Ryzen 5 1600 vs i7-7800X: Ryzen 1080p Gaming Issues Explained; 7700k still King

[mr moose]

8 hours ago, XenosTech said:

Everyone should lol. Why pay more for an intel cpu when you can get the amd one for cheaper with the same performance in multi-threaded tasks unless they're specifically AVX based where intel would be better for the performance.

 

 

 

That's just it, I am looking to upgrade in a few months, right now I am watching everything from the yet to be released ryzen3 all the way to the x299 stuff.  I can save a solid $150 by going with a ryzen5 1600 over the 7700K, but I keep jumping between that and an i5, So depending on how AMD holds up and with the ryzen3 over the next few months I still need to see all CPU's compared in reviews.

[XenosTech]

5 minutes ago, mr moose said:

 

That's just it, I am looking to upgrade in a few months, right now I am watching everything from the yet to be released ryzen3 all the way to the x299 stuff.  I can save a solid $150 by going with a ryzen5 1600 over the 7700K, but I keep jumping between that and an i5, So depending on how AMD holds up and with the ryzen3 over the next few months I still need to see all CPU's compared in reviews.

It's the same for me now but I'm stuck with the 1600 or 1700 since I no longer want to just swap the cpu in my build and then buy my gf a new pc. I decided to I'll build a ryzen based system since that'll be a better use of money than buying a 6700k or a 7700k to run vm's and play games

[mr moose]

44 minutes ago, XenosTech said:

It's the same for me now but I'm stuck with the 1600 or 1700 since I no longer want to just swap the cpu in my build and then buy my gf a new pc. I decided to I'll build a ryzen based system since that'll be a better use of money than buying a 6700k or a 7700k to run vm's and play games

 

The only thing I keep forgetting is the slight premium for AMD mobo's.  Here it's about $30 for the base model mobo's.  That's still an over all saving though.

 

 

 

[XenosTech]

11 minutes ago, mr moose said:

 

The only thing I keep forgetting is the slight premium for AMD mobo's.  Here it's about $30 for the base model mobo's.  That's still an over all saving though.

 

 

 

Not bad it's probably $100 more or so if I buy local so something like the crosshair hero would be $400 usd or so 

[Dark Force]

This is interesting - Hardware Unboxed tweeted a warning about OC'ing Core-X CPUs a couple of days ago:

 

[ravenshrike]

5 hours ago, mr moose said:

 

That's just it, I am looking to upgrade in a few months, right now I am watching everything from the yet to be released ryzen3 all the way to the x299 stuff.  I can save a solid $150 by going with a ryzen5 1600 over the 7700K, but I keep jumping between that and an i5, So depending on how AMD holds up and with the ryzen3 over the next few months I still need to see all CPU's compared in reviews.

What monitor are you going to use and what are the top five games you play?

[mr moose]

Just now, ravenshrike said:

What monitor are you going to use and what are the top five games you play?

Haven't settled on a monitor yet, that'll be part of my upgrade.  Most likely I'll be playing SC (if it ever releases finished). Outside of that I only play games when they take my interest.  So it could be everything or it could be just SC and ESO.

[ravenshrike]

Just now, mr moose said:

Haven't settled on a monitor yet, that'll be part of my upgrade.  Most likely I'll be playing SC (if it ever releases finished). Outside of that I only play games when they take my interest.  So it could be everything or it could be just SC and ESO.

Then go Ryzen. The only significant advantage the 7700k gives you is higher framerates at 1080p when not GPU bound or frame capped. As the framerates are almost always above 80fps on the 1600 anyway, the only way you would notice a difference is playing twitch FPS's competitively on a 120hz+ monitor. Playing other games the difference is not noticeable during normal gameplay, apart from the infrequent stutter in certain games like GTA 5 and Fallout 4 which are longer when they do occur on the 7700k vs the 1700 but still less than 200ms.

[mr moose]

1 minute ago, ravenshrike said:

Then go Ryzen. The only significant advantage the 7700k gives you is higher framerates at 1080p when not GPU bound or frame capped. As the framerates are almost always above 80fps on the 1600 anyway, the only way you would notice a difference is playing twitch FPS's competitively on a 120hz+ monitor. Playing other games the difference is not noticeable during normal gameplay, apart from the infrequent stutter in certain games like GTA 5 and Fallout 4 which are longer when they do occur on the 7700k vs the 1700 but still less than 200ms.

 

I'm going to wait until the ryzen 3's are out and benched and compare them to the intel side of things first.  I have plenty of time up my sleeve.

[Mooshi]

5 hours ago, XenosTech said:

It's the same for me now but I'm stuck with the 1600 or 1700 since I no longer want to just swap the cpu in my build and then buy my gf a new pc. I decided to I'll build a ryzen based system since that'll be a better use of money than buying a 6700k or a 7700k to run vm's and play games

Do what I did. My GF inherited my 6700k system because her PC was dying and I built a Ryzen 1600 build since the 7700k is hardly an improvement over the 6700k it replaced.

[XenosTech]

4 minutes ago, Mooshi said:

Do what I did. My GF inherited my 6700k system because her PC was dying and I built a Ryzen 1600 build since the 7700k is hardly an improvement over the 6700k it replaced.

She was getting this i5 regardless of what I did. My original plan was to buy a 6700k then get her everything else minus the cpu but ryzen looks good enough that I can give her most of this system as is minus the 1070 and all the extra drives I added in when I went bat shit crazy with the vms. She only plays 1 game so this i5 and a 460 (or what ever she wants to buy) would go a long way

[suits]

I like how all the intel fanboys have to keep reminding themselves that the 7700k is still the best gaming CPU. Yeah it is. Everyone knows it. I also HAD to delid my 7700k to get actually good performance out of it because of the shit Job intel did. you're still comparing apples to oranges and I'm still sick of intel screwing me. I'll be sticking with ryzen on my content pc until intel stops screwing me. 

[JoeCoke]

200$ vs 400$?

[mr moose]

13 minutes ago, suits said:

I like how all the intel fanboys have to keep reminding themselves that the 7700k is still the best gaming CPU. Yeah it is. Everyone knows it. I also HAD to delid my 7700k to get actually good performance out of it because of the shit Job intel did. you're still comparing apples to oranges and I'm still sick of intel screwing me. I'll be sticking with ryzen on my content pc until intel stops screwing me. 

do you consider everyone who reiterates a point to be fanboys, or just the ones who still talk about 7700K?  Are the people who keep reminding us the ryzens are better at multi threaded workloads just fanboys? 

[suits]

5 minutes ago, mr moose said:

do you consider everyone who reiterates a point to be fanboys, or just the ones who still talk about 7700K?  Are the people who keep reminding us the ryzens are better at multi threaded workloads just fanboys? 

AMD fanboys are terrible too but intel fanboys lately have been the armpit of the tech community lately. All this thread had to be about was 1600 vs 7800x. Done. 

[XenosTech]

14 minutes ago, 2Buck said:

200$ vs 400$?

what's so odd about that ?

[mr moose]

10 minutes ago, suits said:

AMD fanboys are terrible too but intel fanboys lately have been the armpit of the tech community lately. All this thread had to be about was 1600 vs 7800x. Done. 

except that the original article included the 7700K in all tests and discussion.  The 7700K is a legitimate part of this discussion. Not sure why you take umbrage with that.

[Taf the Ghost]

1 hour ago, suits said:

AMD fanboys are terrible too but intel fanboys lately have been the armpit of the tech community lately. All this thread had to be about was 1600 vs 7800x. Done. 

I included the 7700k statement in the title because it was in the test. The testing results were generally expected, but they point to some really deep technical aspects about the way the current CPUs operate in gaming.

[mr moose]

1 hour ago, Taf the Ghost said:

I included the 7700k statement in the title because it was in the test. The testing results were generally expected, but they point to some really deep technical aspects about the way the current CPUs operate in gaming.

not only that, but it is a CPU still available and a becasue it has been so popular it provides a good reference point when trying to understand how good/bad other CPU's are performing. 

 

As the title says, it's still king of the hill, so if a CPU $120 cheaper performs only slightly worse (overclocking aside) then that is information worth knowing. 

[porina]

Only just saw this thread, all the more interesting since I've started OCing a 7800X yesterday. I see the cache has been touched upon, but I wonder if the game benchmarks so far included cache overclocking or not? The L3 cache speed has tanked compared to previous and increasing the cache clock also helps with ram bandwidth a fair bit. I haven't tried in anger on my sample but a report elsewhere suggests there is significant headroom available to OC the cache clock. I saw some odd behaviour when overclocking just the CPU without touching the cache, that resulted in stupidly low Prime95 speeds although aida64 bandwidth results seemed within expectations. That survived reboots, and went away only when I changed cache clock to something else then back to stock. I suspect is is more likely a bios bug (I have TUF mk2) but haven't proved it.

[kiska3]

18 minutes ago, porina said:

Only just saw this thread, all the more interesting since I've started OCing a 7800X yesterday. I see the cache has been touched upon, but I wonder if the game benchmarks so far included cache overclocking or not? The L3 cache speed has tanked compared to previous and increasing the cache clock also helps with ram bandwidth a fair bit. I haven't tried in anger on my sample but a report elsewhere suggests there is significant headroom available to OC the cache clock. I saw some odd behaviour when overclocking just the CPU without touching the cache, that resulted in stupidly low Prime95 speeds although aida64 bandwidth results seemed within expectations. That survived reboots, and went away only when I changed cache clock to something else then back to stock. I suspect is is more likely a bios bug (I have TUF mk2) but haven't proved it.

Steve did overclock the mesh on the 7800x:

But I don't think he overclocked the cache

[porina]

52 minutes ago, kiska3 said:

Steve did overclock the mesh on the 7800x:

 

But I don't think he overclocked the cache

I believe the mesh/L3 cache are tied. I'll watch that once I'm back from a short trip out. I'm also running GTAV bench in some combos to see if that shows something. Only done one run for now and I've never seen GTAV look smoother. Just tried my sample of 7800x at 3000 cache at stock voltage, a 50% increase, but doing more stability testing on it for now. 3100 is instant crash. It has made a difference to both Prime95 4096k FFT throughput benchmark and aida64 ram/cache bandwidths, I'll look into how significant a change later.

[PCGuy_5960]

7 hours ago, Dark Force said:

This is interesting - Hardware Unboxed tweeted a warning about OC'ing Core-X CPUs a couple of days ago:

Please, Steve damaged ground pins, he probably had bent pins or uneven mounting pressure. This had nothing to do with overclocking the CPU.

[Sauron]

I can't believe we're comparing a 220$ cpu to a 350$ and a 400$ cpu respectively in 1080p gaming. Who are these benchmarks even for? Tech news outlets need to wake the heck up and realize 1080p gaming isn't relevant anymore when benchmarking a cpu, and it definitely  isn't relevant when the price difference is so high. We finally have powerful hexa cores in the mainstream segment and all these people can think about is "hurr durr muh games"... no wonder intel has been milking the same old quad cores since sandy bridge. The only people buying a 350$ cpu exclusively for games are those with more money than sense, and they most likely aren't going to run 1080p, making the difference even smaller.

 

As for the performance gap, what do you expect when one cpu is running 900mhz faster than the other one and a third of the cores on the ryzen aren't even being used? Looking at the average it's actually very impressive that the ryzen 5 was able to get within 10% of the 7700k's performance considering the massive clock speed difference.

[porina]

4 hours ago, kiska3 said:

Steve did overclock the mesh on the 7800x:

But I don't think he overclocked the cache

Here's my GTAV results. My test system is 7800X fixed at 4.3 GHz. Mobo Asus X299 TUF mark 2 bios 0402. Ram is 4x4GB running at either 2133C15 or 3200C16 in quad channel. Modules are single rank. GPU is an Asus 980Ti originally with reference blower, now water cooled so no more thermal throttling. It is running at factory OC whatever that is. Game settings were initially reset to default (which for this system puts most stuff on high/very high) making sure DX11 selected, vsync off, fullscreen.

 

Tested in 4 conditions: ram at 2133 and 3200, cache at 2000 and 3000. It looks like increasing either cache or ram gave a boost, and enabling both gave the highest framerate. There's not much between doing just cache or ram, but ram has a slight edge. Then again, the cache is essentially free performance when you OC it. The difference between 2133-2000 and 3200-3000 for the five passes are: 17%, 7%, 14%, 9%, 14%. I have to wonder how much of this is CPU as opposed to GPU limiting...