Ryzen 7 1700 vs i7 7700k for gaming

gavinh87 · 18 Jul 2017 at 20:02

Other than mindblanktech video I haven't seen anything new. I can post a 3dmark result up if you want to compare but I am GPU limited with a 1070.

humbug · 18 Jul 2017 at 21:13

Sampsy said:
No, it's a bad benchmark because the setup skews the results in favour of Ryzen. Nobody should be comparing overclockable chips at stock, especially in the case of the 7700k where if you weren't going to OC you'd just buy the 7700. The entire point of that CPU is to overclock it. It's like saying "Hey look! If we make the 7700k slower than the 1700x it's slower than the 1700x! Amazing!".

Interesting that you removed my line in the quote before responding to it, you obviously knew it addressed that nonsense which why is why you removed it.....

So here is what i said, again.

Nothing at all wrong with that comparison, the 7700K overclocks by 15% while even the highest clocked Ryzen still overclocks by 10%, the one he used was a 1700X which overclocks a little more than that, the added bonus with Ryzen is higher memory performance adds even more to its gaming performance.

Both chips are overclockable and both gain about the same performance from it so your argument as to why its not a fair comparison for Intel is complete rubbish.... is it you just don't like the results?

voidshatter · 18 Jul 2017 at 22:22

humbug said:
is it you just don't like the results?

Ryzen crusaders should love the results given that Ryzen has claimed victory in 1 out of 6 games.

Can't wait to see more games optimised for multi-thread programming in the near future, before TR4 widely defeats the purpose of AM4 for number of cores

Oh wait, performance price ratio? How about Intel Pentium G4560 for gaming?

Sampsy · 18 Jul 2017 at 23:16

gavinh87 said:
The percentage of people that overclock are much lower than you think. People on these kind of forums are a minority.
How about the videos of 3.9ghz ryzen Vs the 5ghz 7700k? Are those skewed for ryzen too?

Nope I don't see why that would be skewed. If Ryzen wins in a fair test then it wins. I'm not biased against Ryzen as Humbug is implying. I just don't think comparing CPUs which are intended to be overclocked at stock is a fair, representative test, especially when they overclock differently.

humbug said:
Interesting that you removed my line in the quote before responding to it, you obviously knew it addressed that nonsense which why is why you removed it.....

So here is what i said, again.

Both chips are overclockable and both gain about the same performance from it so your argument as to why its not a fair comparison for Intel is complete rubbish.... is it you just don't like the results?

I removed that line because I was in a rush earlier today and felt that given I had only a few moments to post a response that I would focus on the part of your post I disagreed with most. So I quoted only that section to be clear what I was responding to.

Of course Ryzen can overclock too and a typical overclock % isn't wildly different between the CPUs. But it isn't the same and if you're benchmarking, especially to advise which is the best CPU to buy you want to replicate the intended usage scenario as closely as possible.

Running at stock might well get you in the ballpark of the relative overclocked performance, but why get in the ballpark when you can just overclock the CPUs in the first place. Running benchmarks on CPUs at stock when you know people should be overclocking those CPUs needlessly adds inaccuracy to your results and makes your benchmarking less useful.

I'm not sure why you're so ready to call me biased here. I don't have a horse in this race and I think I'm being perfectly fair. I don't particularly care if the 7700k or 1700/x is the better gaming CPU as I don't own nor am I planning to buy either. Neither am I trying to claim one is faster than the other. I just think comparing those CPUs at stock is not a fair comparison.

humbug · 19 Jul 2017 at 01:40

Sampsy said:
Nope I don't see why that would be skewed. If Ryzen wins in a fair test then it wins. I'm not biased against Ryzen as Humbug is implying. I just don't think comparing CPUs which are intended to be overclocked at stock is a fair, representative test, especially when they overclock differently.

I removed that line because I was in a rush earlier today and felt that given I had only a few moments to post a response that I would focus on the part of your post I disagreed with most. So I quoted only that section to be clear what I was responding to.

Of course Ryzen can overclock too and a typical overclock % isn't wildly different between the CPUs. But it isn't the same and if you're benchmarking, especially to advise which is the best CPU to buy you want to replicate the intended usage scenario as closely as possible.

Running at stock might well get you in the ballpark of the relative overclocked performance, but why get in the ballpark when you can just overclock the CPUs in the first place. Running benchmarks on CPUs at stock when you know people should be overclocking those CPUs needlessly adds inaccuracy to your results and makes your benchmarking less useful.

I'm not sure why you're so ready to call me biased here. I don't have a horse in this race and I think I'm being perfectly fair. I don't particularly care if the 7700k or 1700/x is the better gaming CPU as I don't own nor am I planning to buy either. Neither am I trying to claim one is faster than the other. I just think comparing those CPUs at stock is not a fair comparison.

If you agree there isn't much difference between the CPU overclocked vs stock what is the problem? the majority of people actually do not overclock at all anyway, and as you agree it doesn't make much difference to the performance between the two.
Hell if you're overclocking the chips why not also tweak the memory speed and timings? that may narrow the gap between Ryzen and the 7700K to nothing, that would also be fair given its only right that the best possible performance should be got out of the Ryzen chip as well as the 7700K.

RBImGuy · 19 Jul 2017 at 08:52

few months with the ryzen 1600
best buy you can make with a cpu today
Intel laps glue, goes heating, voids warranty if you overclock and sometimes you dont need to oc either for it to heat up.
then the stuttering when games overload the cpu cores.

Yea really happy camper here paying half what Intel asked for 4 cores vs 6.
dont be a dumpkoff

chroniclard · 19 Jul 2017 at 08:59

Comparing at stock is the only fair comparison.

Phage · 21 Jul 2017 at 09:32

chroniclard said:
Comparing at stock is the only fair comparison.

I'd agree. Especially as the results of OCing vary widely chip to chip and may not be representative of the final performance of a chip that anyone may buy.

humbug · 21 Jul 2017 at 10:07

Look at these huge gains you get from tweaking the memory performance, from 2133Mhz to timings tweaked 3400Mhz is 40% higher performance, if reviewers spend a little time properly setting up the Ryzen systems they would look very very diffrent, Intel would look very vulnerable not just in the HEDT space.

Phage · 21 Jul 2017 at 10:33

I'm too much of a noob to play with the timings on RAM. I assume the DOCP profiles will cover this ?
My 3600 Corsair Kit seems very happy at 3200 when selected in the BIOS

humbug · 21 Jul 2017 at 10:39

Phage said:
I'm too much of a noob to play with the timings on RAM. I assume the DOCP profiles will cover this ?
My 3600 Corsair Kit seems very happy at 3200 when selected in the BIOS

Mostly yes, but even with higher latency timings your still going to get a 20% boost over the 2400/2667Mhz ram Reviews like using for Ryzen.

Phage · 21 Jul 2017 at 10:52

That sounds like a 'yes' ? Certainly feels better

humbug · 21 Jul 2017 at 10:54

Phage said:
That sounds like a 'yes' ? Certainly feels better

DOCP profiles will set better than stock timings but not to the best you could get out of it tweaking it yourself, so yes but not 'as' good.

Phage · 21 Jul 2017 at 10:57

Makes sense, thanks

///Rage · 21 Jul 2017 at 11:55

humbug said:
Look at these huge gains you get from tweaking the memory performance, from 2133Mhz to timings tweaked 3400Mhz is 40% higher performance, if reviewers spend a little time properly setting up the Ryzen systems they would look very very diffrent, Intel would look very vulnerable not just in the HEDT space.

To a noob like myself considering a 1600X upgrade, whats the difference between3200Mhz and 3200Mhz LL?

Ive seen some ram advertised here as 3200mhz CL14.

karik666 · 21 Jul 2017 at 12:13

Here is the 1600 vs 7800x and 7700k in 30 games.

Panos · 21 Jul 2017 at 12:46

humbug said:
Look at these huge gains you get from tweaking the memory performance, from 2133Mhz to timings tweaked 3400Mhz is 40% higher performance, if reviewers spend a little time properly setting up the Ryzen systems they would look very very diffrent, Intel would look very vulnerable not just in the HEDT space.

Yep, but unfortunately they still test them with 2666 ram like Hexus this week. And pointing similar charts about the perf gains and that reviewers are at fault, even in this forums we are classed as AMD zealots. As someone put it on the other discussion couple of days ago.

btw where you got the image from?

humbug · 21 Jul 2017 at 12:54

///Rage said:
To a noob like myself considering a 1600X upgrade, whats the difference between3200Mhz and 3200Mhz LL?

Ive seen some ram advertised here as 3200mhz CL14.

LL is Low Latency, where the ram has been tweaked from something like CL16 to CL14

humbug · 21 Jul 2017 at 13:00

Panos said:
Yep, but unfortunately they still test them with 2666 ram like Hexus this week. And pointing similar charts about the perf gains and that reviewers are at fault, even in this forums we are classed as AMD zealots. As someone put it on the other discussion couple of days ago.

btw where you got the image from?

Can't remember... but Googling i found this.... scroll down a bit https://forums.anandtech.com/threads/how-is-ryzen.2507141/

Panos · 21 Jul 2017 at 13:05

karik666 said:
Here is the 1600 vs 7800x and 7700k in 30 games.

The WOT benchmark matches what I write months now. On a single thread game like that I get more fps (around 18%) on the 6800K and 1700X at 4ghz than the 6700K @4.8ghz.

because I don't play only the game but using TS to talk to my team mates. And having observed the thread usage on the MSI overlay I can see how hard the 4cores are pushed for sharing their resources over 6 or 8 cores cpus.

Nice also to see someone using 3200mhz on his reviews. The guy is pretty good on his videos.