
- Thread starter Boomstick777
- Start date
Yes but not by that much, the IPC difference between Broadwell-E and Ryzen is know to be between 0 and 12% to Ryzen, not 30% and the 7700K is not 180% faster than both.
Something is not right with ATOS.
Don't get me wrong, I'm not saying that the FX is not smoother than my i5 4670k. I know it can be.
It's the user's opinion/description of their experience that I find funny and makes me think it's more self-justifying because of new shiny and costs.
It's particularly disappointing because Pcgameshardware do sterling work in my experience with lots of detail and thorough testing. They found a bug in Doom which caused Nvidia cards to get lower performance for example, I think it was something to do with Vsync. Their game tests are usually really good, so I don't think they are a jumped-up tech site. However, this article was clearly not up to their usual standards and they should have done more research before publishing.
The latest Adored TV video was extremely interesting I have to say, and very confusing. Could Nvidia be deliberately sabotaging performance when used with Ryzen? I don't understand how in areas where the game appears to be CPU bound with a Nvidia GPU there is such a big difference when using 2x AMD GPU's.
Soldato
- Joined
- 18 May 2010
- Posts
- 23,042
- Location
- London
Very very unlikely. During AMD 's demo's of Ryzen they where using a GTX 1080.
Plus if that was in someway something Nvidia where doing and it was possible for AMD to prove it, then Nvidia would be getting sued.
Plus why would they want to make their own cards perform worse on a certain architecture.
They want to sell more cards not less.
This argument would also apply to the 6900K given the vast difference between it and the 7700K.
Doesn't add up to me
Exec1: Okay here's the plan! We gimp our GPUs when running with an AMD CPU, thus making their CPUs look bad!
Exec2: Won't that also make their GPUs look good? You know the things we actually compete against?
CEO: You're fired Exec1.
This argument would also apply to the 6900K given the vast difference between it and the 7700K.
Doesn't add up to me
What got me was the relatively small difference with the 7700K between a Nvidia card and an AMD card compared to the massive difference between the Ryzen with the Nvidia and AMD cards. While I know Ryzen cannot match the 7700K for pure IPC it is not, what, 30 odd percent behind.
That with AMD GPU's the chips perform as you would expect with Intel having a slight but not massive edge, compared to with a Nvidia GPU the Intel chip performing a little worse (which I assume is down GPU bottleneck) but AMD chip is performing so significantly worse especially under DX12 is what made me question what was going on.
For those who havent watched the video here is the graph which was most interesting I thought
What got me was the relatively small difference with the 7700K between a Nvidia card and an AMD card compared to the massive difference between the Ryzen with the Nvidia and AMD cards. While I know Ryzen cannot match the 7700K for pure IPC it is not, what, 30 odd percent behind.
For those who havent watched the video here is the graph which was most interesting I thought
Empty spoiler
its not the IPC, Kaby Lake is about 5% better IPC, the real difference is the difference in clock speed in low threaded situations.
I know what you are saying with the differences by using AMD vs nVidia GPU's, it does looks suspect but they would also have to be gimping Intel's Broadwell-E, bar a few tweaks they are pretty much the same chip, it doesn't make any sense.
Something is not right with AOTS but i don't think its deliberate in any way shape or form, its something clearly wrong and AOTS need to investigate.
I think I fixed the spoiler, hopefullyEdited my post too
Yup i see it, it sure is interesting, someone pass this on to AOTS, they need to see this..
Don't know if it's just me (it very often is) but this was benchmarking Tomb Raider. Unless you mean the same sort of benchmarking should be done on AOTS
So it could be what i stated a while back ...AMD may have some magic hot sauce running Ryzen + crossfire this may bode well for the coming Vega then too ...
Stuff of nightmares
we could be drifting into .... Intel / nvidia game only and AMD game only i hope not making pc platform into consoles
Stuff of nightmares
we could be drifting into .... Intel / nvidia game only and AMD game only i hope not making pc platform into consoles
Soldato
- Joined
- 18 May 2010
- Posts
- 23,042
- Location
- London
So it could be what i stated a while back ...AMD may have some magic hot sauce running Ryzen + crossfire this may bode well for the coming Vega then too ...
Stuff of nightmares
we could be drifting into .... Intel / nvidia game only and AMD game only i hope not making pc platform into consoles
But what if you mix AMD CPU with Nvidia GPU?
Soldato
- Joined
- 25 Nov 2011
- Posts
- 20,675
- Location
- The KOP
Really starting to like his channel, goes into much more detail then other so called big channels.
But what if you mix AMD CPU with Nvidia GPU?
You would only get half the game running at best... worse case.... "this game is not compatible with your hardware please use the appropriate hardware stated in the system requirements " or a blank screen
...It states on the back of the box / system requirements AMD only or NVIDIA / INTEL only
Also you are not allowed to buy a green blue case this is Intel / Nvidia only while for a AMD system case red only
its the corporate lawI will stop
TPU just released a massive review of differing Ryzen RAM benchmarks: https://www.techpowerup.com/reviews/AMD/Ryzen_Memory_Analysis/
2133-3200MHz covered at differin CAS and also pretty much all the speeds in between.
FPS difference decreases as the resolution rises due to games being GPU limited not CPU but at 1080p there's a 10FPS or higher difference in some games.
Still their conclusion was it's not really worth buying faster RAM as the increase in performance doesn't warrant the much higher increase in cost. They even said to go for 32GB 2133MHz for size future proofing over faster 16GB kits lol.
2133-3200MHz covered at differin CAS and also pretty much all the speeds in between.
FPS difference decreases as the resolution rises due to games being GPU limited not CPU but at 1080p there's a 10FPS or higher difference in some games.
Still their conclusion was it's not really worth buying faster RAM as the increase in performance doesn't warrant the much higher increase in cost. They even said to go for 32GB 2133MHz for size future proofing over faster 16GB kits lol.
Last edited:
Soldato
- Joined
- 1 Jun 2014
- Posts
- 5,109
I don't always agree with him, but he's one of the very few doing investigative analysis. Everyone else is "blah blah here's some numbers, and a word from our sponsor".
Soldato
- Joined
- 25 Nov 2011
- Posts
- 20,675
- Location
- The KOP
Yep.. Its one reason I like DigitalFoundry videos, They showcase so much more.
We are now in an age that benchmark results with Graphs need to be dropped and not used anymore they don't show the big picture.. I have banged on about this for well over a year now!