Principled Technologies Says it Messed Up Intel 9th Gen Testing

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar


So you're telling me AMD's cooler doesn't allow maximum boost even at stock? Hmm... sounds like AMD made some poor decisions and are wasting money on inadequate coolers and/or falsely advertising boost performance.
 


To be clear here, we aren't talking about game mode disabling hyperthreading on the 2700x. It actually disables cores, to make turn it from an 8 core 16 thread chip into a 4 core 8 thread chip.



For 1-2 cores it will hit and sustain max boost with the stock cooler. But anything more demanding than that and it starts to produce more heat and throttle down. It's still boosting above the base clock but still 200-300mhz less than the max xfr boost clock. And to be fair, AMD advertises XFR as going hand in hand with better aftermarket cooling solutions. So essentially, PT decided to use the stock cooler, which means no XFR at all, only precision boost.
https://www.amd.com/en/technologies/sense-mi

XFR2 allows the clocks to boost as many cores as possible regardless of workload as long as the temps are under 60C. So an aftermarket cooler is necessary. The temps on the stock cooler are too high.

Still 10x better than the Intel Cooler, but nowhere near as good as a higher end Noctua.

 
for the cooler bit, boxed one from AMD is "pretty good" for what it is, but, many aftermarket ones are SIGNIFICANTLY better. the boxed one from AMD is able to keep the cpu cool enough, but, will not allow it to be "the best it can be" either, whereas that Noctua one certainly is capable of letting it be nearly as best it can be via air cooling alone.

design of XFR and such AMD did for Ryzen is very intelligent and if provided with "enough cooling" will scale the clock speeds as best it is able to do (will not throttle because of poor cooling)

if Intel did NOT include a boxed cooler, so be it, but test BOTH using the SAME cooler (or comparable) to be as "unbiased" as possible, not use one that is "only" capable of stock TDP and slap a different much higher end cooler on the other and call it "good enough"..lazy mofos

this whole situation reeks of Intel paying off a 3rd party to make themselves look awesome and AMD look cruddy (do not expect anything less of Intel or Ngreedia to be honest)

for the memory testing as well, XMP is an Intel thing, not an AMD thing (A-XMP or DOCP sometimes is, but often enough sets timing much looser which impacts performance vs manual setting it) tester needs to manually set things properly to show them "as they should be" vs showing "worst case" amazing how everything is or seems to be done to show Intel at it's best (if not cherry picked) and also subsequently putting an anvil on AMD.

agree on the Game Mode though, if AMD "knew" this causes performance problems with Ryzen chips and was meant for Threadripper or EPYC they should have done an Agesa update that disables it so users will always get the best performance possible vs a potential crutch of performance loss.

weird called game mode when enabling it hurts game performance LOL.

also as far as using the 2666 ram or whatever, Intel has generally scaled much better at lower speeds and is able to use much higher speeds as well because they have prioritized the way their memory controller works, I would not be at all shocked if they chose this EXACT speed because it absolutely benefits the Intel chips but is on a "bad" memory divider for the Ryzen systems.

anything Intel can do to "get ahead" they will do...if they REALLY wanted to show how "good" their product was, they would have given it to 3rd party "verified" unbiased review sites to show the merits this way here potential buyers are not being "tricked" into their BS games....show the product as it is (warts and all) at least then they can proudly stand up and say "we are not trying to hide anything"

I quite wish this "industry standard" BS the facts to make sales would be buried to never be seen again, but nah, we got Intel pulling crud like this and Nvidia being anything but a corporation that does everything they possibly can to BS things, to use software/hardware "tricks" to make sales instead of making the best possible product they can (multi-multi billion $ companies might as well be selling knock off brand t-shirts from back of van style as of late..is pathetic they have to resort to being this way)
 
if Intel is the one that "paid them" to do it, then Intel should be "proud enough" to stand behind their product and not enforce "parlour tricks" to capitalize performance...hell if it were ME, I would say "you know what, my product is that awesome, you will use this awesome liquid cooler for mine and theirs to show how amazing we are compared to them"

more costly, yes, but, real world conditions from enthusiast standpoint no matter what you throw at it speaks volumes of why you can call yourself a premium solution justifying a premium price (hell I would go one step further and tell the reviewer to get 2 or 3 "off the shelf" so there is very little chance of "cherry picked" chips messing with average consumer expected results)
 
I think a fair comparison is Intel with no cooler and AMD with the stock cooler. I mean all these people are saying it's AMDs fault for providing a baseline heatsink. I think it's Intel's fault that the chip melted because they didn't provide one at all. That's fair.

Jesus Christ people. How far up Intel's ass are you? I'll just go ahead and say you'd be fine with liquid nitrogen cooling which would essentially permanently run the Intel chip in boosted mode against the AMD chip with, and I repeat, A BASELINE COOLING SOLUTION, as in you know, BARE MINIMUM COOLING. I mean, that's fair right?

Does anyone around here know that you should keep test conditions as close as possible for valid results? No? Then you should stfu.
 
Guys I see this review as a very positive one for AMD. What it clearly states is that a Ryzen 5 2500X(4 cores 8 threads) with stock cooler will deliver ~85% of the performance of an Intel Core i9-9900k + Expensive Noctua Cooler for less than half the price. hahahaahahaha OMG!!!!!

Thats terribly depressing for Intel.
 
The packaging for the new Intel Core i9 reminds me of Windows Vista which had better graphics than WinXP but the functionality was sooooooooo much worseeeeeee!!!!

Just because it's in a fancy box, it ain't gonna sell... when the performance is only a little higher than a R5 2500X with a stock cooler ...lol
 


Well then. Why don't we have them test the Intel offerings with what they come with (or not) just to be fair and equal? If it's in the box, it's good. If it's not included, well... good luck with that.

/s
 


Yeah but that's the problem, the cooler that comes with the Ryzen 2700x is not advertised as "BARE MINIMUM COOLING". How far up AMD's butt are you?

If the stock cooler that is included with it is labeled as "the ultimate cooling solution", I am going to assume it at least supports the stock clock/boost speeds. Changing the coolers should have no impact as long as they are good enough to support the stock boost thermals.

If AMD's stock cooler cannot, then well maybe they shouldn't say they included "the ultimate cooling solution" with it 😛. Maybe AMD can put out a notice to all retailers "please cross out 'ultimate cooling solution' on ryzen 2700x box and replace with "this cooler blows (Well actually, it doesn't!) and we don't even trust it enough to run stock speeds in benchmarks, just throw it out and buy a different one. Happy gaming. P.S. when you're gaming, disable gaming mode."

Joking aside, I don't have a dog in the race, I like the competition, but aside from not enabling XMP on the AMD side, I don't really harbor any ill feelings towards the benchmarks. If I saw "game mode" as a cpu setting, I'd automatically assume it was for, well, gaming.
 


If you can't take heed of the numerous flaws in the testing methodologies and well documented inconsistencies by some of the more respected tech YT folks and are willing to accept the "findings" of this test given the aforementioned, you're not seeing the forest for the trees. Seriously.

Go watch Steve's video on Gamer's Nexus. He makes a point-by-point case of all the flaws in this test which clearly skewed the outcome heavily in Intel's favor.
 


If I were at AMD I'd probably do just that. Then I could say, see. Intel is overpriced and doesn't even come with cooling.
 


And perhaps you're right. And that being the case, I'm glad that they are retesting it. Nothing wrong with more information.
 


And he should be doing just that. We need independent reviewers who can give us the real results.

What if the i9-9900K had the same Wraith cooler? Would the results have been any different?
 
I could be wrong, but I don't think a smaller cooler would make a difference for Intel. They've got solder TIM now, and even without it I know I tested an i7-8700K at stock with a Cryorig H7 and it never even hinted at thermal throttling. Which is why it doesn't really matter to me. Intel has higher thermal limits. AMD has lower limits. AMD suffers. That's the way the cookie crumbles.
 
As always
Intel is full of shit, dishonest, doing market manipulation
and cheating customers by fake CPU generations
9th generation(same shit as 5th, 6th, 7th 8th generation): Or 999999th generation
 
Have seen the interview with pt along games nexus. The co founder did his best to beat around the bush. But in true honesty i believe him on certain issues. No co founder wants his firm to go down like this. Btw PT is as old old firm. I kinda knew of them long before the the pentium 4 was developed. They r good people.

There results however r not truly correct as amd systems can and DO deliver more.

Its just they now have to get their house in order and respond properly, else their creditability will never EVER be restored.
 
If this review was legit to begin with they would just used whatever was in the box, or not in the box, and done a 1:1 compare. But this company did not and should have lost all the trust they built up over the years due to being BOUGHT by Intel to fudge these results.

Also, Game mode is for Threadripper/EPYC due to multi-numa nodes being present on those CPUs. Ryzen based Am4 CPUs only have a single NUMA and game mode does NOTHING for those CPUs. The fact this poor excuse of a benchmark team enabled game mode and used it to skew their test results should just tell everyone how far down Intel has fallen since AMD got their shit together. I can't wait to see what bullshit piss poor companies like this use to debunk AMD at 7nm. Since 7nm is going to be a 13.5% IPC jump and 800mhz above 12nm its going to be a fun shit show to watch.
 
This is nothing at all, compared to the designs with the spec violation being the INTENDED designs, no bug, according to the ex-CEO!!!
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom