Read more.Quote:
Worth it over AMD's Ryzen? We find out.
Printable View
Read more.Quote:
Worth it over AMD's Ryzen? We find out.
Interesting but not surprising results. If you're purely gaming then Intel is a lot better because of its single core performance is a lot stronger. But in other workloads the Ryzen either matches it or beats it.
I'm waiting to see how Zen 2 will be able to compete with Intel in terms of single core perf but it probably still won't match it but hopefully should get a lot closer.
Part of the Spectre fixes was cutting HT as it was one of the main vectors. It will take till 2020 for all the fixes to be pushed to hardware with Intel chips. Thus why Coffee Lake and beyond will not have HT.
A lot stronger? The results show 168 vs 153 fps in F1, under 10% difference and staying well withing the comfortable range of a variable sync display. Not what I would call "A lot", and that's assuming lightly threaded games which not all of them are.Quote:
Originally Posted by AlvieM
The only big difference in there is the PiFast results which are utterly meaningless because there isn't any other workload that I have ever come across that is like it.
Edit: I like how the first section is titled "Conclusion", good for confusing people who just skip to the end :D
I guess "a lot stronger" is an exaggeration but it is still significant in my opinion.Quote:
Originally Posted by DanceswithUnix
*Shrug*, you must be way more sensitive to frame rates than I am if you can tell a 10% difference, specially with adaptive sync.Quote:
Originally Posted by AlvieM
But then I chose largely for work use where I want lots of threads, but what started my last upgrade was Elite Dangerous in VR which on the Rift seems to eat as many threads as you can throw at it.
I'd love to see a Real World benchmark, with the gamer's reality: Discord open and voice chat on, 2 or 3 chrome tabs running, Steam, keyboard and mouse software loaded in the systray, and eventually streaming to Twitch.
Its true that poorly optimised games will favour less threads, but with the Real World setup you'd be hammering the CPU with a load of different apps... its not just one game anymore.
That would be a very interesting test to see.
I was about to post similar, plus at higher than 1080p the CPU's are neck and neck basically. Nothing significant about that in my opinion.Quote:
Originally Posted by DanceswithUnix*Shrug*, you must be way more sensitive to frame rates than I am if you can tell a 10% difference, specially with adaptive sync.Quote:
Originally Posted by AlvieM
But then I chose largely for work use where I want lots of threads, but what started my last upgrade was Elite Dangerous in VR which on the Rift seems to eat as many threads as you can throw at it.
Probably not in that title and resolution but there will be others where the FPS will be lower and those the difference might be noticeableQuote:
Originally Posted by DanceswithUnix
Obviously not every game but perhaps those optimised for Intel / single core.
Why no comparisons with the previous gen?
It bugs me whenever people conflate 1080p results with a top-end GPU with all 1080p results. You're giving the impression that a mid-range system is better off with intel ("it's better at 1080p, and that's what I play at"), when an actual test with a mid range card won't show a difference at 1080p between AMD or intel (like the 4k results). The actual difference is high-refresh-rate gaming (where the CPU overhead is greater) - the resolution this shows up at is a function of GPU, not CPU
Hi there,Quote:
Originally Posted by Usernamist
Because we chose to update the drivers and try different settings for some benchmarks. If you look under the Test Methodology page, there is a link to older editorial with a plethora of CPUs, including the near-identical Core i5-8400. This was an as-today evaluation.
I have yet to see any results which look like gameplay will be impacted by having a Ryzen CPU. There are more comparisons here:Quote:
Originally Posted by AlvieM
https://www.techspot.com/review/1829...ryzen-5-2600x/
I'm ignoring the overclock results as you can't rely on those, I think they make the 2600X look artificially too good and personally I don't overclock.
Far Cry looks pretty bad for AMD there, until you see that the 2600X is getting minimums of 73fps at 1080p, and if you are on 1080p that is either maxing out your 60Hz monitor or in this house comfortably handled by FreeSync. But some interesting comments in that review:
Battlefield V:
Quote:
The experience with the 9400F for the most part was very good, but whereas the 2600X was silky smooth at all times, the 6-core Intel CPU suffered odd frame stuttering here and there.
So with the Intel chip you have to overclock memory to get smooth gameplay?Quote:
Using DDR4-2666 memory the 9400F was a little jittery in Monster Hunter: World
Intel still hanging tough on 14nm+++++++++ ?
Get the #uck outa here. Mean while AMD is at
7nm moving towards 5nm, not that, that's
important thing to mention here...........
The obvious thing to point out here is that frame rates in games are fairly useless in that normally, investment in the GPU is the plan of choice rather than upgrading a CPU for FPS. 1080 is chosen here as most of the time these workloads are GPU bound and so you'd not spend more on a CPU for frame rates except in very specific games.
Yeah, Intel still seems to be having a lot of problems with the 10nm and 7nm manufacturing processQuote:
Originally Posted by albert89
I guess it trades blows and both have advantages and disadvantages though. The average person can notice more than 60Hz though even though not everyone owns a high refresh rate monitor.Quote:
Originally Posted by DanceswithUnix
"Though still faster than AMD for light-load applications and lower-res gaming... "
- What the hell does that even mean? That's Intel lackey style reporting.
Light load & low res gaming means when the CPU is basically barely more than idle or doing only single threaded loads isn't it?
That is a ridiculous justification for a modern CPU. Why would anybody want to consider speed in such unrealistic conditions. In this day and age, all users would benefit from multi-tasking and multi-threaded performance because we all run multiple apps at all times. Yes, even when gaming, you would want your communication apps and browsers and notifications to be running.
Its the same reason why Androids are ruling the world of personal computing by utilizing multiple (up to octa-cores) core CPUs to give a good user experience.
isn't it kind of silly to still only make avg and mins in 2019? Where are the 1% and 0.1% lows?