Confirmed: Core i9 CPU Series will get soldered heatspreader

If a game only uses X amount of threads and all those threads are locked at 4.3ghz, that is the max possible performance possible on that CPU.

Threads with no use are downclocking which is normal behavior for any CPU not constrained by TDP limits.
So yes, the argument it's not overclocked is a poor one at that due to the fact it's already near its limit of clock speeds on the threads that the game is utilizing.

Which I say it is a fair comparison considering the intel parts can overclock much more than the other ryzen CPUs.

 

Once more. No matter what you think, 2700X in that video is not overclocked.

 

And no matter what you think, the results are comparable.
Due to aforementioned reasons above.

So the argument is a wash.

And lasty, yes PBO is overclocking.
You can argue with AMD if you don't agree with it.

Precision Boost Overdrive (PBO) is an opportunistic automated overclocking mechanism found in various AMD processors that pushes the system power budget beyond its rated specifications in order to allow Precision Boost to act more aggressively and achieve higher performance. As with all other overclocking mechanisms, using this feature voids warranty.

As such, it is hardly different than Intel MCE with end results.

 

That's been my experience as well. When I saw the unused cores downclocking in the video I immediately recognized that it was not manually overclocked, and the video's description confirmed this. On the other hand, the Intel CPUs were running at 4.8 GHz, which is beyond what even MCE can do, and the video description confirmed this as well. So the AMD CPUs were running with no manual overclock (only PBO enabled) while the Intel CPUs were manually overclocked to 4.8 GHz (no MCE).

If the Intel CPUs used MCE then it might have been an apples-to-apples comparison, but this was not the case. At any rate, it seems @Agent-A01 found a video that showed a large difference and posted it without checking (I guess it looks much better than the 10%-14% difference found by techspot).

 

PBO is automated overclocking, a 4.8 manual clock speed is hardly a OC since all SKUs can hit 5ghz easily(90% of them will do it)

Since PBO maxed out the clock speeds on the relevant threads with usage it's a fair comparison.

Lastly, techspot isn't testing multiplayer the same way, nor do you know exactly what they are testing.
They could be running around in a single player match.

The only way to test it perfectly is have 2 players join the same match, one dies and spectates the other to get a mirrored screen on each setup.
Doesn't matter if you don't like the results lol.

 

https://www.amd.com/en/products/cpu/amd-ryzen-7-2700x

Down there, you have specifications. Guess what, PBO behaves within those specifications and does not breach any.
PBO is mechanism of last mile clock vs temperature vs power limit. Nothing more. (From real world experience, it does give 25~50MHz higher clock than XFR(2) alone.)

Calling PBO overclock is same to calling XFR(2) overclock. But both just allow CPU to go above 3.7GHz which is base clock of 2700X and both behave within AMD's official specifications.

 

PBO is an extension to AMD's normal Precision Boost that offers an alternative overclocking solution to manual overclocking by automatically bumping the SoC power budget when possible in order to give Precision Boost more headroom to act more aggressively, thereby achieving higher performance. PBO opportunistically increases the chip power budget by sensing SoC information (such as power, current, voltage, temperature) as well as motherboard subsystem VRM configuration and modifying that info in order to further push the performance.

PBO is capable of raising the following limits:

• Package Power Target (PPT) - the power delivery to SoC
• Thermal Design Current (TDC) - sustained current deliverable from motherboard design
• Electric Design Current (EDC) - peak current deliverable from VRMs during transient

After raising the limits mentioned above, Precision Boost is given more headroom to act more aggressively. Due to Precision Boost granularity, both single-core and multi-core performance benefits. Unlike manual overclocking that might require non-ideal frequencies for all cores in order to get better performance, Precision Boost can dynamically overclock single-thread or a few cores higher than when more cores or all cores are active.

Precision Boost Overdrive - Auto overclocking that invalidates the CPU warranty - not activated by default. Needs activation from BIOS or Ryzen Master and works on some high end motherboards only.

PBO is an overclock, period.

I believe you are mixing up your terms.

Precision boost 2 is enabled by default.

But in the end, it doesn't matter as it is in fact overclocking, which you can read any review saying the same thing.

 

I'll close it here:
i7-8700K Base clock is 3.7GHz; Turbo 4.7GHz => Video All cores 4.8GHz => Overclock
2700X Specifications: Base clock 3.7GHz; "Turbo" 4.3GHz => Video Cores dynamically clock within 2.2 ~ 4.3GHz range => No overclock

Your statements: AMD CPU is overclocked => You lie!
Your statement: 4.8 manual clock speed is hardly a OC => False again, clearly lie!

You are liar! You attempt to deceive people! It is behavior which has no place in this forum. Take it elsewhere.

 

FYI, cheap ass entry level boards have PBO too. As far as PBO limits for current go.
You do not have to change them above 105W. It will still clock within Standard Thermal and power limit.

BTW, I wonder what guys around say about calling someone overclocker upon he/she increases power limit for CPU only

 

Dude, you are dense as a rock.

You can argue all you want, take it to AMD if you don't agree with their own words, quoted by an AMD community leader.

I'll bold it once more since you are blind/dense.

Much like traditional overclocking.....As you can see, Precision Boost Overdrive is sort of a “best of all worlds” approach to overclocking that manual mode usually doesn’t offer.

As much as a hardcore fanboy of AMD, you don't seem to understand that PBO will scale threads that are in use to max clock speeds with the ability to downclock/idle other threads not in use for better performance where it counts.

And to respond to your "You lie" statements.

Your statements: AMD CPU is overclocked => You lie!

Nope. In fact, AMD themselves just proved you wrong.

Your statement: 4.8 manual clock speed is hardly a OC => False again, clearly lie!

Not in the least, considering coffee lake can easily hit 5ghz, with good skus being able to hit 5.3ghz.

I'm closing it here, you can argue with yourself now lol.

 

Much like any insane, you repeat same lie over and over again, expecting us to finally believe you.
(Maybe you got analogy of "like" in your quotation. Maybe not, since you would not suit your lie.)

 

Calling AMDs Robert Hallock a liar.

 

Oh, one of us do not get difference between "like" and "as". I wonder if it is me since you are native speaker. Would be tragic if it was you.

 

Again, the Intel CPUs were manually overclocked while the AMD CPUs weren't. How you call that a "fair comparison" is beyond me (a fair comparison would have been for the Intel CPUs to use MCE). At any rate, the video is put into doubt by results from professional testers, who tested with even higher clocks on the 8700K (5 GHz) but did not come up with a similar difference.

So now you're casting doubt on the results of a legitimate tech site? Yes, let's believe an obscure video on YouTube over professional testers - a video that is barely gotten a thousand views and a total of 13 likes! Frankly, I would take the word of professional testers over a random YouTube video any day of the week.

 

Don't you see it? i7-8700K stock speed for all cores under load is 3,7GHz. When all cores run 4,8GHz permanently, it is not 30% OC. It's actually not OC at all.
When PBO puts one core of 2700X mere 25MHz above 4,3GHz, it is OC. It may be just 0,58% above standard 4,3GHz, but that's OC for sure.

 

There are multiple videos showing similar results using the tried and true spectating method.

Secondly, with zero data on how they test it, where and how, results in fortnite are hugely dependent on area.
Techspot is landing somewhere that has no heavy action to do a benchmark. A place where repeatable results are doable.

I can drop tilted towers with 40 people and drop fps to low 60s
Considering you don't play that game, of course you don't know what you're talking about.

So I could keep posting videos again but you'd just do the whole saying over again since anybody other than techspot is obscured.

Lastly, PBO maximizes clock speeds on cores that are being utilized. AMD says it's a traditional OC but with 'intelligent' abilities(downclock unneeded threads, etc)

So in this sense, setting a manual clock speed wouldn't make a difference anyways since PBO keeps the threads under-load maximized.

But in the sense of comparison, 4.3 is the max the 2700x can do, 4.8 is not the max an 8700k can do.
That's why I say it is fair.

They are both overclocked, PBO is just a smarter and more efficient way to OC compared to MCE.

Now you're just digging deeper holes.

8700k is 4.3 all-core boost, MCE enabled is 4.7 all-core.

 

https://ark.intel.com/products/126684/Intel-Core-i7-8700K-Processor-12M-Cache-up-to-4_70-GHz
I wonder what's wrong with you. Quite a few people have to call your nonsensical statements for you to even acknowledge that 4.8GHz on all cores is OC. And then try to at least deflect with 4.3GHz whatever which was not even part of it. Bring more clocks nobody cares about because they were not in that video.

You have not had argument in this entire thread. Just loads of baseless nonsense and telling others that they are digging their holes.

Only hole I may have been digging would be calling you idiot and you reporting me for it. But did I? Probably not as it is not needed at this point anyway.

Things you are pushing lately. If anyone ever considered your knowledge of any value... they are likely reconsidering that point of view.

At this point, do you think that if you actually managed to make any sound argument, it would connect? No it would not. You lost me completely. (Not that you could have made any sound argument and persuade anyone about that entire AMD is OCed, intel is not.)

 

Once again, you are - rather incredulously - downplaying professional testers in favor of random, individual YouTube videos! Whether it's Techspot, AnandTech or Guru3D, these tests are done by those who benchmark and review products for a living, and yet some video from some unknown person is more authoritative? Jeez, why even read product reviews at all then? (just search YouTube for random videos to base your purchase on!) I find it amazing how you continue to dig this hole.

And for the final time, a "fair test" would have used PBO for AMD and MCE for Intel. How much you can push each chip is not the issue, nor is the efficiency of PBO vs MCE - it's that one test used manual overclocking while the other didn't. Frankly, if this was done by a professional reviewer they would be rightly chewed out by their audience (from both the Intel and AMD fans).

 

Actually fair comparison would be to test both with stock settings BIOS sets up for CPU. So likely no BPO and no MCE. While almost everyone on this forum would enable them, masses which would consider buying those chips would not even know where to begin.
So, fair comparison would be what average Joe gets.

 

I'm not saying youtuber's are any less or more credible than websites that do reviews.

You are downplaying everyone else. If they aren't techspot, anandtech, etc they are not credible; that is asinine thinking considering the results are commonly seen across many reviewers/youtubers.
Are you going to say the same thing when I post more results? Probably.

I get it, you don't like the results.

I've already said it, FPS is heavily dependent on player count and area tested which allows for many testing in many different possible ways.

I'd agree with you on the fair test part, but in this instance it would have made no difference either way in performance.

Ryzen is capped off around 4.2-4.3ghz whereas coffee lake is capped around 5.3ghz.
Besides, AMD markets PBO as a superior, more intelligent way compared to manual overclocking.

We can agree to disagree on fair testing, but you can't pick and choose a single review over the countless other reviews especially considering the fact that there are GPU and CPU bottlenecked areas.

Edit: I just skimmed both techspots main fortnite review and the 2700x vs 8700k review.

Since they reviewed 44gpus in the main one, I can say the definitely landed somewhere that nobody lands at for repeatable results.
They did the same thing for 2700x vs 8700k review.

"For the benchmark test we dropped into the game at the exact same location every time. The test started the moment we hit the ground and we then followed the exact same 60 second path each time and reported the average result from a three-run average."

Running around at one spot with no action is not a real-world scenario.

Not only that, they ran epic settings.
Which makes it mostly GPU constrained.

Nobody plays with epic settings, all competitive players run low with textures on high, or similar.
Low settings puts more stress on CPU

So yeah, my point is reinforced.

Useless review done by techspot.