6700k or Ryzen for a new build

[Sigo]

Hey all, I have a chance in getting a new 6700k for £225 or I can get a R5 1600 fro £199. Ive seen benchmarks of said cpu's but I would like some feed back on people that have use these.
I would like to know if anyone had a 6700k and moved to a ryzen build and if they felt it was worth it. All I tend to do with with this pc is play games, stream and record with fraps with a little video editing. Thanks for you input and time.

[Baelic]

I'd probably lean a little closer to the 6700k, given your potential use-case

[Evildeffy]

In simple terms how much streaming, recording and video/audio editing do you do?

If you have an overwhelming %-age to gaming (say 80% gaming and 20% rest) I would probably grab the 6700K and just enjoy.
If you have a closer to equilibrium... then the Ryzen's extra cores WILL outpace the 6700K.

Overclocked or not matters little, of course it would be faster.

But that said you won't find anyone that went from a 6700K to a Ryzen 5 1600 and say it was worth it UNLESS they WILL use those cores in pure production/content creation, if you as a gamer go from a 6700K to R5 1600 ... well.. you have too much money and should give me it all!

[potis]

Hey all, I have a chance in getting a new 6700k for £225 or I can get a R5 1600 fro £199. Ive seen benchmarks of said cpu's but I would like some feed back on people that have use these.
I would like to know if anyone had a 6700k and moved to a ryzen build and if they felt it was worth it. All I tend to do with with this pc is play games, stream and record with fraps with a little video editing. Thanks for you input and time.

As Evildeffy already mentioned the core parts, if you are mostly gonna game, 6700K, the passive 4.2GHz makes a difference, no OC etc needed for the most part.

If you are gonna "Fire up a single player 2 hours a week" and work the rest, then the Ryzen could be slightly better depending if your program actually uses them cores, i would still go for the 6700K either way personally.

Its been almost 2 months and still updates and microcode fixes and memory problems.

No sensible human being with a clue would go from 6700K to a Ryzen build unless they are seriously making that much money from video editing that the few minutes/hours faster translates to that much money.

[Sigo]

Thanks for the reply's you have helped me decide to jump from my 8320fx to skylake

[nToxik]

6700K (with a Z270) motherboard. Most motherboards have built-in overclocking that pretty much overclock themselves with one button push.

[Dilbon]

No point getting the i7 really, get a B350 motherboard and the Ryzen. Gamingwise they are equal (meaning i7 is faster but you won't notice it) but in productive things Ryzen is better.

[Deleted]

No point getting the i7 really, get a B350 motherboard and the Ryzen. Gamingwise they are equal (meaning i7 is faster but you won't notice it(...).

Unless of course you play a game where you do notice, i.e. games that are limited by the cpu's single thread performance and having low fps. Yes those are rare, but they do exist.

[moremana]

Unless of course you play a game where you do notice, i.e. games that are limited by the cpu's single thread performance and having low fps. Yes those are rare, but they do exist.

Still wont notice it. Ryzen competes with Intel, less maybe 10%~.

[Deleted]

Still wont notice it. Ryzen competes with Intel, less maybe 10%~.

Well over-clocked it's ~25% in clock speed plus whatever the IPC gain is in the game's case, so I'd say we're looking 15%+ fps difference. Now if your game already runs at 200fps then indeed it wouldn't be noticeable.
However for games struggling at 30 fps, 5fps are noticeable, and if you're running with v-sync on the difference might be further amplified, even for games close to 60 fps.

[Lathais]

Well over-clocked it's ~25% in clock speed plus whatever the IPC gain is in the game's case, so I'd say we're looking 15%+ fps difference. Now if your game already runs at 200fps then indeed it wouldn't be noticeable.
However for games struggling at 30 fps, 5fps are noticeable, and if you're running with v-sync on the difference might be further amplified, even for games close to 60 fps.

Except that you can't compare clock speeds across architectures like that. Real world benchmarks show the difference in games to be closer to 2-5%. Even at 30FPS, you are not likely to notice a 5% difference, as that's only 1.5FPS. If you are running V-Sync and getting 31 FPS, yeah, you may notice a difference.

There have even been double blind tests done where people do not notice the difference. In one that gets posted here the only hiccup any of the 3 mentioned was one particular spot in one particular game that stuttered. They were kinda shocked when they found out that it was one of the intel systems stuttering not the Ryzen. Other than that one spot in one game though, no discernible difference.

[Deleted]

Well over-clocked it's ~25% in clock speed plus whatever the IPC gain is in the game's case, so I'd say we're looking 15%+ fps difference. Now if your game already runs at 200fps then indeed it wouldn't be noticeable.
However for games struggling at 30 fps, 5fps are noticeable, and if you're running with v-sync on the difference might be further amplified, even for games close to 60 fps.

Need to stop you right there, if you own a Ryzen 1600 or the Intel 7600, and you are struggling to run at 30 FPS normally, thats a badly made game you ask for a refund on, these class of hardware on high - very high settings will be getting you 60 FPS with a mid range GPU pretty much all the time. You're comparison is null for this class of hardware, not to mention if you are running on ultra, you are likely to be GPU bound first but foremost, you just notch down settings.

[Deleted]

Except that you can't compare clock speeds across architectures like that.

You can to some extend if you consider the average IPC which is slightly worse for Ryzen (if you disregard AVX2). Sure there may be some synthetic benchmarks that run a very limited instruction set yielding in a vastly different ICP. But in general that metric is pretty spot on in my experience fro single thread performance.

Real world benchmarks show the difference in games to be closer to 2-5%.

Real world benchmarks vary quite a bit and usually none of the games tested are single threaded CPU limited.

Even at 30FPS, you are not likely to notice a 5% difference, as that's only 1.5FPS. If you are running V-Sync and getting 31 FPS, yeah, you may notice a difference.

Never said you'd notice 5%, but 20% means 6 fps and that is noticeable even when not taking v-sync into the picture.

There have even been double blind tests done where people do not notice the difference.

I've seen the video - none of the game tested was limited by single thread performance and they were running quite a bit faster than 30fps.

[ComputerNerd]

Just a heads up, some people are going to tell you that in 5-6 years most games will perform better on systems that have more cores. There isn't any real evidence for this right now, so take any of those comments with a grain of salt.

It has certainly gained more traction recently, but I agree that an assumption is wrong.
Not all games will benefit equally, and it depends on the willingness of the developer.
It isn't a minor or quick change to implement, going from single to multi-threaded.

[MatthewOU2015]

Go Ryzen for your new build

[ComputerNerd]

It has gained more traction, but the benefits vary greatly from title to title. Add in the additional overhead to make a game that utilizes the extra cores and there's currently not a ton of incentive for companies to invest in it.

Yes, the incentive really depends on the returns.
Is it going to improve performance enough, which is dependant on how parallel the processes are.
Is it selling enough/will it sell enough to justify the update.

Ashes of the Singularity is perhaps as close to an ideal case scenario as we will get.
Though even then there will be limits to the scaling.

[Deleted]

Need to stop you right there, if you own a Ryzen 1600 or the Intel 7600, and you are struggling to run at 30 FPS normally, thats a badly made game you ask for a refund on, these class of hardware on high - very high settings will be getting you 60 FPS with a mid range GPU pretty much all the time. You're comparison is null for this class of hardware, not to mention if you are running on ultra, you are likely to be GPU bound first but foremost, you just notch down settings.

The 30 fps is not even on ultra settings and it is not the GPU that is limiting here.

The game simply depends heavily on single thread performance because they were unable to split their main thread into multiple independent tasks or offload to the GPU.

Now that would certainly qualify as "badly made" if we were talking about the bazillion-th CoD clone.
However in this case it's a high fidelity civilian flight simulator, so I'm not sure how difficult it is to make that work.
I suspect it mainly has to do with the smaller development team and being a non-mainstream type of game, but of course it's also possible that this type of game doesn't lend itself to multi tasking like shooters do.

In any case I haven't seen any alternative that does it better, so I can either abandon this type of game and bask in the glory of the difference between Ryzen and i7 being unnoticeable, or admit that this game thrives on single thread performance and I need to get an i7 to get the best performance... although I'd love to get a 1700.

[potis]

But @Vegas82 is correct.

We had this discussion back in 2011, nothing changed for 2017.

Battlefield and 2-3 more games were the only games utilizing more cores "properly", this is 2011-2012.

Battlefield 1 and 2-3 more games (relevant games) are the only games utilizing more cores "properly" this is 2017-2018.

There is no reason to get a Ryzen at the moment if you are gonna game only.

Sounds bad but it is how it is in the end.

All the most played games in the world couldnt care less about extra cores simply because they are old engines and wont be updated any time soon, an i3 at 4Ghz is cheaper and owns any Ryzen at the moment cause its clocked so low.

I know we all strive for competition so both sides release better stuff but this is getting ridiculously biased towards AMD nowadays and suggesting stuff which arent right.

WoW/CS:GO/Dota/LoL and a few other "CORE" games which are most of the games mentioned in here and in every pc building forum prefer single IPC+ Clock to work better, they couldnt care less about 1234223 cores.

We whine that Intel releases only 5-10% upgrades over years but suddenly being 5% behind in IPC and 500Mhz-1Ghz which results in a good much more % in raw difference in most games at the same time is somehow acceptable now.

Not sure how suddenly a "3.7Ghz" overclock is consider "okay".

Not sure how suddenly it became "acceptable" that Ryzen has memory problems 2 months after releasing, and still has microcode updates.

Not sure how every test must be made with A NON OVERCLOCKED INTEL BUT OVERCLOCK THE AMD TO BE FAIR!!

[MrPaladinGuy]

This thread annoys me.

On the subject of games,

The only time Ryzen (with DDR4 3200) ever appears to beat a 7700 is in a handful of situations. Many of the pro AMD benchmarks you see have a GPU bottleneck then the AMD fans just go "LOOK HOW CLOSE THEY ARE" "WHEN YOU'RE THAT HIGH WHO CARES MOAR COARS ERMAHGURD".

The performance difference ranges from a few frames to significantly more.

https://www.youtube.com/watch?v=TDvk9_iTq6Y

https://www.youtube.com/watch?v=HZPr-gNWdvI

Do I need to post more? I used DF as proof since they're one of the last unbiased places around.

You shouldn't take my word for it, look for benchmarks from others, but pay attention to a multitude of factors; settings, what games were used, if said games are CPU-bound and when, if there's a GPU bottleneck available, ram speed, etc.

There's also a famous 1700 vs 7700k video going around with an obvious GPU bottleneck that AMD fans are eating up on Reddit like a fat kid discovering cupcakes.

Is Ryzen worth it for for some people?

Yes.

Is it the best?

Not even close in many cases.

Like I said, the unbiased general consensus is if you want pure gaming performance for current games or old games and want to gain every frame you can (144hz displays), go Intel, if you want reasonably likely future proofing and improvements in other aspects, go Ryzen.

And for the record in the last week a channel named Joker Productions was caught falsifying a 1700 vs 7700k benchmark with Prey where the Intel benchmark had shadows enabled but the Ryzen one didn't. The video was quickly removed after it appeared on OCN.net

http://www.overclock.net/t/1629676/y...#post_26072392

Also, proper multithreaded programming is incredibly complicated, that's why most games still use one thread as the main thread.

Think of this. If it was so easy then why isn't every game using it, especially with most games aiming at a console being the main platform and those consoles use very low freq multicore CPU's where extra threads would help significantly.

I could say so much more, but I'll leave it at that.

[potis]

I love how someone else was posting the exact same thing i was posting

@MrPaladinGuy

This forum is turning like some Leftist forum, is that how they call all these multi-gender rights and random stuff i am not sure? but with AMD.