Intel Core i9-9900k 8c/16t, i7-9700K 8c/8t, i7-9600k 6c/6t 2nd Gen Coffee Lake CPU's + Z390

Principled Technologies Responds to Hardware Unboxed
Hardware Unboxed
Published on Oct 10, 2018

Spoiler: Click to See Video

Intel's i9 9900K vs Ryzen 2700X Gaming Benchmarks Are Misleading, Period
Jason Evangelho, Oct 9, 2018, 04:23am
https://www.forbes.com/sites/jasone...700x-gaming-benchmarks-are-misleading-period/

"A meaty set of competitive benchmarks for Intel's newly announced 9th Gen Core i9-9900K are in, pitting the powerful mainstream gaming CPU against AMD's Ryzen 2700X. There are plenty of them to chew through. Nineteen in fact, comprising many of the most popular PC games out there. Unfortunately, they're all bogus."

https://twitter.com/VideoCardz/status/1049735983661494275
https://twitter.com/VideoCardz/status/1049732385745641472

 

The video by AdoredTV, posted by hmscott dealt with this. It even mentioned that AMD has bought publicity services from them. That still means NOTHING on whether the test results were conducted fairly or skewed to Intel's benefit, nor the history of nerfed benchmarks to focus less on how consumers use chips while including tests that would benefit Intel, nor the clear problems with their testing, which even they published an apology for.

First, you need to consider all 16GB DIMMs are dual rank, not single rank, which puts extra stress on the IMC, making it harder to clock the memory higher and to tighten timings. There is zero reason to say this is normal. Then, filling two DIMMs per channel further puts stress on the IMC and can make compatibility harder to achieve applying timings to the ram, hence being almost the worst situation you could put AMD's Ryzen in. Putting on DOCP then changing clocks does theoretically mean that it was running the XMP timings, which are Intel optimized timings, while downclocked to the speeds for the chips running at stock, 2933 on Ryzen and 2666 on Intel. Putting aside the effects of sub-timings or the fact Intel's timings generally are looser than AMD's for XMP (compared to lines like Flare X), it still seems to have been unrealistic.

Nowadays, a solid testing would be setting both up like this, but with 16GB of ram in dual channel (2x8GB single rank DIMMs), or 32GB dual channel single rank max (still do not like it, nor has any testing I've ever seen shown any ram bottleneck for gaming if you have 16GB of ram). Now, since no current games use more than 16GB, it would have been fine to do only 16GB of ram. But willfully doing 64GBs when the state of gaming is known, memory prices are known, and it being well known that dual rank DIMMs perform worse on Ryzen platform overall, while doing it two DIMMs per channel, you know you are sandbagging intentionally on memory. This is bad, but not as bad as the game mode or the cooler.

Meanwhile, buying secondhand does not benefit Intel as a new sale. It benefits them on the purchase of their chipset, but not for the higher margin CPU. So, buy for your needs. If that means supporting AMD, so be it. If that means getting a second hand Intel chip, so be it. With the lower price you are paying, and depending on your needs, it may be that the used 8700K fits right in with your needs. Do what is right for you and what your needs are.

Don't be so obtuse. When Intel was selling their 8-core chips for $1K when the 1800X was released, Intel had nothing in the market to compete with the AMD 8-core chips, and that was priced at HALF of what you could get from Intel. You cannot just point to a price without consideration of the market as a whole. Intel responded by putting the 10 core to $1K, cutting PCIe lanes on their 8 and 6 core chips, and putting the planned 12-core at $1200. They then reacted to AMD's news of entering the HEDT market by announcing the 14-18 core skus, basically Xeons unlocked, but that would come in like 4 months or 5 months after the computex announcement with only the 10-core on sale immediately and the 12-core dropping in like August, if I recall correctly.

Then AMD released their HEDT platform and decimated prices. Then, after release, further reduced prices over time.

So, you can technically compare it to that, if being disingenuous. Instead, if you look at the market as it stands, not the pricing in isolation, compared to a CPU that at the time undercut the market drastically in order to justify today's pricing, then you are being intentionally obtuse.

Market isn't everything and as we have seen, the market has sold off AMD on good news for AMD. It isn't an efficient market, and if you think it is, you haven't studied the emerging literature on the topic. I suggest you do research on that before pointing to something that does not have more than a correlation with corporation's performance, which is beyond simply CPU performance.

You are correct that AMD has not really made inroads into the high end laptop market yet.

But I would caution and recommend a deeper analysis of capitalism, which involves incentives. Until AMD gets more mindshare with consumers, consumers will, rightly or wrongly for their needs, purchase based on what they think will be proper for them. This means AMD will not get the sales in that segment, which creates an economic incentive to put that money elsewhere. This was seen with gaming GPUs. They kept pouring money into trying to compete at the high end. They also were mismanaging their IP which sat on shelves, etc. Su came in and took that IP off the shelves, moved the money to compete with high end GPUs over to CPUs, used revolutionary tech to introduce multi-die chips that did better than when Intel glued two cores together and called it a dual core CPU, and saved the company from a 2018 bankruptcy. The point of that story is that you follow the incentives (and capitalism creates a LOT of perverse incentives, including Intel's long history of anti-competitive, anti-consumer behavior, with this testing being the most recent stick on that pile).

 

https://www.principledtechnologies....ng_PC_gaming_processor_study_interim_1018.pdf

In no way does this company sound like an Intel front or shill company. In fact they seem genuine and legitimately interested in finding real results. Sounds like they are correcting some errors or running things in a different manner to appease some. In the ending the results will likely change a bit, but we will still see the 9900K being the best CPU for gaming and multitasking at the mainstream level. Still expensive, but still the best for now.

Also I think someone said Intel was charging $600 for their 6 core processor when Ryzen launched. I have doubts, I paid (looking at my newegg invoice) $389.99 for my 6 core 5820K in March 2015. Microcenter was charging far less for the same chip. I feel $499.99 for my 9900K with 8 cores in 2018 is fully justified. AMD is cheaper sure, but you get less performance period.

Edit:

Just as I suspected the 6800K 6 core Broadwell-E hovered around $400-420 for almost all of March 2017 which was launch month of the Ryzen CPUs.

https://pcpartpicker.com/product/Td98TW/intel-cpu-bx80671i76800k?history_days=730

 

I think AMD is smart to stay out of the laptop business anyway. The direction things are headed (poorly engineered, anemic, anorexic, BGA, Max-Q dungbook throttleboxes) shows the whole idea of "high performance" laptops is gradually dying. These turds that are circling the drain have become little more than glorified bulky Windoze OS X smartphones. In some cases, they might even be an insult to smartphones. I think it would be stupid on their part to get sucked into the vortex of notebook filth along with Intel and NVIDIA at this point.

There is so much ludicrous marketing noise at this point that a great deal of the "information" available is either worthless, or it is biased because computing is becoming far too political, with all sorts of pundits and spin-meisters offering their two cents and sometimes getting uppity whenever they encounter dissent. It was kind of comical for a while, but it's starting to get old. The ones that stand to lose the most in all of this crap are the ignorant consumers that don't have enough knowledge or experience to make a good decision. So, they turn to their idiot friends on Facebook to help them decide what is best.

I agree with you. Based partially on the interview in the GamersNexus video, I'm leaning toward incompetent mingled with ignorance rather than bias. If they were not incompetent it wouldn't have the appearance of bias because they'd be smart enough to conceal it well enough to not get caught so easily with their pants down. Let's face it... some people (including self-proclaimed professionals) suck at benching. They should stick to what they are good at, but not all of them know they suck at it. I stand by my belief, which I express every time something shiny and new surfaces in the tech world. The "professional" reviews (about 90% of them) and benchmarks are inaccurate, flawed or biased. Early adoption based on that information isn't very smart and I won't believe most of what I see, hear or read about performance unless it originates from someone that shares my interests, skills, opinions and passions.

Then, we have those poor ignorant souls that are so confused they want to scream. Maybe they don't have any friends on Facebook with enough baseless bias to help them decide what is best.

 

They (Intel the sponsor and curator of the tests + patsy) carefully picked the mis-tweaks that most negatively affected AMD 2700x performance - everything that Intel thought they could get away with, "shucks, we didn't know, they did it - not us!", and most Intel fanboi's would want to believe that to be true and go with it.

Intel's most offensive offense is publishing skewed benchmarks 11 days before independent benchmarks are allowed to be published for comparison. Making Intel's BS benchmarks the only data for deciding whether to Pre-Order.

Intel's behavior is an extremely egregious example of why it's never a good idea to Pre-Order based on vendor claims. Nvidia was just as bad, but at least with Nvidia it was a lack in intelligible data accompanied by obviously obfuscated claims.

Putting all of the mis-tweaks in the testing documentation is the cleverist stroke of genius. Documenting the skewing right there in the test plan makes it look innocent. Hiding BS in plain sight is Intel's specialty - don't forget to look under the desk for the Phase Change Cooler, "ohhhh, we forgot to mention it, sorry!", Pfffft!!

Game Mode is known to have been done for ThreadRipper to shut down cores / CCX's, same happens on Ryzen 2700x, this is the most damaging.

Memory speed and configuration, easy to blame on innocent incompetence.

Cooling, not using the same cooling for both CPU's to even the comparison.

I don't recall seeing Intel Multi-core Enhancement being mentioned... I'd check for that too, as soon as independent reviewers can publish their findings.

Game settings as required to be selected for testing and not explicitly listed, obviously tuned to benefit Intel, with embargo on 9900k benchmark publication there is nothing that can be given in response - until at least Oct 19th.

And more, and more, given the expertise of the tests creator - Intel - and their well instructed henchmen, there is no benefit of the doubt that should be given, as Intel has already counted on taking advantage of their fanboi's gullibility.

With Intel's well documented history of playing dirty pool, getting caught with huge fines assessed, and legally binding requirements to include consumer protective disclaimer text in their benchmark claims - forever...how can we trust Intel?

Intel can't seem to help themselves - is their sneaky low down cheating behavior ingrained in their corporate DNA? Is this cheating behavior something that is simply never going away?, can we never trust Intel?

Or, will Intel stop their unethical cheater corporate culture? Will it take a top down expulsion of all of the cheating cheaters making Intel's culture so untrustworthy? Will it take government intervention? Another FTC slap down?

How does Intel stop looking like cheating cheaters, that can't stop cheating?

Intel has been more reliably able to deliver on their dirty tricks than actually following through and delivering 10nm through to high volume production, which at this point also feels like it is going to take forever.

The crazy thing is that a fair even playing field test of the stock 9900k against the stock 2700x would have shown a smaller difference - but more than enough for their fanboi's to self-justify paying 2x as much for the 9900k.

None of Intel's shenanigans have ever been necessary. It seems like at every opportunity Intel snatches defeat from the jaws of victory. Intel are simply fooling themselves, and looking like fools while doing it.

There is nothing innocent about Intel. I don't trust Intel nor their well paid henchmen.

Core i9-9900K, benchmarking controversy, adios mainstream hyperthreading? | The Full Nerd Ep. 71
PCWorld
Streamed live 12 hours ago
Join The Full Nerd gang as they talk about the latest PC hardware topics. Today's show is all about Intel's Core i9-9900K, the news, the benchmarking controversy, and is this the end of mainstream hyperthreading?

 

It is a smart move by Intel to reduce the number of threads in favor of cores. AMD's forte has been its multi-threaded performance. If Intel would try and compete by increasing the number of threads its mainstream processors can handle, it will further stimulate the development of software dependent on many threads, thereby boosting the performance of Ryzen. Think about it, game developers won't suddenly use 16 threads as their target when only the Intel flagship processor can process them. They'll still aim for a maximum of 8 threads (= 7700K, but also 9700K with its 'pure' eight cores).

 

It's an interesting thought, but more likely Intel dropped i7 hyperthreading because they had to in order to differentiate the i9-9900k so they can charge a much larger premium.

A 9700k with 8c/16t selling next to the 9900k 8c/16t wouldn't allow both to coexist in the same lineup, especially if Intel wants to charge $100+ premium for the 9900k. Both would likely OC similarly, within a small margin of difference, too small to charge more for the 9900k.

It would make no sense to have 2 sku's with 8c/16t in the same line up.

Given that simple fact, if the 9600k had 6c/12t like the 8700k 6c/12t, then there would again be 2 sku's with the same core / thread count, and just like the 8700k will likely beat the 9700k 8c/8t in some tests, so the 9600k 6c/12t would compete directly with the 9700k 8c/8t.

The smart move would have been to not introduce the i9-9900k sku at all, and release the 9700k with 8c/16t, 9600k 6c/12t, etc - a normal sku by sku upgrade path from 8th gen to 9th gen desktop CPU's.

So, no it's not a smart move on Intel's part to have completely screwed up the 9th generation core / thread positioning, going back on a decade of i7 hyperthreading co-existence, and instead introducing an "i9" that is really the i7 expected.

The only way it would be a smart move is if Intel simply can't build 8c/16t CPU's in high quantity, but can build 8c/8t dies successfully in large quantities. Given those production limitations Intel would need to put out these odd ball core / thread sku's.

Rather than a smart move, Intel might have been forced to implement this eclectic survival tactic. Smoke and mirrors once again from Intel instead of coming clean on the situation?

Intel has likely shot themselves in the foot again, and have taken the shine off of the 9700k / 9600k, making both seem "broken" or "cut down" in comparison to the 9900k and the previous i7 generation hyperthreaded CPU's.

 

Leo asks WHAT is INTEL WORRIED ABOUT ?
KitGuruTech
Published on Oct 11, 2018
Leo has a very important question to ask everyone today. 'What is Intel worried about?.'


What Reviewers Think of Intel's Launch, ft. 7 Reviewers
Gamers Nexus
Published on Oct 11, 2018
We spoke with 7 reviewers at the Intel 9900K & 9980XE launch event to ask what they thought of Intel's new CPUs -- good or bad, all commentary was on the table.

Spoiler: Click to See Video

Ad: Buy EVGA's RTX 2080 XC Ultra (Amazon: http://geni.us/Wk5e) or 2080 Ti XC Ultra (Newegg: http://geni.us/xKfae)

In order of appearance, this video features Bitwit (Kyle), Paul's Hardware, Joker Productions, PCWorld (Gordon), Techgage (Rob), Hot Hardware (Marco), and Tech Report (Jeff). Their respective websites and outlets can be found at the links below:

Bitwit / YouTube: https://www.youtube.com/channel/UCftc...
Paul's Hardware / YouTube: https://www.youtube.com/channel/UCvWW...
Joker Productions / YouTube: https://www.youtube.com/user/jerzybak...
PCWorld / YouTube: https://www.youtube.com/user/PCWorldV...
Techgage / Website: https://techgage.com/
Hot Hardware / Website: https://hothardware.com/
Tech Report / Website: https://techreport.com/

 

9900k has had its shine taken off with the "independent study" marketing shenanigans. Its exposed attempt at "arms length" third party verification ... where the entire length of the arm is buried up unPrincipled Technology's backside moving the "independent" puppet's mouth

 

If you think about it, it really doesn't matter, it doesn't affect the performance of the 9900K once you actually own it - it is still gonna be the best gaming CPU, and probably for a lot of other applications too - you can't argue with 8 hyperthreaded super fast overclocked cores! If you don't like the way Intel might be conducting their business, if you think the whole Principled Technologies thing is a wicked scam then people can vote with their wallet even if they lose out on not owning potentially the best CPU for their needs. I'm kinda agnostic towards it all, if you know your stuff & research it then you're not affected - you get to choose the best products for your needs without the biased marketing of Intel & AMD, which lets face it...happens.

 

I hope those of you who are morally grandstanding about this will indeed put your money where your mouths/keyboards are spend it elsewhere on future hardware purchases.

 

Supporting critical investigative journalism into questionable practices of tech giants isn't morally grandstanding. You don't forfeit your right to buy from a company if you criticize certain aspects of that company.

 

That's perfectly fine, but as soon as you (the general/collective) start telling people to vote with their wallets, I expect them to abide by that same standard.

 

Well, I've got a 1950X I've owned over a year, so if I'm included, there you are. Would I talk about Intel being right for someone or not? Of course. Does the little bit faster justify the larger expense? Generally, no. There are tonnes, but not always. Serious gamers are looking at 1440p more often and the frame rate at medium 1080p settings is not realistic for those buying on only FPS in a couple titles without understanding holistic value. Especially when sales inflated process so much that you could nearly buy a 7700K build AND either a 1700 or 1700X or 2600X build for capture card use for what a new 9900K costs with a good capture card. Well you peg the quad core from Intel in demanding games? Yes. But you also have way more power than any one chip.

Would I have them look at BW-E X99 platform and the HEDT platform for AMD over the 9900K? Hell yes, due to the value proposition. I even tools bennyg to not boycott a used 8700K due to his morals as the second hand purchase doesn't benefit the company he cares to act against on moral grounds.

I just think that the 9900K is bad value and people would be better served with either the 2700X, the 8700K, or the 9700K, generally. Leave the flagship, especially at inflated prices, to professional gamers, to overclockers, etc. For anyone else, the value isn't there and you are lighting your money on fire.

I've explained before why 10nm Intel will suck and if able to wait for 7nm from TSMC, placed at around 2.5x as dense as Intel's 14nm, considering current Intel pricing and expected availability placed around 6-8 months out for AMD, one should wait. If you want to burn money or have a need or hobby, then sure, but I still recommend reflection before then. Anyone that has a problem with me telling people such a thing should stand back and look at what I'm trying to say.

Sent from my SM-G900P using Tapatalk

 

I'm salty about it precisely BECAUSE I have no other option being a Clevo LGA owner

I do not like supporting, indirectly by purchase, companies that choose to be dishonest and manipulative. Yes, the 9900k will be the best gaming CPU. Which just means Intel didn't need to do this with their performance benchmarking. Their concern has to be the value proposition compared to the cheaper 2700x, and misrepresenting that by their paid "study" is an intentional strategy to deceive the consumer.

 

Here's likely the first article out in response about this Intel / Principled Technologies mess after PCGamesN published them (original article missing...):

Commissioning Misleading Core i9-9900K Benchmarks
Botched Results, That's What You Get
By Steven Walton on October 9, 2018
https://www.techspot.com/article/1722-misleading-core-i9-9900k-benchmarks

Comments

 

Principled Technologies Issues Updated Test Results
by Tomshardware.com | Paul Alcorn October 12, 2018 at 7:45 PM

Principled Technologies issued an updated game testing report to respond to the benchmarking controversy surrounding its recent gaming benchmarking report. The original report, which was paid for by Intel, pitted AMD's Ryzen processors against the Intel's new lineup of Coffee Lake Refresh processors. A closer examination of the results revealed a few test conditions that obviously could skew the results in favor of Intel's processor, including using a less-capable CPU cooler on AMD's chip and disabling half the cores on an AMD Ryzen processor.


Principled Technologies issued a statement about the new reports, stating that it had retested all AMD processors in Creator Mode. As expected, Creator Mode "yielded the best gaming performance on the 2700X." The Threadripper processors provided the best performance in the Game Mode testing used in the original report. Hit the link above to read the full statement.

---------------------------------------------

Enthusiasts also raised the alarm about the cooling solution. As we know, less-capable coolers can impact performance, but the company stuck with the stock AMD cooler yet again in the retests, leaving the potential issue unresolved. A beefier cooler on the AMD processor could afford it an advantage by allowing it to take full advantage of its XFR boost frequencies. As we've proven in the past, improved cooling benefits both AMD and Intel's chips by allowing the processors to operate at their Boost frequencies more frequently, and then maintain the heightened clock speeds for longer periods of time. The company should have retested with a more capable cooling solution, or provided test data to prove the cooler didn't impact the test results. We're reaching out to Principled Technologies on its decision to stick with the stock cooling solution.

In short... This means the Ryzen R2700x can't take full advantage of its XFR boost frequencies even at stock clocks with the included cooler. Aka a useless free feature thrown out by AMD. Wort Nothing! What a shame. Why not forget skip the cooler in the package and instead lower the Cpu prices equal what the cooler cost AMD? On top they could save money on decrease the glossy package (smaller box = cheaper) + lower the shipping cost.

 

I read the pdf that you linked with the updated tests of the 2700X in Creator Mode vs Game Mode. Unsuprisingly the 2700X performed better in pretty much all games when in Creator Mode. One thing the testing company didn't do, they didn't provide a conclusion/summary showing the overall gap between 2700X vs 9900K when 2700X was in Creators Mode - they showed the performance gap on each game, but they didn't provide an average difference over all games. So it doesn't clearly highlight how much of a boo-boo they made by not running Creators Mode during the first round of their tests, unless you go through each game and then calculate your own average difference of all games. Loosely, I remember seeing where previous differences were sometimes 50% between 2700X and 9900K, well now with Creators Mode those differences seem to have decreased to say 20% - I can't be bothered to go through all their games to calculate the overall average now with Creators Mode, but they should have done that themselves, but I guess they didn't want to highlight how much of a mistake they made by running the 2700X in Game Mode previously.

 

No need for that. The number will be published by the reviewers Be you sure

 

Ha, yes, of course! We'll keep an eye out!

 

Re the re-released PT report, where they corrected the main flaw and the performance deficits are much reduced in certain games
They also wasted their time testing the threadrippers in Creator Mode as if that was going to show anything that wasn't already known.

Deceptive Benchmarks Corrected, Intel Still Doesn't Get It

Gamers Nexus
Published on 12 Oct 2018
Principled Technologies has corrected testing that Intel published, previously showing Ryzen 7 in an unfair light.
This follows a growing saga of Intel bending the truth about its i9-9900K CPU, to a point which is completely nonsensical when considering its existing gaming lead.

 

I thought Steve might summarise the Creator Mode testing by giving an overall average difference (which I was talking about in my previous post), but unfortunately he didn't do that. No new info in that video, but interesting commentary on Intel, etc.

 

The best recommendation(s) I've seen are to ignore the Intel paid benchmarks - "fixed redo" or not - and wait for independent benchmark reports before deciding how best to invest your funds.

Like every other new release - skip the Pre-Orders altogether. Most early sales are going to be price gouging anyway, so it's better to wait for purchase until the prices settle at MSRP.

Intel has built up the illusion that they "just can't make enough", so there will be "shortages" - I'd skip that malarkey altogether and when Intel is playing straight again - products are reliably in stock long enough to purchase at MSRP, wait for a sale.

 

New PT Data: i9-9900K is 66% Pricier While Being Just 12% Faster than 2700X at Gaming
by btarunr Friday, October 12th 2018 22:47
https://www.reddit.com/r/Amd/comments/9nsfql/new_pt_data_i99900k_is_66_pricier_while_being/

Deceptive Benchmarks Corrected, Intel Still Doesn't Get It
https://www.reddit.com/r/Amd/comments/9nr4hh/deceptive_benchmarks_corrected_intel_still_doesnt/

WayeeCool 57 points 4 hours ago*
"LOL. It went from a 40% - 60% advantage, to the 9900K being only 10% to 15% faster than the 2700X in most modern games..."

At this point this "checking on Intel" BS is probably eating into their real 9th generation testing time, which is more important to do well.

These guys have their own schedule of tests and history of comparisons to bring together, without redirecting focus and effort on "Intel and Principled Technologies Bogus Journey".

I hope all of the reviewers stop letting Intel influence the direction of their testing time.

No, THIS is the Fastest Gaming PC Possible! - October 2018 Builds
$3000 build is the 9900k Build
Paul's Hardware
Published on Oct 12, 2018
No, THIS is the Fastest Gaming PC Possible! - October 2018 Builds

Spoiler: Click to See Video + Build List / Links

2:26 ► $1000 8-Core Gaming PC (GTX1070) - http://bit.ly/2A86E0C
5:24 ► $800 8-Core Gaming PC (RX580) - http://bit.ly/2PwaB4o
7:56 ► $3000 Fastest Gaming PC Possible - http://bit.ly/2NF1pZw

►► VOTE FOR NEXT MONTH
https://www.strawpoll.me/16627518

My "Builds" Playlist: https://www.youtube.com/playlist?list...

2:26 ► $1000 8-Core Gaming PC (GTX1070) - http://bit.ly/2A86E0C
AMD Ryzen 7 1700 3GHz 8-Core Processor - https://amzn.to/2CFQMos
Asus STRIX B350-F GAMING ATX AM4 Motherboard - https://amzn.to/2CH22AK
G.Skill Ripjaws V 16GB (2 x 8GB) DDR4-3200 Memory - https://amzn.to/2pQX6Bd
SanDisk Ultra 3D 500GB 2.5" Solid State Drive - https://amzn.to/2CHl3Do
Gigabyte GeForce GTX 1070 8GB Mini ITX OC Video Card - https://amzn.to/2pSld2k
NZXT H500 (Black) ATX Mid Tower Case - https://amzn.to/2pO2a9i
Corsair CX (2017) 550W 80+ Bronze Power Supply - https://amzn.to/2CJyH8Q

5:24 ► $800 8-Core Gaming PC (RX580) - http://bit.ly/2PwaB4o
Same parts as the $1000 system, with these swapped for GPU/RAM/Motherboard:
ASRock X370 Pro4 ATX AM4 Motherboard - https://amzn.to/2QMBMIj
Radeon RX 580 8GB Video Card - https://amzn.to/2pQWdso
ADATA XPG GAMMIX D10 8GB (2 x 4GB) DDR4-3000 Memory - https://amzn.to/2AGA6x1

7:56 ► $3000 Fastest Gaming PC Possible - http://bit.ly/2NF1pZw
Intel Core i9-9900K 8-Core CPU ($530 pre-order at Amazon) - https://amzn.to/2pSmlmA
ASRock Z390 Taichi Ultimate ATX LGA1151 Motherboard - https://amzn.to/2pRSYAW
G.Skill Trident Z RGB 16GB (2 x 8GB) DDR4-3600 Memory - http://bit.ly/2On33nR
ADATA ULTIMATE SU650 960GB 2.5" Solid State Drive - https://amzn.to/2wWoCQT
Asus GeForce RTX 2080 Ti 11GB ROG Strix Gaming OC Video Card - https://www.asus.com/Graphics-Cards/R...
Lian Li PC-O11 Dynamic White Tempered Glass Case - http://bit.ly/2OtpSCD
EVGA SuperNOVA G3 650W 80+ Gold Modular ATX PSU - https://amzn.to/2pQugkp

Build A Great Gaming Rig - For The Cost Of A 9900k
GearedInc
Published on Oct 13, 2018
For the cost of the new i9 9900k from Intel you can build a great entry level gaming rig capable of playing anything at 1080p.

Spoiler: Click to See Video + Build List / Links

Parts List - https://pcpartpicker.com/list/pLdmWD Total $584

Amazon Affiliate Links - Some options/prices are different
Ryzen 2600 https://amzn.to/2Cdf4Fd
120GB SSD https://amzn.to/2RL8yuE
1TB WD BLCK https://amzn.to/2NEdkXu
550 Watt PSU https://amzn.to/2EiWA97
Case https://amzn.to/2CfMVh0
Motherboard https://amzn.to/2EgUtm1
RX 570 https://amzn.to/2CbairU

Patreon https://www.patreon.com/GeardInc Support me directly through Patreon.

 

I plugged the results into a spreadsheet
and played a little bit to produce one result per game
then averaged the % differences

9900K over 2700X Game Mode = ~24%
9900K over 2700X Creator Mode = ~16%

i.e. the gap between the two reduced by about 1/3rd

Cherrypicking the CS:GO result means Intel could still claim "up to 50%"



I'd upload the spreadsheet but this forum rejects .xlsx file extension

 

Upload it as a *.txt and tell people to change the extension to *.xlsx when they download it.

Charles

 

Didn't work as a txt so here it is as a zip

 

Intel is trying to get the Ryzen platform looked at as reincarnation of the Bull Dozer line of processors. If they pull this off then the market would look at 7nm as a futile effort to catch up in performance to their own cores. it would be as if 10nm was not even needed to stay competitive. With all the announcements and open letters from Intel it has successfully taken the wind out of AMD stock prices.

Because AMD is so quiet on the 7nm front. This both for Epyc and Ryzen there is no additional info. So Intel easily is just being given the benefit of the doubt and everything is just being eaten up. Just do not expect the assault to end soon as FUD in the media is a powerful tool, maybe not forever but maybe just long enough.

 

It's important to remember one of the complaints not addressed is that while the resolution is 1080p, the game settings are not highest / Ultra, and critics say the effect is like it is running the games at 720p.

Intel skewed the result by doing this again, as with increase in resolution the gap narrows significantly between the 2700x (all Ryzen / TR really) and the 9900k (Intel).

So if you are gaming at 1080p or 1440p or 2160p Ultra, these numbers aren't giving you the right details you need to decide whether to spend 2x $$$ on the 9900k vs 2700x, because you won't notice the difference at all or enough to matter.

Adding 1% to the average with your totals really doesn't help much either, even at "720p"...

It should also be mentioned that the Ryzen 1st generation 8c/16t CPU's - 1800x, 1700x, 1700 are all much cheaper than the 2700x / 2700, and will be just that much better price / performance overall vs the 9900k / 9700k / 8700k.

 

I think it's fair to test gaming performance of CPUs when games are operating at lower resolutions or settings to achieve say the 144 fps zone, which is where people who have high refresh rate monitors find the relevance. It also puts more emphasis on the CPU rather the GPU, which again is fair because that's what's being tested. If they were testing with GPU bottlenecks all the time, then there would be no difference between CPUs.

 

It's an artificial creation - a configuration that no one would reasonably use to game - so the results are not based on actual use, so it's useless.

Set up the hardware and games evenly optimized like you are going to use them in real life, then take the comparison, otherwise it's an artificially skewed view created only to sell Intel CPU's.

Still no satisfaction from Round 2 of " Intel and Principled Technologies Bogus Benchmarking Adventure", the slightly better sequel.

Intel still got it wrong.

 

Intel should have launched the 8/16 CPU with a few different SKUs of the same CPU but a few different clock speeds.

They could have launched a lower clocked one at $329, and then launch a mid grade one at $399, but saved the binned top tier chip at $499. I guess that would be milking the customers a bit though, and no company has ever, EVER done that. Nope, not a single time in recent history can I think of this ever happening. I guess at least with the Intel chip you are getting HT for your $120.

 

See, if Intel launched the 9900K at between $380 and $420, I wouldn't have said a word or I would even recommend it. Even though known information puts performance Delta around 16%, let's just go with 20%. The 2700X is around $320. Add 20% to it ($64), you are at $384. Even at $420, paying around 30% more for 20% (16% by what we currently know) more performance, that is still in the range of reason. You slap 20% more on that, now at 50% more for under 20% performance, you lost me. You then add 10% initial gouging to $510-520, you are at 60% additional cost for 20% tops, which that gap closes once you peg 1080p settings to ultra like most do.

Then, we saw an additional $60 added on, bringing it to around $580 per chip, or about 80% over the cost of a 2700X. At that point, picking up the cheapest 8700K or 7700K and building a companion build for capture starts making sense if you are a steamer. For a non-professional gamer, that is an insane price. For an overclocker hobbyist/enthusiast, if not XOC, value isn't there, but if you got the money. For a professional overclocker, you've got no choice, you have to have it.

This is what I mean by value proposition. Meanwhile, of it is all the same silicon and all still unlocked, if they left the ht in, they would see the same result of the 1700 vs the 1700X on sales, something remedied in the second gen by closing the price gap a bit, giving certain other tweaks, etc.

Sent from my SM-G900P using Tapatalk

 

I 100% agree at the scalp prices that are currently being asked it's not really worth it. That isn't the fault of Intel however as retailers can charge whatever they want. Look at the Vega launch, where were those low priced GPUs that AMD announced? Everyone came to the defense of AMD and blamed retailers. I'm just glad I locked mine in at $499.99 but I got lucky and grabbed it in the first few minutes. The poor souls paying $579.99 plus taxes from Newegg are victims of daylight robbery. The demand isn't there this time around vs the 8700K and that very evident by the availability of preorders on both Amazon and Newegg, or it could mean Intel has been stocking up for quite some time as to not repeat their "paper launch" from 8th gen. I think we will see prices normalize and drop fairly quickly after launch. I also expect to see some decent deals at places like Microcenter for those of us lucky enough to have one near by.

 

Testing games at 144fps is far from artificial, it's high refresh rate gaming (with panels going up to 240Hz and likely even higher in the future - there's even theoretical value in 1000Hz displays when it comes to motion clarity, see Blur Busters website) . Besides, like I said, if you test in a GPU bound scenario then it doesn't matter what CPU you have, just buy anything that's above the bare minimum required for your fps. Another argument for testing at higher frame rates, and therefore lower settings or lower resolutions, is that it does highlight which CPU can push the most frames, which is useful to know the full extent of the 'gaming power' of the CPU, which then would show you how future proof those CPUs could be for gaming, with the ones that have the most overhead spare being the ones likely to be most future proof. It's insane to test the performance of gaming CPUs under GPU bound scenarios - that's just obvious.

 

Noooooo!!

Let's not codify Intel's ******** by starting to use their artificially tilted scores meant to sell 9900k Pre-Orders!

Sheeeesh man, really? After all of this you are gonna fall for it? Wow...

Let's wait for real tests done by the reviewers we know will set up the play on an even ground, and do reasonable testing on a wide range of games - ones that favor each CPU in equal measure, and give us a nice chart of % difference with averages unskewed for a variety of resolutions and usage modes.

October 19th can't come soon enough...

 

Chill. This was to run the numbers with easily dealt with multiples of $32 up the stack for analysis. It showed at each price point where the percentage deviated further and further from 20%, and the performance Delta, as we all have seem, is even lower than that. So, at $420, it is really 30% more cost for about half the performance percentage boost. By $480, you get 16% more performance while paying 50% more, or three times the amount of the performance. At this point, they surpassed the 30% performance on the 2080 Ti over the 1080 Ti while paying 70% more. At 60%, or the $520 area in price, they are paying, as a percentage, 4x what additional performance they would get over the market competitor. Once you are at $580, you need your head examined, period. You've got more money than sense.


Sent from my SM-G900P using Tapatalk

 

I understand, you need a number to plug into your fantasies - Intel knows this, that's why they did it. Let's see, what number makes sense to use between now and October 19th?

Ok, how about we halve the percentage and round down instead of round up?

Why give any ground to Intel's BS? Instead let's make a penalty for their bad behavior.

And, to keep it fun and friendly, let's make the score a joke that makes everyone laugh, and realize it's probably not real, but doesn't help Intel sell squat.

Just until the real numbers come out. Then we can find out the real deal - in the many modes of usage.

Let's make it 10%.

Probably, realistically more accurate for most setup's and users, a difference no one will notice.

Most games tap out at 4 - 6 cores full usage, so the difference going from 8700k to 9900k adding 2 cores for most games won't make it a worthwhile upgrade for most existing 8700k gaming owners.

Same goes for existing 8c/16t AMD owners. Same goes for new AMD 8c/16t owners, those trying to decide whether to save money and still get great performance, or blow 2x as much for the 9900k to get 10% better FPS.

Yeah, 10% sounds generous.

How do your numbers work considering 10% improvement, on average?

 

At 10% improvement, I give up to $380-400 as the recommend, but that is because Intel excels at single threaded performance in other tasks and the 5GHz will help with that. $420 is looking way less palatable, and $450 is insane. Above that price, it shouldn't even be a conversation as it makes no sense. So it moves my top price down a bit for my recommend and the ludicrous price down a lot.

I also found the bulldozer analogy apt above on their goal, along with using 50% cherry picked numbers to then push the 50% increase in price when Intel knew better. I believe that is internal pressure for margins with issues on volume sales. But that is a different discussion and analysis.

But, at the 20% increased cost on 10% performance over Ryzen 2700X, I'd still give a buy only because of speed and IPC giving boost to single threaded and lightly multi threaded workloads. If AMD had better IPC negating the Intel speed advantage, I'd have a different answer (which Zen 2 rumors don't bed brought into this thread other than a potential wait and see what AMD comes out with if a person can wait to build their next system). But that fits more mainstream workloads, single threaded or lightly multi threaded. But, it should be mentioned, those mainstream users don't yet use all that any 8 core CPU has to offer in ordinary activities outside of gaming. Once they realize they can do more, we may see more content creators emerge, like we have since the Ryzen launch kicked off this rush to more cores. That changed game design, streaming on a single PC, etc.

But that is my honest opinion, at $380, the 9900K, whether 10%, 16%, or 20%, would have been something I would recommend at that price.



Sent from my SM-G900P using Tapatalk

 

Everyone remember last October 2017 when Micro Center wanted $500 for the 8700k at launch? MSRP was $359, most retail prices later subsided to ~$420 = ~16.667% markup

9900k MSRP is $488 (tray) | $ 499 (box) x 1.1667 = $570 | $ 580
Today's Newegg price is $ 580

MicroCntr charges sales tax, here it's 6%:

• example $520 CPU = $550 (tax)
• (common to get ~$30 off motherboard bundle)(nice)

MicroCntr reliably undercuts Newegg prices (partial tax offset), there are three MCenters in my area, I love the place to death as does everyone, but at launch I suspect they want a piece of the pie too. It'd be no surprise if they ask $540+ ($572 tax). A 'legit' price (relative to Newegg) may be $520 ($550 tax). What to make of that compared to say ShopBLT's $520 (no tax) pre-order is to shop smart online 1st, or, pay a tad more for the great MCntr in-store experience as I normally do --> one typically runs into a bargain on other related hardware at a discount for possible $ offsets = pleasure, not pain

Newegg's $580 9900k? 16% markup --> it's no more outlandish than 8700k's 16% markup, but certainly feels outlandish ... ... ...

edit to new reply below:

Yes, in a vacuum is what I had in mind when I posted + $370 for 8700k was much more palatable. 8700k exactly a year ago in those 1st few months: initial 'scarcity', scalping, then leveling down to a 'rational' ~16% markup. Ryzen's subjective worth relative to CovfefeLake isn't my concern; that's being fielded as it should be & makes for good reading here ... Principled Technolgies blunder has been gold

 

If looking in a vacuum, yes. Just looking at the price alone, not market competition, then it seems more reasonable. But the 8700K was available at times this year for $370 or less. Prices shot up during the shortage and since Intel's numbers and pricing was released.

Edit: also, the 8700K must be compared to the first gen Ryzen at that time.

Edit 2:


Sent from my SM-G900P using Tapatalk

 

Something else to look forward to on October 19th, Microcenter's Intel sponsored pricing bundles.

Any word on MicroCenter 9th Gen (9900k) pre-order availability?
https://www.reddit.com/r/intel/comments/9mq9gf/any_word_on_microcenter_9th_gen_9900k_preorder/

"Just talked to someone on the phone who pretty much had no useful information on this for the Dallas location.
Has anyone heard anything different from their local spots regarding pre-ordering specifically?"

Pyromonkey83 4 points 4 days ago
"I too contacted my local MicroCenter in Denver and their only response was that they will not be taking pre-orders at this time. Last year they also did not accept pre orders for the Coffee Lake launch, so I'm not too surprised.
To my knowledge, the only pre-order they have ever done in the past was the recent Nvidia launch, and that obviously went very poorly all around. It's possible they have decided to forgo pre-orders all together and stick with launch day rushes like they did in the past."

QLJ1986 2 points 3 days ago
"**** MICROCENTER! Dont even bother dude! Last year with the coffee lake release I stood in line and once they opened the store they only had a few 8600ks. I heard that employees bought the 8700ks. I would not count on them to have any in stock. I stood in line for hours for nothing..."

terencecah 1 point 4 days ago
"When is the actual i9 release date?"

aragorn18 3 points 4 days ago
"Oct. 19th"

 

Joker gets his first direct from Intel review kit.

Watch for the sly RTX 2070 "accidental" shot...

i9 9900K UNBOXING + EVGA Z390 FTW Motherboard
Joker Productions
Published on Oct 13, 2018
Unboxing the reviewer's kit for the Intel i9 9900K, 8-core CPU. In addition to that an overview of the EVGA Z390 FTW motherboard.

Spoiler: Show Index

Pinned by Joker Productions
Loque 16 hours ago
Time stamps for ease of access:
00:05 King Pin Boss Intro Pose !
00:24 Commercial break, take a tic tac !
00:58 Japanese Porn !
01:37 That's what she said #1
02:01 Who knew The tesseract had intel inside !
02:40 The actual CPU!
03:20 That's what she said #2
04:00 We almost see where Joker lives
04:00- 04:19 Joker leveling up his dagger skills
05:00 Joker takes the dress off the motherboard really slow
06:15 Tacky little things!
07:16 The actual motherboard !
08:00 Check out what's in the "rear"!
08:25 Check Out the sizes !
09:14 Take a moment to gaze at the lovely decoration on the wall behind Joker!
10:37 A moment of gratitude from Joker while he does his King pin pose for the outro !

My Response to Intel & Principled Technologies
Science Studio
Published on Oct 13, 2018
It's not every day that we see a commissioned company publish such incompetently-acquired benchmarks... but who's really to blame here? Why did Intel do this to begin with? This is my response to both Intel & Principled Technologies.

Spoiler: Click to See Video

 

Buildzoid shares many of my opinions on pricing for the Intel chips, my opinion on the HT vs no HT on the 8700K vs 9700K, and the value on the 2080 Ti. Thought worth sharing.

Sent from my SM-G900P using Tapatalk

 

as far as king of the hill for gaming it is well known there is on real threat from AMD 12nm. I again say Intel is trying to say to investors that the new CPU's are so advanced and fast that 7nm is not going to give Ryzen or Epyc the performance advantage. So far the media and investors seem to be eating it up. They did this with the FUD so far in advance of the NDA so even if there were 7nm info there is nothing to compare it too.

In the end if the writing is clearly on the wall then they will do the best they can to scramble up the information where ever they can with misinformation. Would you expect any less?.

 

If the 8700K at stock is ahead, I wonder how the 9900K will fair since it's faster than the 8700K. Again this is at stock and we all know the 2700X will boost itself (auto overclock) to dang near maximum out of box to compete.

 

You didn't listen to what he said. He said, clearly, neither were overclocked. That means precision boost overdrive was not on on the 2700X. That means you are reading in what you want to hear.

So you are wrong on how Ryzen works, are misstating his testing conditions, etc. Watch the video again please or wait for his video of the 9900K overclocked against the overclocked results on the 2700X.

I'm amazed you think he would make a mistake like that without discussing testing conditions after his rants about Intel's paid testing. Go on with that nonsense.

Sent from my SM-G900P using Tapatalk

 

Thanks but I listened to him just fine.

@9:19

He clearly states that even with an H100i V2 the Ryzen CPU would have only run at max an additional 100Mhz, clearly stating the CPU is already boosting to max out of box on it's own. If a beefier cooler were to be used it would boost 100mhz further, right from his mouth.

Tell me again how I didn't listen?

 

100MHz left in the tank means it isn't at Max. Period. That would be like saying that you are benching your 8700K at 5.1, but your sample could bench all at 5.2.

Now, if you are referring to AMD not overclocking much beyond the XFR frequency, then say it. But say it clearly without exaggeration.

Sent from my SM-G900P using Tapatalk

 

I'm not quite sure just how much more clearly I could have stated that for you.