Intel Stands Behind Controversial Tests That Favored Its CPU Over AMD's

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
No paid company is going to be objective. That is really a red flag. Though intel has been solid performing compared to many amds over the last 10 years. I haven't had any ryzen experience yet especially after a couple years which is where intel usually really shines.
 

No, they're saying that the Intel CPUs were tested with a top-performing, $80 premium third-party cooler, while testing the 2700X with the cooler that comes in the box at no additional cost, probably with a default fan profile. By that logic, they should have tested the Intel CPUs with no cooler at all, since Intel certainly didn't supply an adequate cooler with them. : P The AMD boxed cooler is arguably "adequate" but due to the nature of XFR boost, the 2700X may have managed an extra 100Mhz or so had the same third-party cooler been used, though that's only a small part of what was wrong with their testing. Just looking at their numbers between the 8700K and 2700X compared to existing reviews, it should be obvious that there's some pretty dodgy testing methodology going on there. And since Intel put in place an NDA that restricts reviewers from posting performance numbers for the 9000-series until the CPUs are released, these deceptive numbers that Intel paid for are all potential customers have to go by in the mean time.

And keep in mind, once you figure in the cost of an $80 cooler, and probably a more expensive motherboard, you're looking at paying twice as much for a 9900k and its supporting components. If someone is building a gaming system with no budget limitations whatsoever, perhaps paying a few-hundred dollars more for slightly better performance might be considered worthwhile. Anyone else would probably be better off putting that money toward their graphics card instead, where it would likely make a much larger performance difference in most games.
 
  • Like
Reactions: T0mTh3T0t3m


What is amusing about this is that with the exception of a couple games the crippled half a 2700X chip was still roughly in the same ballpark as the 8700K.
 
OF COURSE Intel stands behind the results...they commissioned them and doing as Intel pretty much always does (more so af of late) THEY choose the settings, how to test, what setting to use on others products to always show themselves "in the best light" they absolutely did NOT need to do this....Let a good product stand on it's own merits (bot the good and the bad)

let the customer/consumer know all the ins and outs of it.

If they (and Nvidia) still did not "clue in" there really is no help for them, and people who keep throwing their money at them (allowing them to continue being this way with no real consequence) are utter fools IMHO.

"stand behind the tests" what foolish putzes.
 
Intel is all lies and deceit. Sure, I can also replicate the rigged tests in my lab and and claim that my results are consistent with Principled Technology...duhh. I also think that the i9 9900K will post better results in single core performance for old-school games in 1080p which nobody paying $500 for a processor will run. But for $150 more than a Ryzen 2700X with no upgrade path once once AMD releases 7nm Zen, is Intel worth it? Heck no! All this smoke and mirrors and lying needs to come to an end. Instead of fixing their act Intel continues to scam their customers. BK already got fired, what else is new with Intel?
 
"Intel Stands Behind Controversial Tests That Favored Its CPU Over AMD's"...and "Tom" can't find the spot covering this criminal act. Intel does not "Stand Behind".. Intel is ignoring its (alleged) crime in its answer, Tom.. Intel is not responsible for the AMD test machine configuration, but Intel cannot claim it was not deliberately done.
 
The worst is that they tapped down the setting to medium at 1080p to maximize the effect of the CPU bottleneck.

And basically, they did that with a 1080 TI. This is a totally bogus Use Case scenario. Nobody will put more than 2000$ of hardware and play at 1080p at medium settings.

First off, if the GPU is anything below a 1080 GTX, this gap just get smaller and smaller as the GPU become once again the bottleneck.

If only this was just a bunch of benchmarks, but no, Intel needed to try spread lies by manipulating numbers. Guess what, the only number that was important for them to manipulate was the price, however they failed miserably on that.

Basically, Intel gaming superiority is only a real thing with a minimum of 500$ GPU at 1080p. this is the only Use Case scenario. At 1440p, the GPU is bottleneck and you will be lucky to get anything close to 5-10%... and we are not even talking about 2160p which doesn't even make a difference.

The 9900k could be having a MSRP of 480$, it doesn't change the fact that it is selling for 580$ which is almost twice the price of the 2700x. At this point, the TR 1950x is almost the CPU that it should be compared to in the price bracket.
 
>that gap has narrowed significantly since Ryzen 1, and at 4K it has all but been eliminated.

It hasn't been eliminated, the bottleneck has been shifted from the CPU to the GPU. This doesn't mean that AMD CPUs are just as fast as Intel's.

If you run a game at 4K where the CPU is the bottleneck like a significant minority of games then Intel is still going to offer higher performance.
 
Just to be clear: Nobody expects the Intel CPU to be sluggish in these tests. The contested part is that its advantage over current CPUs is exaggerated, and possibly by a great margin.

They used a bunch of GTX 1080, one for each CPU.
As far as I can tell they compared eight different systems side by side, not sharing components between them. It's known that for example videocards can have more than 5% variance in performance between cards in the same batch, and here it's not even clear if they used cards of the same SKU.

 
AMD Ryzen 7 2700X and other Zen CPUs is to hard and complicatet to tune, there is not many people who will or like to fiddle with complicating settings in the BIOS, so it's a fair comparison, maybe AMD should write on the box: "Only for professional hardware nerds"
 
I suppose the main objective was to justify the price of the 9900k. Of course it will be faster. The question is whether it makes sense to pay so much for it compared to much more affordable competition. Even from their own processors. That's what matters to Intel lol

In any case I believe that at this point there's plenty of reason not to pay that much for a CPU just for gaming. For most people that is.
 
If you have the chance, check the interview Gamer's Nexus did with the Co-Founder of Principled Technology (on GN's YouTube Channel). It's an uncut, transparent interview.
 


No more hard or more complicated than any Intel CPU. Theres plenty of people who like to fiddle with settings, hence this entire community, and the larger market out there for big coolers so people can push their equipment. But aside from the fact there is no fiddling involved in sticking a Ryzen CPU into a board and having it just work, exactly the same as building any system with an Intel CPU.
 
GN is local to PT, so they went over & the co-founder agreed to sit down & talk it out. i was surprised... the guy seems genuine & open to learning from the criticism. i have serious doubt now that leadership at that company knew of wrongdoing or purposely skewed the testing-
...but...
-it's still entirely possible that after leadership attained the contract from Intel, employees responsible for the actual testing either; do not fully understand testing methodology, or received heavy amounts of "guidance" from an Intel rep. either way, there are a few things in their testing that can be easily argued as mistakes, and definitely a few choices that are very hard to imagine were not deliberately made to hobble the 2700x results.

even the slightest oversight by Intel should have resulted in corrections, yet they chose to publish & promote the obviously skewed results anyway.

https://www.youtube.com/watch?v=qzshhrIj2EY
 


This is true. With the exception of the fact that they used a superior HSF cooler on Intel. Yeah, Wraith isn't a bad cooler, its a good one in fact, but it isn't at the same level as the Noctua used on the Intel CPU.
 


Paid testing, reviews, and studies should all be looked at closer due to the fact the results are being bought many times.

While Principled hasn't admitted to a paid result (just paid testing), they have came forward admitting they goofed... and are retesting. I somehow suspect that Intel will still quote and favor the first set of results.
 
Intel this is all they know. They have screwed AMD so many times in the past its a wonder AMD is still with us. I would expect no less from Intel as this test was flawed from the get go its like they are using a corvette to a old broken down crippled car... Maybe that would be some of the reasons I went with amd this time around, Besides intel still has all those bugs and yeah the updates to fix said bugs did slow down the intel chips. So pay more get less and be Happy!
 


Yeah, in the past I used to wonder if Intel let AMD survive just enough to counter anti-trust claims...though that's certainly not the case these days.
 

Yes it will just work,just as it did just work in this benchmark,the point he tried to make was that you loose way too much performance from things that normal people wouldn't even think about.
Not putting in 4 dimms into a 4 dimm slot system? Wouldn't even think twice about doing that and yet it has a big impact on performance,same goes for the dual ranked ram argument, I wouldn't even think to check if the ram I got is dual ranked or not.

 


Err what? Putting 4 DIMMS in had nothing to do with the memory issue, its that they allowed it to use the default configuration which is ultra conservative, compared to going into the BIOS and setting XMP in the Intel system thereby giving it faster speeds and more aggressive timings. The memory issue had nothing to do with having 4 dimms or dual ranked ram.

Sorry the argument doesn't hold water. Ryzen and Intel both require the same exact tweaks to get the best performance out of the installed processor and memory, just like they always have and always will, the only way to avoid such tweaks on BOTH systems is to not care about memory speed and use baseline 2133 mhz ram.
 
Ah come on guys, you know you want that 14nm+++++++++++++++++++++! Just kidding. I can't reply directly on mobile so to the guy saying people don't want to fiddle with settings and Ryzen is too complicated. You posted the exact same thing on wccf. Ryzen works just as fine as Intel. There is no fiddling if you don't want to but what's the point of building a PC if you don't want to play with it. That's half the fun. I go for performance per dollar and right now that's AMD.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom