HOW is THIS for Gamers??

[Emily Young]

45 minutes ago, Canada EH said:

Yes, people are lazy and I would say even Linus and his team are lazy when it comes to reporting to their viewers.

What more would you want us to report?

[D13H4RD]

5 hours ago, Paddi01 said:

is the 8700k still the best gaming cpu?

Nope. In terms of raw performance, the 9900K is. 

 

Not by a whole lot in games, mind you, but still enough for it to technically be the best in terms of performance. 

[JoostinOnline]

The i3/i5/i7 video has been inaccurate for at least a year now. It's BS that he's claiming it's a new occurrence.

[JCHelios]

 

9 hours ago, JoostinOnline said:

The i3/i5/i7 video has been inaccurate for at least a year now. It's BS that he's claiming it's a new occurrence.

 

Not that I minded losing hyperthreading on the Core i3 line, since it got upgraded from dual-core to quad-core, but yeah.  

 

[Sakkura]

7 hours ago, GabenJr said:

What more would you want us to report?

Would help if you reported accurately on the price difference vs. the 2700X and 8700K. Your video claims the 9900K is 30-40% more expensive, when the truth is it costs 49% more than the 8700K and a whopping 90% more than the 2700X.

 

Would also help if you picked a motherboard that wasn't VRM throttling...

[Nogr]

I'm curious, will you test this addressing the issues mentioned here 

Hardware Unboxed explained thermals

[Ross Siggers]

14 hours ago, Canada EH said:

I laughed at his facial expression LOL

EXPOSED. THE GIG IS UP LINUS.

[MattMatt]

5 hours ago, Nogr said:

I'm curious, will you test this addressing the issues mentioned here 

Hardware Unboxed explained thermals

Wow! This is a HUGE deal, I know nothing about VRM limits or what that even is really. Time to learn more about this as it cries out as a huge deception tactic from intel to me. I always thought any motherboard could fully utilise a stock chip that was socket compatible.

[Roamable]

19 hours ago, SolarNova said:

I mean seriosly, for gaming , an i7 sandybridge with a good OC .. maybe 4.8-5ghz, or 4.6-4.8ghz for a 6c/12t chip, is more than capable even now.

 

3

This is the thing, so much "old" hardware is more than capable BUT the hardware being released constantly is to keep feeding the industry and churning profits. Making people buy more regularly basically. 

 

Maybe I'm too contrarian for a lot of folks when I say that, but seriously games haven't advanced so quickly that hardware needs constant upgrades.

 

I upgrade every 3-4 years and even then it's not because I can't play the newest games in max settings.

[Hunter259]

7 hours ago, Sakkura said:

Would help if you reported accurately on the price difference vs. the 2700X and 8700K. Your video claims the 9900K is 30-40% more expensive, when the truth is it costs 49% more than the 8700K and a whopping 90% more than the 2700X.

 

Would also help if you picked a motherboard that wasn't VRM throttling...

Would help if you even knew what you were talking about. This board actually uses the TDP limits unlike many boards now and in the past.

 

[Sakkura]

12 minutes ago, Hunter259 said:

Would help if you even knew what you were talking about. This board actually uses the TDP limits unlike many boards now and in the past.

 

And the cTDP is throttling the CPU because of weak VRMs. Try again.

[Princess Luna]

19 hours ago, Enderman said:

No...... watch the video.

The 9900K is.

Actually it's looking like the i7 9700K is because Hyper-Threading is being derp on games.

[Enderman]

10 minutes ago, Princess Cadence said:

Actually it's looking like the i7 9700K is because Hyper-Threading is being derp on games.

Just because the 9700k performs better in one or two DX12 benchmarks does not mean it is the best overall.

Literally at the beginning of the benchmarks they put a "weighted gaming performance" chart where you can CLEARLY see the 9900K is better.

[Canada EH]

1 hour ago, Enderman said:

Just because the 9700k performs better in one or two DX12 benchmarks does not mean it is the best overall.

Literally at the beginning of the benchmarks they put a "weighted gaming performance" chart where you can CLEARLY see the 9900K is better.

Yeah this man is right, plus the games they use and the spots in the games they use.

Plus whats the algorythm they use for the "weighted" perf

[JoostinOnline]

10 hours ago, JCHelios said:

 

 

Not that I minded losing hyperthreading on the Core i3 line, since it got upgraded from dual-core to quad-core, but yeah.  

 

It makes me wonder if he really didn't know.

[1GNAS]

eh, donno if okay to ask here, cause youtube comments is a cesspool. But classic question, first background music song name?

[Hunter259]

12 hours ago, Sakkura said:

And the cTDP is throttling the CPU because of weak VRMs. Try again.

You are making an assumption that that is why and not just the way Asus has always set it up.

[leadeater]

On 10/23/2018 at 6:00 AM, Sakkura said:

And the cTDP is throttling the CPU because of weak VRMs. Try again.

If you set the cTDP higher which other reviewers have done (tested 95W and 300W cTDP) on the Asus Maximus XI Hero the performance and Turbo Frequencies increase accordingly, that motherboard handles it just fine and the VRMs do not throttle, temperature or otherwise.

 

Intel default SVID sets long term Turbo TDP to 95W and Turbo Boost Short Power Max to 120W with Turbo Boost Power Window of 16 seconds. This is the Intel defined spec, that's actually not what most motherboard set as default and configure/override to settings that are not Intel default.

 

You are seeing a drop off in performance because the Turbo Boost Power Window has been meet so TDP throttling in engaged, that's a CPU thing not a motherboard thing and with Intel SVID set every single motherboard will do the same.

 

You can see it in the GN review on the Power over Time graph for the Fortnight streaming graph, it drops to ~95W after the boost window is met. Hardware Unboxed latest video also shows this and talks about exactly this, some of his opinions I don't fully agree with but he correctly identifies what is actually going on and the settings that cause it.

[Sakkura]

1 minute ago, leadeater said:

If you set the cTDP higher which other reviewers have done (tested 95W and 300W cTDP) on the Asus Maximus XI Hero the performance and Turbo Frequencies increase accordingly, that motherboard handles it just fine and the VRMs do not throttle, temperature or otherwise.

 

Intel default SVID sets long term Turbo TDP to 95W and Turbo Boost Short Power Max to 120W with Turbo Boost Power Window of 16 seconds. This is the Intel defined spec, that's actually not what most motherboard set as default and configure/override to settings that are not Intel default.

 

You are seeing a drop off in performance because the Turbo Boost Power Window as been meet so TDP throttling in engaged, that's a CPU thing not a motherboard thing and with Intel SVID set every single motherboard will do the same. 

But what you get out of the box with most motherboards is what's most relevant to test, in my opinion. It's also what's more in line with the performance Intel is making people expect, even if it's technically outside their specification.

 

And when boards do set a lower TDP that would presumably be related to having weaker VRMs (like many Z370 boards do compared to Z390 models). Like a 300W cTDP is no joke on a plain old 4-phase board.

 

Intel probably should have set a higher TDP for the 9900K. 95W is technically possible, but those advertised turbo speeds are going to be hard to sustain.

[leadeater]

1 minute ago, Sakkura said:

But what you get out of the box with most motherboards is what's most relevant to test, in my opinion. It's also what's more in line with the performance Intel is making people expect, even if it's technically outside their specification.

 

And when boards do set a lower TDP that would presumably be related to having weaker VRMs (like many Z370 boards do compared to Z390 models). Like a 300W cTDP is no joke on a plain old 4-phase board.

 

Intel probably should have set a higher TDP for the 9900K. 95W is technically possible, but those advertised turbo speeds are going to be hard to sustain.

Setting the Intel SVID was an active choice by those that did it, I don't know what the defaults actually are for that board but it's not Intel SVID.

[leadeater]

9 minutes ago, Sakkura said:

But what you get out of the box with most motherboards is what's most relevant to test,

For a motherboard review yes, not for a CPU review. The problem is, as demonstrated by all this "tech drama", that you can't accurately compare reviews because as it seems and has been mostly known for a while now every motherboard sets different parameters.

 

I don't think it's much of a burden on reviews to run stock tests in a CPU review as actually stock, Intel's stock, then also test motherboard defaults as well as reviewer optimized defaults i.e. Overclocked.

 

You shouldn't be reviewing the motherboard in a CPU test.

[Sakkura]

1 minute ago, leadeater said:

For a motherboard review yes, not for a CPU review. The problem is, as demonstrated by all this "tech drama", that you can't accurately compare reviews because as it seems and has been mostly known for a while now every motherboard sets different parameters.

 

I don't think it's much of a burned on reviews to run stock tests in a CPU review as actually stock, Intel's stock, then also test motherboard defaults as well as reviewer optimized defaults i.e. Overclocked.

 

You shouldn't be reviewing the motherboard in a CPU test.

Well, yeah, but the CPU performance is dependent on the motherboard. Since most motherboards will be more aggressive than Intel's official spec (see also MCE, or even the non-K OC shenanigans), I think what they standardize on for testing should be more representative of that. Intel's official spec and SVID may say one thing, but it's getting further and further from what most people actually experience.

 

Otherwise Intel will continue to try and have their cake and eat it. So many cores, so many clocks, so few watts.

[leadeater]

9 minutes ago, Sakkura said:

Well, yeah, but the CPU performance is dependent on the motherboard. Since most motherboards will be more aggressive than Intel's official spec (see also MCE, or even the non-K OC shenanigans), I think what they standardize on for testing should be more representative of that. Intel's official spec and SVID may say one thing, but it's getting further and further from what most people actually experience.

 

Otherwise Intel will continue to try and have their cake and eat it. So many cores, so many clocks, so few watts.

Oh I agree it's more representative but I still think testing Intel's SVID should be done, that's going to be the only way that Intel will actually look at changing their TDP specification. If it's widely shown by every reviewer just how performance constrained the CPU is under Intel's spec that might just be enough.

 

Also using Intel's SVID spec you can't actually reach the all core turbo of 4.7Ghz, from what I've seen you need ~160W to reach it and the default 120W is far off from that. 

[App4that]

On 10/22/2018 at 10:00 AM, Sakkura said:

And the cTDP is throttling the CPU because of weak VRMs. Try again.

Those "weak" VRMs allowed Steve at Gamers Nexus to hit 5.3GHz on his 9900k during his live stream overclocking of the 9900k.