Good Ram/CPU/MoBo bundle?

[jerubedo]

12 hours ago, valdyrgramr said:

Well, that last one was pretty bad...

Oh that definitely seems erroneous, yes. There was definitely an issue with the 8700K results. Okay I'm fine throwing away the last screen shot then. The other data is still in line, though. I was just grabbing every example I could find I've eliminated that last screen shot from the post.

[jerubedo]

3 minutes ago, valdyrgramr said:

Well, what makes the others credible compared to a platform like GN?

Steve's data shows the same thing. In Far Cry 5 it was even, margin of error. Look at F1, though:

 

This is right from the video you posted.

 

The 9600K is showing a 6% gain over the 8700K, yet they have the same boost clock and the 8700K has more cache.

[jerubedo]

18 minutes ago, valdyrgramr said:

That was only in one game every other game the 8700k was primarily beating it.  The 8700k also still has the advantage on the multithreaded side.

It was one of 2 games tested. Civ VI turn time is not a measure of FPS. So it was just Far Cry 5 and F1. That's why I used the other videos. It's 8 games and 13 games. In those videos some games are tied as well. But overall, more of them look like Steve's F1 than not.

 

So this also brings us back to the question of WHY does the 9600K perform better than the 8700K at equal clocks and at when the 8700K has more cache. @Herman Mcpootis that question still stands.

 

EDIT: Oops, I just noticed that Assassin's Creed Origins is in there as well, however that game notoriously favors more threads. Even so, they are close to each other.

[jerubedo]

11 minutes ago, valdyrgramr said:

The 8700k won in AC Origins, the turn time test, and FC 5.  Out of those 4 tests, the 9600k won a single test.  Point is you're wrong.  You're cherry picking.

I'm not. Again turn time is not a measure of FPS, so you HAVE to eliminate that one. AC:O favors threads and is a rarer case. 

 

That leaves us with Far Cry 5 (functionally even) and F1: a big lead for 9600K

 

This data source agrees with Steve on Assassin's Creed which gives it credibility:

 

 

However, this same data source has the 9600K on Hitman 2 with a huge lead, Project Cars II with a decent lead, Battlefield V tied, GTA V with a decent lead, Witcher 3 tied, marginal lead for Arma III (just outside margin of error), Shaodow of the Tomb Raider tied.

 

So again, where the 9600K doesn't tie, it pulls ahead. WHY?

 

@valdyrgramr that's a total of 5 of 9 games with the 9600K leading.

[jerubedo]

11 minutes ago, valdyrgramr said:

It's not a rarer case you're trying to dismiss it to fit your claim.  There's a bit of a difference.  You also never answered the question of how your source is as credible as Steve.  Having a check next to your name on a yt account doesn't make you credible compared to a reputation.  So, that's what I am still waiting for you to answer.  Secondly, OP wants to multitask while gaming if you even bothered to read that.  OP would better off with more threads in that scenario.  Which, the 8700k has more of.

No I AM including it. That's why I said above that 5 of NINE games still performed better on the 9600K. I was only pointing out that that particular game is a rarer case since it's actually taking advantage of more than 8 threads. Only a handful of games do that.

 

As for reputability, again, this is all he does. He posts hardware comparisons all day long and with all different configurations. And his data has historically (AND in this case as well as I pointed out with Assassin's Creed) matched up with Steve, Jay, Linus, and Paul. So in my book his data is just fine. If you want to disagree, that's fine too. But data is data.

 

OP wants to multitask? Having a browser open on a video on screen 2 and word open on screen 3 hardly qualifies as multitasking. You could do that on an i3 without much of an impact. If he wanted to render a video while playing, then I'd say hands down Ryzen.

 

 

[jerubedo]

11 hours ago, valdyrgramr said:

That doesn't make you credible having a solid reputation does.  And, yes your previous data has been wrong hence why I am questioning it.

I posted one (possibly) erroneous source, which I added as an edit, by the way, after initially posting only the first three. I just happened to find it and it said the same thing so I added it. There were still THREE other sources which all corroborate each other. When there's data corroboration amongst multiple sources that usually confirms said data.

 

11 hours ago, valdyrgramr said:

That doesn't make you credible having a solid reputation does.  And, yes your previous data has been wrong hence why I am questioning it.

Furthermore, when his historical data matches data from the big names, it certainly DOES lend credibility.

[jerubedo]

@valdyrgramr, let's put aside credibility for a second. I'd like to really discuss the question that @Herman Mcpootis brought up about how this performance differential is happening. Let's look at just Steve's data for a sec. So yeah, AC:O wins on the 8700K because of hyperthreading. Far Cry 5 is tied and it is also known for utilizing 8 threads, which might be why they tied. But F1 favors frequency and IPC (like the other games where the 9600K wins in the other video) and the 9600K wins. And by a decent margin too.

 

Both the 9600K and 8700K boost to 4.3GHz. The 8700K had 3MB more cache. The architecture is the same.

 

11 hours ago, valdyrgramr said:

 

My other problem with it is the lack of information about the person doing it.  I've only heard of this person just now, haven't seen all this lining up the you keep claiming about it, and crap like this...

It doesn't come off as a reliable source.  Simply saying things like, "It's all he does!"  Doesn't make the source credible. 

Credible or not the guy plugged in his 9600K, and his RTX 2080, benchmarked it, SHOWS US THE GAMEPLAY with active charts, and then posted the results. It's still data. I doubt it's wrong. Not with the amount of testing he does. And if you want to see data matching the big names, take this one for example:

 

 

 Compare those results to reviews for the 8700K or the 1080 Ti and they are all in-line.

[jerubedo]

1 minute ago, valdyrgramr said:

Which could be faked.  We don't know, and that's why credibility matters.

I mean, I trust it myself. If you don't want to, that's your right. So then let's forget any of the videos I posted. We can just talk about Steve's data. 

 

AC:O was a victory for the 8700K because AC:O uncharacteristically uses all 12 threads, giving the 8700K a nice wide birth.

 

Far Cry 5 tied because 8 threads are used, giving the 8700K a 2 less thread advantage. Which means that the 9600K is STILL making up ground because it only has 6 threads.

 

F1 was a victory for the 9600K because it only uses 4 threads and favors IPC and frequency. 

 

That all points to there's some sort of IPC gain on 9th gen without any architecture change. 

[Princess Luna]

i5 9400F only makes sense on a b360... it's fine for an alternative to the ryzen 5 2600 although the latter has better use if multi-tasking is more present on your use-case.

 

i5 9600K is a bad value proposition, you are spending the premium tag already, great z390 motherboard and aftermarket cooling, might as well grab the i7 8700K which has more to it, clock for clock there's little to no difference in gaming, but there's enough of an improvement in multi-tasking.

 

the i5 is already near it's max usage just like older 4 cores i5 already were all maxed out in the latest AAA games, Hyper-Threading cuts the processor usage in half gives you much more breathing room for other stuff, it's worth it personally speaking.

 

I'd still favor a cheaper i7 8700 locked on b360 build over going all expenses on an i5 9600K from my personal experience even with a 1080 Ti there's not 'bottleneck' matter, you can still be pretty much gpu bound always still.

[jerubedo]

6 minutes ago, valdyrgramr said:

But, the frequency on stock is the same while the boost is slightly higher on the 8700k.  So, the frequency shouldn't be a considered factor only the IPC should be.

What? The all-core boost clock for both chips is 4.3GHz. That's why I said clocks are equal. In gaming, the all-core boost is what is used.

 

Sources:

 

https://www.tomshardware.com/reviews/intel-core-i5-9600k-coffee-lake-cpu,5922.html

 

https://www.gamersnexus.net/guides/3077-explaining-coffee-lake-turbo-8700k-8600k

 

And by the way: if the 8700K did have a higher all-core boost, it would make even less sense that the 9600K would ever win.

[jerubedo]

10 hours ago, Princess Cadence said:

i5 9400F only makes sense on a b360

I agree, but we need to recommend Z390 since B360 may not support 9th gen out of the box.

 

10 hours ago, Princess Cadence said:

i5 9600K is a bad value proposition, you are spending the premium tag already, great z390 motherboard and aftermarket cooling

I'm not sure where you are, but here in the US the 9600K is only $264 and you can pair it with the Z390 UD for $109. The 8700K is $365 (and I guess you COULD use the UD for that too or a Z370 for a similar price). That's a $100 difference and for a STRICTLY gaming rig, the extra cost doesn't make much sense.

 

10 hours ago, Princess Cadence said:

the i5 is already near it's max usage just like older 4 cores i5 already were all maxed out in the latest AAA games, Hyper-Threading cuts the processor usage in half gives you much more breathing room for other stuff, it's worth it personally speaking.

I agree about the breathing room, however the older i5s are only struggling in a hand few of games like Assassin's Creed Origins and a few other games. Those games are still far and few between and I wouldn't call it a trend yet. Plenty of AAA releases are still coming out preferring frequency/IPC over threads. Once that shifts on a big scale then I'd get on board the more thread train for gaming.

 

10 hours ago, Princess Cadence said:

I'd still favor a cheaper i7 8700 locked on b360 build over going all expenses on an i5 9600K from my personal experience even with a 1080 Ti there's not 'bottleneck' matter, you can still be pretty much gpu bound always still.

I certainly agree about not being bottlenecked either way. And the 8700 here is $294, making the 9600K a better value still.

 

10 hours ago, valdyrgramr said:

I thought you were talking about the turbo before by previous wording.  I don't remember you saying the all-core boost.  Unless I missed that.

Well I am. When a CPU turbos in gaming, it uses the all-core turbo. So both CPUs are running at 4.3GHz in a gaming load.

[jerubedo]

1 minute ago, valdyrgramr said:

I meant the other turbo where the i7's is 4.7 and the i5's was 4.6.  Not exactly the same thing.  I came into this a bit later.

That's the single core Turbo Boost.

 

So for turbo boost:

 

i7:

1C: 4.7GHz
2C: 4.6GHz
3C: 4.5GHz
4-5C: 4.4GHz
6C: 4.3GHz

 

i5:

1 Core - 4.6
2 Core - 4.5
4 Core 4.4
6 Core 4.3

 

all-core, of course, being the 6 core from above.

 

So now that we are on the same page, any reason the 9600K is performing better at the same frequency and with less cache?

[jerubedo]

3 minutes ago, valdyrgramr said:

I'm aware I just based something off of an assumption since I came into this late.

Ah, got it  

[FinalFantasyBros]

47 minutes ago, valdyrgramr said:

Which could be faked.  We don't know, and that's why credibility matters. 

Are you really trying to imply that both remaining sources have gotten together, put together similar fake results and then posted them? I agree that the third source was questionable, and @jerubedo admitted that after taking a closer look as well (I guess they posted hastily). The other 2 look good though. In looking at all the data thats here I have to agree with them. The 9600k performs better in many games than the 8700k.

[jerubedo]

10 hours ago, FinalFantasyBros said:

Are you really trying to imply that both remaining sources have gotten together, put together similar fake results and then posted them? I agree that the third source was questionable, and @jerubedo admitted that after taking a closer look as well (I guess they posted hastily). The other 2 look good though. In looking at all the data thats here I have to agree with them. The 9600k performs better in many games than the 8700k.

Eh, it doesn't matter. Honestly we're arguing over such small performance differences. At most we're talking about 6-7% and on average more like 5% (when there is a difference). At the end of the day, ALL of the data out there shows that in gaming a 9600K at least matches an 8700K. I was just pointing out that there IS data out there that shows the 9600K BEATING the 8700K in certain games, and in more than just a few (and I'm also still wondering why). But I digress, we're just getting off topic at this point. My original point still stands which is relevant to the OP:

 

For GAMING:

 

The 9400F will still beat the 2600X even when it's OCed for $5 cheaper (CPU+Mobo in the US)

Of course this means it also beats the 2600 by an even bigger margin, but for $20 more (CPU+Mobo in the US)

It even beats an OCed 2700x by a small margin for much less (in the US and most parts of the world).

The 9600K will still match (or occasionally beat) an 8700K for $100 less (in the US)

 

And in that regard review sites back that up. Hopefully this is enough reputable sources to back it up (and note that the 9400F is faster than the 8400. I won't argue about how much faster since no professional sites reviewed it, but it does sport 100MHz more , so the FPS difference is even greater than what's shown below).

 

The results on the left from Tom's is for a manually set all-core 4.2GHz OC on both the 2700x and the 2600x. If you want to see the stock performance as well, here's the original article. I only included the OCed since it's what's more relevant here: the 9400F beating the 2700X/2600X with an OC.

 

 https://www.tomshardware.com/reviews/intel-core-i5-9600k-coffee-lake-cpu,5922-5.html

 

Here's the other articles I used:

 

https://www.anandtech.com/show/12625/amd-second-generation-ryzen-7-2700x-2700-ryzen-5-2600x-2600/15

 

https://www.techpowerup.com/reviews/Intel/Core_i5_9600K/13.html

 

PLEASE feel free to look through the entirety of the  articles to prove I didn't cherry pick. This is the case in almost every game with only 3 results favoring Ryzen out of 22 total tests across these 3 sites

 

@Herman Mcpootis @valdyrgramr @seoz @LienusLateTips if 3 major sources doesn't convince you, I don't know what will. 

 

 

 

It's also very hard to argue with this:

 

Again note that the 9400F performs even better than the 8400

[jerubedo]

2 hours ago, valdyrgramr said:

@jerubedo Your techpowerup source is pretty bad.

How is an 8500 doing better than an 8700k?



This happened with Ryzen as well in another game where a 1600x beat an 1800x.

In this same source, a pentium beat an i9.  A 2200g beat a 2400g.  This is a very bad source.

Firstly, techpowerup is a large and reputable source. Secondly all of those results are within margin of error. It just means that there was not enough of a CPU bottleneck to have any meaningful difference in the results between any of those CPUs. This happens all the time in very GPU bound games. Which CPU shows up on top of luck if the draw in those cases.

 

And again their results match both Anand and Tom's in cases where they tested the same games. 

 

Namely Rise of the Tomb raider which was tested on both Anandtech and Techpowrup. 8400 gets 148 on Andand and 149 on Tech. 2700x/2600x pulls ahead by a good amount because they are using much more powerful 3200MHz C14 RAM vs Anand who used 2933 C16. 

 

To call Techpowerup a bad source is absolutely absurd. 

[jerubedo]

24 minutes ago, valdyrgramr said:

I went over the charts in your source and every chart was bs in it.

AC Origins the 8th and 9th gen i5s beat the i7s of both gens.  BF 1 1600x beating a 2700.  Again with this Ryzen 5 beating that Ryzen 7 in Civ.  8th gen i5s beating 8th Gen i7s in F1.  I can keep going, but this doesn't make any sense rather you like it or not.  

 

^0 fucking sense.  You're telling me a pentium is better than a 9900k?  Well, I might as well go buy that pentium considering the savings.  e.e  edit = didn't realize Steve tested a different year F1.  But still, these charts are really bad.  There's even a couple of charts where a 1300x is a beating a 2700 in gaming.  XD

What aren't you understanding about GPU bound examples? The pentium does not do "better" here. It does evenly. All of these results are within 3 FPS of eachother. And at those high levels of FPS that's margin of error. The CPUs just aren't a bottleneck at all. Do you understand the concept? 

[jerubedo]

7 minutes ago, valdyrgramr said:

I am seeing higher frames from a Pentium than an i9.  More is better.

Margin. Of. Error. They are functionally the SAME. And it's because the GPU is not being bottlenecked by any of the CPUs.

[jerubedo]

5 minutes ago, valdyrgramr said:

It isn't the same the number is higher.  These kind of results are apparent in every chart.  That makes that charts bad in general.

The number is indeed not the same. But they are FUNCTIONALLY the same if you calculate the standard deviation and then figure out the % error. At this point it's apparent that you aren't familiar with how margin of error works, nor how a lack of CPU bottleneck in any capacity can yield equal results across any number of CPUs. It could even be the case you'd see an older dual core "beating" the 9900K as well if they had tested that. 

 

At the end of the day all 3 major sources demonstrate my point. 

[LienusLateTips]

5 minutes ago, valdyrgramr said:

Not really, but whatever helps you sleep at night.  I am going to go buy that pentium now since it is superior according to your source.

Nonono, at this rate you should be going for a CELERON. C. E. L. E. R. O. N.

[jerubedo]

16 minutes ago, valdyrgramr said:

Not really, but whatever helps you sleep at night.  I am going to go buy that pentium now since it is superior according to your source.

Read Steve's article on margin of error. Again I can't stress enough that none of these processors beat each other. They were all functionally the same. 

 

You're picking apart source after source to prove your point. How about Tom's then? Or Anand? Do you have any data that counters my point? Show me data that shows a high number of games where ryzen beats the 8400. If you can do that THEN you have an argument. 

[jerubedo]

35 minutes ago, valdyrgramr said:

I wasn't part of the other CPU argument, so I am not sure why you're including me in that.  I am also amazed you missed the point.

I got your point: it's that data should be questioned and that reliability of source is important. I can understand that point MORE with the videos previously posted. But TechPowerUp? Come on. The "bad" data you've pointed out I've explained. They are GPU bound scenarios where virtually all the CPUs will perform within margin of error.