Speculation: i9-9900K is Intel's last hurrah in gaming

[TheELF]

Right now AMD can make chips that are 90% of Intel perf and offer them from launch at 25-40% cheaper than Intel can and STILL make money hand over fist because their yields are so good. That's game. 2700x is their top chip, but it's the most mainstream and it's cheap to manufacture. The 9900k is more expensive to make so it has to be a halo product which in turn means sales volume will be lower. Intel cannot compete on price because of their monolithic dies. AMD is winning because Intel is chasing their strategy. Everything Intel has done since Ryzen launched, in terms of product segmentation, has been reactionary. AMD beating Intel to a new node just compounds the issues Intel faces.

Neither the 9900k or the 2700x are the breadwinners for intel or amd,those are just halo products they use to make headlines.
What makes money for both companies (outside of server) are the cheap CPUs that are being bought in huge amounts,like for example intel's quad cores or for amd the 4/8 6/12 cores.
Look at the hardware.fr bench someone linked to earlier,intel's i3-8350k has a huge advantage in performance per core in games as well as in apps compared to amd's 2200,it's even faster in both games and apps than the 2400 that has SMT enabled,this is far far away from 90% of intel's perfromance.
https://www.hardware.fr/articles/965-3/performances-jeux-3d.html

GAMES

Spoiler

APPS

Spoiler

 

[Timorous]

I actually think Zen 2 would come pretty close IPC wise to Skylake when it comes to gaming. A lot also depends on how AMD much AMD can improve latencies as well. As it stands, I believe Zen+ is about 10% behind per clock, or thereabouts. I'm not sure if it can actually overtake Skylake at gaming per clock, I don't have a crystal ball. But I'm not even going to try to extrapolate an estimate from an unrelated and unknown scientific workload and try to guess how much effect that has on actual gaming performance.

The main thing I'm more skeptical about is the ability of Zen 2 to clock to 5GHz. I guess even 4.5GHz with Skylake level IPC would bring Zen 2 'close enough' (within 5% actual in game framerates) to a 9900K that it doesn't really matter what CPU you choose for gaming. But to actually exceed a 9900K like you confidently predicted in the OP and subsequent posts, I have serious doubts. As I said, its one thing to draw close, its another thing to beat a 9900K in the majority of the games to declare the 9900K 'dethroned' as the 'best gaming CPU'.

I don't think we are *that* far off in our expectations - we're only talking about a 10% swing. I think Zen 2 would probably be 90 - 95% 9900K level whereas you think it would be 100 - 105% 9900K level? Would that be a correct assumption on my part?

Techspot did an IPC comparison between the several CPUs which can be found here https://www.techspot.com/article/1616-4ghz-ryzen-2nd-gen-vs-core-8th-gen/page3.html. From this we have some numbers we can use to try and get a rough estimate.
@4ghz the 8700K is about 11.4% faster than the 2600X @4ghz. (same speed, same cores, same threads)

In a previous post you showed that an increase of 21% in clockspeed represented a 3-4% increase in framerate for the 8700K. Lets just say that performance increase = 20% of the clockspeed increase. In doing that the framerates of a 5Ghz 8700K should increase by about 5% over a 4Ghz 8700K.

Adding it all up gives a 17% advantage to a 5Ghz 8700K over a 4Ghz 2600X. Based on the Techspot 35 game comparison @1080p between a 5Ghz 8700K and a 4.2Ghz 2700X giving only a 9% advantage to the 8700K I think that 17% figure is probably about right if not slightly favourable for the 8700K.

This puts the 2600X@4Ghz at 85.5% of the gaming performance of a 8700K@5Ghz meaning to match your low end estimate Zen2 only needs to beat Zen+ by 5.3% which based on the IPC comparison linked is less than the performance uplift going from Zen to Zen+ (6.35%). That does not seem plausible. Getting up to your top end estimate would take an 11.1% performance increase which just needs a 10% IPC increase and a 10% clockspeed increase to achieve and both of those seem doable for a 2nd iteration architecture on a new node.

I think it could land anywhere between 95 - 105% of the 9900K in gaming. If its like Zen+ and the average IPC uplift is lower than the gaming IPC uplift or the clockspeed increase is 20% or more then it could actually be a good chunk faster than the 9900K in gaming, if not then even worst case I expect it to be within touching distance for gaming and faster for non gaming.

 

[beginner99]

The problem with Skylake-X with gaming appears not to be the switch to exclusive but rather that the L3 latency is slow due to being on the mesh.

Pretty confident Intel increased the L2 because of AVX-512 (and nothing else), and probably is increasing it in Icelake for the same reason.

Exactly. And back after Skylake-X was released at least a few people on this forum mentioned that overclocking the uncore eg. the mesh, gaming performance increased a lot.

 

[DrMrLordX]

Don't dismiss TSMC as a low-performance device manufacturer. They are aiming for leadership in HPC and to compete squarely with Intel. AMD chose them ahead of GF.

I won't, but also don't dismiss GF's clockspeed capabilities with their HP processes. The only reason why anyone expected GF 7nm HP to help Matisse hit 5 GHz is that it was supposed to be the same process IBM had intended to use for POWER10. Now that entire deal is scuttled. It's unlikely that AMD set 5 GHz design targets for Rome with TSMC, and it's likely that we'll see CCXs with those design targets in Matisse now that Matisse is being fabbed by TSMC.

No idea how or if an older TSMC 20nm node relates to all this, since that may have been a custom node with a high-clockspeed design target for a specific client. Kind of like GF's 32nm HP SOI which AMD never used . . . or GF's 14nm HP which AMD also never used.

 

[Zucker2k]

Techspot did an IPC comparison between the several CPUs which can be found here https://www.techspot.com/article/1616-4ghz-ryzen-2nd-gen-vs-core-8th-gen/page3.html. From this we have some numbers we can use to try and get a rough estimate.
@4ghz the 8700K is about 11.4% faster than the 2600X @4ghz. (same speed, same cores, same threads)

In a previous post you showed that an increase of 21% in clockspeed represented a 3-4% increase in framerate for the 8700K. Lets just say that performance increase = 20% of the clockspeed increase. In doing that the framerates of a 5Ghz 8700K should increase by about 5% over a 4Ghz 8700K.

Adding it all up gives a 17% advantage to a 5Ghz 8700K over a 4Ghz 2600X. Based on the Techspot 35 game comparison @1080p between a 5Ghz 8700K and a 4.2Ghz 2700X giving only a 9% advantage to the 8700K I think that 17% figure is probably about right if not slightly favourable for the 8700K.

This puts the 2600X@4Ghz at 85.5% of the gaming performance of a 8700K@5Ghz meaning to match your low end estimate Zen2 only needs to beat Zen+ by 5.3% which based on the IPC comparison linked is less than the performance uplift going from Zen to Zen+ (6.35%). That does not seem plausible. Getting up to your top end estimate would take an 11.1% performance increase which just needs a 10% IPC increase and a 10% clockspeed increase to achieve and both of those seem doable for a 2nd iteration architecture on a new node.

I think it could land anywhere between 95 - 105% of the 9900K in gaming. If its like Zen+ and the average IPC uplift is lower than the gaming IPC uplift or the clockspeed increase is 20% or more then it could actually be a good chunk faster than the 9900K in gaming, if not then even worst case I expect it to be within touching distance for gaming and faster for non gaming.

The data is flawed to begin with, and worse, games don't scale lineally. Take the i5 8400 with it's limited 3.8GHz clock, for example, and compare it with both the 2700x and 8700k. It'll mess up your analyses above.

 

[Despoiler]

Neither the 9900k or the 2700x are the breadwinners for intel or amd,those are just halo products they use to make headlines.
What makes money for both companies (outside of server) are the cheap CPUs that are being bought in huge amounts,like for example intel's quad cores or for amd the 4/8 6/12 cores.
Look at the hardware.fr bench someone linked to earlier,intel's i3-8350k has a huge advantage in performance per core in games as well as in apps compared to amd's 2200,it's even faster in both games and apps than the 2400 that has SMT enabled,this is far far away from 90% of intel's perfromance.
https://www.hardware.fr/articles/965-3/performances-jeux-3d.html

GAMES

Spoiler

APPS

Spoiler

I'd argue that the people buying a 2200 or a 2400 are doing so with the intention of not buying a GPU. It's comical how bad Intel's iGPUs are. https://www.techspot.com/review/1574-amd-ryzen-5-2400g-and-ryzen-3-2200g/page4.html Also DDR4 2400 @ CL15 for everything that uses DDR4? That's not normal for either camp. I also didn't see what resolution they did these tests at. I could have missed it. I also checked the numbers before I put the claim up. https://www.techspot.com/review/1655-core-i7-8700k-vs-ryzen-7-2700x/ Realistic Intel @ 5.0Ghz vs AMD @ 4.2.

 

[dark zero]

I may be *completely* wrong but my opinion is that Intel "struggles to keep up with the development stride of AMD" not because of incompetence but because - like EVERY large company - they have bureucracy problems; they are too big and too dumb to do things right.

Until AMD starts whoppin' they' ass, at which point they will pull another C2D miracle out of their ass. By which i mean "out of the R&D department".

TLDR AMD can put out a better cpu than Intel; they cant "beat" Intel.

But now there are 3 more opponents to watch, 2 of them on ARM and one of them in x86. The x86 one is VIA. If they do the things right, they might catch Intel up way sooner than expecting.

Then there is Qualcomm, which is the least of the problems unless we take in consideration the advance of the 2 in one laptops. They might kill their Atom line.

And then there is Apple... For me is the most dangerous of all. Their improvements are no longer ignored and is matter of time to see an ARM laptop from them.

 

[moinmoin]

I feel Intel's clock advantage is being overstated. What counts both for mainstream as well as server and mobile is efficient performance. The 9900K may well be Intel's last hurrah in brute force performance, as indicated by the leaked Gigagbyte's OC guide stating that it needs water cooling and 244.8W for a 5GHz OC while still running hot at 90°C. This is not mainstream but a very particular gaming niche where any remaining OC capability at all cost trumps energy efficiency.

In the upcoming nodes both Intel and AMD need to look for ways to generate more performance at less heat.

 

[Vattila]

now [Intel has] 3 more opponents to watch

Please stay on topic. If you diverge, try at least to relate it to the discussion. Thanks.

 

[LTC8K6]

it needs water cooling and 244.8W for a 5GHz OC while still running hot at 90°C

If that's actually true, then who would even bother to buy the chip?

The chip would be a literal joke.

You could probably do better with the 9800X.

 

[PotatoWithEarsOnSide]

It's winter and it's cold, so time to buy a 9900k.

 

[Timorous]

The data is flawed to begin with, and worse, games don't scale lineally. Take the i5 8400 with it's limited 3.8GHz clock, for example, and compare it with both the 2700x and 8700k. It'll mess up your analyses above.

How is the data flawed? The analysis is for many reasons but its to give a rough ballpark not an exact forecast.

The techspot article also had the 8600K locked at 4Ghz and the 8700K was 2.9% faster although there was a large outlier in the form of Ashes, exclude that outlier and the performance advantage for the 8700K is 0.94% which is within margin of error.

I expect the i5 8400 to be about the same at 3.8Ghz since 4Ghz is only a 5.2% increase in clocks which should represent a 1% performance swing using the scaling from earlier in the thread, again margin of error stuff.

The problem with comparing the 8600K / 8400 to the 8700K or the 2600X is that they do not have SMT which skews the results if any of the games gain or lose performance due to SMT. This is why I compared the 8700K to the 2600X to keep the clock speeds, the core count and the thread count the same. It makes it more reliable to analyse and is closer to a pure architecture comparison rather than an end product comparison.

 

[moinmoin]

If that's actually true, then who would even bother to buy the chip?

The chip would be a literal joke.

You could probably do better with the 9800X.

This has been some time coming since Intel's TDP is very misleading (only covering all core at base frequency which is basically never the case).

As for the numbers, this is the guide straight from Gigabyte, the data is on page 13 and 14.

 

[LTC8K6]

This has been some time coming since Intel's TDP is very misleading (only covering all core at base frequency which is basically never the case).

As for the numbers, this is the guide straight from Gigabyte, the data is on page 13 and 14.

That's been posted several times. I've already said I don't believe it. Plus it's an ES chip. Plus there are some oddities in it, as others have pointed out.

But what's the point? Tomorrow is the day we will all find out.

 

[LTC8K6]

BTW, Intel has always clearly stated how they arrive at the published TDP number.

It's only misleading of you don't read what Intel posts.

Click on the ? next to Processor Base Frequency and TDP and Thermal Solution Specification.
https://ark.intel.com/products/186605/Intel-Core-i9-9900K-Processor-16M-Cache-up-to-5-00-GHz-

 

[DrMrLordX]

It's winter and it's cold, so time to buy a 9900k.

Hell, save some money. Buy some AMD Zoner's old 9590.

 

[Vattila]

The poll now favours the proposition with 44% for and 42% against, with 14% undecided.

Thanks for continuing the discussion with valuable input and opinion. Special thanks to those of you that have provided data to show the current Intel lead in gaming, and to those that have scrutinised the data. The gaming IPC discrepancy between Zen and Skylake/CFL indeed looks big — it may even be 18%, as one post here stipulates. But this, in and of itself, does not say much about Zen 2. You can view a big deficit as a big opportunity for improvement (low hanging fruit) or as an unsurmountable challenge.

So, let me reiterate my main point in a slightly different way — from the Intel viewpoint:

Would you expect 10nm Ice lake to beat 14nm Skylake/CFL? Would you expect it to surpass i9-9900K in gaming? Would you think Intel made wise architectural choices to achieve that? Do you think they could do that without compromising other workloads? Would you be excited? Yes? Then you better get excited about Zen 2, because it was designed to compete squarely against Ice Lake.

Regarding scepticism about frequency, I suspect this is to a large degree coloured by Intel's 10nm stumbles, as well as industry reports about the increasing difficulty in process development. But, without any proof, don't let that make you expect that the latest 7nm process advances cannot match Intel's 14nm process. And please note that 5 Ghz is not a special number. It is just ~15% beyond Zen on the so-so 12LP GF process. As pointed out, 5 GHz has been achieved on multiple diverse processes already.

Regarding gaming IPC, arguing that Zen 2 will stay behind i9-9900K leads to one of these logical fallouts:

1. Ice Lake is better for gaming, while Zen 2 is better for other workloads. Why do AMD and Intel diverge?
   
2. Zen 2 matches Ice Lake, but both will be worse than i9-9900K for gaming. Likely?
3. Zen 2 will fall flat. Ice Lake can do all workloads better, including gaming. This is a fallacy favouring Intel, unless it is argued with evidence.

Which is it?

 

[Vattila]

It's winter and it's cold, so time to buy a 9900k.

Hell, save some money. Buy some AMD Zoner's old 9590.

BTW, Intel has always clearly stated how they arrive at the published TDP number.

​

Please stay on topic. If you diverge, try at least to relate it to the discussion. Thanks.

 

[LTC8K6]

Please stay on topic. If you diverge, try at least to relate it to the discussion. Thanks.

I think the TDP of the chip is well within the scope of the thread, especially since I am only responding to a claim in the thread about the TDP.

 

[Vattila]

I think the TDP of the chip is well within the scope of the thread, especially since I am only responding to a claim in the thread about the TDP.

Ok. I guess so. But I'd like to hear more from you how so. I was thinking a little bit about this, and I guess we may see TDP become more of an issue in reviews, especially if the gaming IPC deficit continues to narrow. For example, reviews may look at "gaming champion at 65W", "best gaming CPU for a quiet system", for example.

I do expect 7nm Zen 2 to be a big improvement in power-efficiency.

 

[happy medium]

Will Zen 2 with a 12% IPC uplift at 4.5ghz best a 9900k at 5.0?

This is what I expect.

 

[epsilon84]

Techspot did an IPC comparison between the several CPUs which can be found here https://www.techspot.com/article/1616-4ghz-ryzen-2nd-gen-vs-core-8th-gen/page3.html. From this we have some numbers we can use to try and get a rough estimate.
@4ghz the 8700K is about 11.4% faster than the 2600X @4ghz. (same speed, same cores, same threads)

In a previous post you showed that an increase of 21% in clockspeed represented a 3-4% increase in framerate for the 8700K. Lets just say that performance increase = 20% of the clockspeed increase. In doing that the framerates of a 5Ghz 8700K should increase by about 5% over a 4Ghz 8700K.

Adding it all up gives a 17% advantage to a 5Ghz 8700K over a 4Ghz 2600X. Based on the Techspot 35 game comparison @1080p between a 5Ghz 8700K and a 4.2Ghz 2700X giving only a 9% advantage to the 8700K I think that 17% figure is probably about right if not slightly favourable for the 8700K.

This puts the 2600X@4Ghz at 85.5% of the gaming performance of a 8700K@5Ghz meaning to match your low end estimate Zen2 only needs to beat Zen+ by 5.3% which based on the IPC comparison linked is less than the performance uplift going from Zen to Zen+ (6.35%). That does not seem plausible. Getting up to your top end estimate would take an 11.1% performance increase which just needs a 10% IPC increase and a 10% clockspeed increase to achieve and both of those seem doable for a 2nd iteration architecture on a new node.

I think it could land anywhere between 95 - 105% of the 9900K in gaming. If its like Zen+ and the average IPC uplift is lower than the gaming IPC uplift or the clockspeed increase is 20% or more then it could actually be a good chunk faster than the 9900K in gaming, if not then even worst case I expect it to be within touching distance for gaming and faster for non gaming.

I understand your calculations, but I think the point that you missed is that the actual 'gaming IPC' of Intel is far more than 10% ahead of Ryzen per clock. It's about 10% ahead at 'reasonable' resolutions like 1080p but as I alluded to earlier, that is because the GPU still greatly influences the average frame rate at that resolution.

Run a game at an unrealistically low resolution like 720P and you'll know what I mean. The closer you get to a 100% CPU limited scenario, the further Intel draws ahead.

For example, using the TPU 2600X 720P gaming aggregate scores:

https://tpucdn.com/reviews/AMD/Ryzen_5_2600X/images/perfrel_1280_720.png

A 2600X is almost 20% behind a 8700K. Sure, it has a slight clockspeed deficit, but even assuming clockspeed parity we are still looking at a 16 - 17% deficit per clock. This is with a GTX 1080 by the way. Run a 2080 Ti at this unrealistically low resolution and I'm pretty sure the 8700K will end up 20% faster per clock, or more. This is the true 'gaming IPC' deficit between AMD and Intel, not the GPU diluted % you see at 1080P or above.

The more you remove the GPU from the equation, the more Intel draws ahead, which is hardly rocket science.

The reality of GPU limitations in real world gaming that prevents Intel from truly pulling away from Ryzen at 1080P and above will also work against AMD trying to surpass Intel in gaming. Any significant advantage AMD can potentially have (most likely IPC as I seriously doubt Zen 2 will outclock a 9900K) will be seriously diluted by GPU bottlenecks.

So basically what is 'saving' AMD these days in gaming will also work against it when it comes to truly outclassing a 9900K.

You literally have to reverse the tables on Intel to take the gaming crown - that is by doing what Intel is doing today WR to AMD, which is higher IPC *and* higher clocks. One without the other wont get it done. It will draw AMD closer, but it wont beat a 9900K outright. That's my prediction and I'll stand by that unless new data is given that may convince me otherwise.

 

[Vattila]

the GPU still greatly influences the average frame rate [at 1080p]

Yeah. As you point out, it is wrong to equate the percentage lead in a 1080p benchmark with the same percentage IPC advantage, even with the competitors at same frequency.

I made a spreadsheet to calculate the numbers, taking the GPU-bound part of the frame time into account. Using a 16% frequency lead for i9-9900K over Ryzen 2700X (5.0/4.3 = ~1.16), and an arbitrary 18% gaming IPC advantage (as claimed by an earlier post here), we get 37% gaming IPS (instructions per second) advantage for Intel. Then, to recreate the PT study result, which shows 16% lead for Intel (14% loss for AMD), I have to make 57% of the frame time GPU-bound.

Spoiler: Gaming CPU benchmark competitive analysis with +18% gaming IPC for Intel

Frequency advantage for Intel: 16%
Gaming IPC advantage for Intel: 18%
Gaming IPS advantage for Intel: 37%

Intel frame time: 10.0 ms 100%
GPU-bound part of frame time: 5.7 ms 57%
CPU-bound part of frame time: 4.3 ms 43%

AMD frame time: 11.6 ms 100%
GPU-bound part of frame time: 5.7 ms 49%
CPU-bound part of frame time: 5.9 ms 51%

Frame rate for Intel: 100 fps 100%
Frame rate for AMD: 86 fps 86%

Benchmark lead for Intel: 16%
Benchmark loss for AMD: 14%

(Bold numbers estimated/arbitrary, the rest calculated.)

If we assume the GPU-bound part of the frame time is even bigger, say 60%, Intel's gaming IPC advantage has to be 20% to arrive at the same 16% benchmark lead.

Spoiler: Gaming CPU benchmark competitive analysis with +20% gaming IPC for Intel

Frequency advantage for Intel: 16%
Gaming IPC advantage for Intel: 20%
Gaming IPS advantage for Intel: 40%

Intel frame time: 10.0 ms 100%
GPU-bound part of frame time: 6.0 ms 60%
CPU-bound part of frame time: 4.0 ms 40%

AMD frame time: 11.6 ms 100%
GPU-bound part of frame time: 6.0 ms 52%
CPU-bound part of frame time: 5.6 ms 48%

Frame rate for Intel: 100 fps 100%
Frame rate for AMD: 86 fps 86%

Benchmark lead for Intel: 16%
Benchmark loss for AMD: 14%

(Bold numbers estimated/arbitrary, the rest calculated.)​

Likewise, we can arrive at a lower IPC advantage for Intel by assuming the GPU-bound part is smaller.

In conclusion, without knowing the GPU-bound part of the frame time, IPC can not be deduced.

 

[epsilon84]

Yeah. As you point out, it is wrong to equate the percentage lead in a 1080p benchmark with the same percentage IPC advantage, even with the competitors at same frequency.

I made a spreadsheet to calculate the numbers, taking the GPU-bound part of the frame time into account. Using a 16% frequency lead for i9-9900K over Ryzen 2700X (5.0/4.3 = ~1.16), and an arbitrary 18% gaming IPC advantage (as claimed by an earlier post here), we get 36% gaming IPS (instructions per second) advantage for Intel. Then, to recreate the PT study result, which shows 16% lead for Intel (14% loss for AMD), I have to make 57% of the frame time GPU-bound.

Spoiler: Gaming CPU benchmark competitive analysis with +18% gaming IPC for Intel

Frequency advantage for Intel: 16%
Gaming IPC advantage for Intel: 18%
Gaming IPS advantage for Intel: 37%

Intel frame time: 10.0 ms 100%
GPU-bound part of frame time: 5.7 ms 57%
CPU-bound part of frame time: 4.3 ms 43%

AMD frame time: 11.6 ms 100%
GPU-bound part of frame time: 5.7 ms 49%
CPU-bound part of frame time: 5.9 ms 51%

Frame rate for Intel: 100 fps 100%
Frame rate for AMD: 86 fps 86%

Benchmark lead for Intel: 16%
Benchmark loss for AMD: 14%

(Bold numbers estimated/arbitrary, the rest calculated.)

If we assume the GPU-bound part of the frame time is even bigger, say 60%, Intel's gaming IPC advantage has to be 20% to arrive at the same 16% benchmark lead.

Spoiler: Gaming CPU benchmark competitive analysis with +20% gaming IPC for Intel

Frequency advantage for Intel: 16%
Gaming IPC advantage for Intel: 20%
Gaming IPS advantage for Intel: 40%

Intel frame time: 10.0 ms 100%
GPU-bound part of frame time: 6.0 ms 60%
CPU-bound part of frame time: 4.0 ms 40%

AMD frame time: 11.6 ms 100%
GPU-bound part of frame time: 6.0 ms 52%
CPU-bound part of frame time: 5.6 ms 48%

Frame rate for Intel: 100 fps 100%
Frame rate for AMD: 86 fps 86%

Benchmark lead for Intel: 16%
Benchmark loss for AMD: 14%

(Bold numbers estimated/arbitrary, the rest calculated.)​

In conclusion, without knowing the GPU-bound part of the frame time, IPC can not be deduced.

Yes, that's precisely the point I'm trying to make, thank you for doing the maths on this.

Those parameters (GPU bound frames) will be reduced if PT had used a 2080 Ti , for example. We'll probably be looking at close to a 20% margin (instead of 16%) in favour of the 9900K with a 2080 Ti at 1080P... which unsurprsingly also matches up closely with the 8700K vs 2600X 720P results from TPU using a GTX 1080:
https://tpucdn.com/reviews/AMD/Ryzen_5_2600X/images/perfrel_1280_720.png

Without having actual data on hand, I'd estimate that the 'GPU bound frames' would be a similar % comparing a 1080 @ 720p and a 2080 Ti @ 1080p

 

[Abwx]

For example, using the TPU 2600X 720P gaming aggregate scores:

https://tpucdn.com/reviews/AMD/Ryzen_5_2600X/images/perfrel_1280_720.png

A 2600X is almost 20% behind a 8700K. Sure, it has a slight clockspeed deficit, but even assuming clockspeed parity we are still looking at a 16 - 17% deficit per clock.

The 2600X is 16.4% behind the 8700K in the link you posted, but if it was clocked at 4.3 all cores and 4.7 single core boost it would still be 16-17% behind , that is at a 16-17% deficit...?.