Intel Core i9-9900K 9th Gen Review

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
This seems to be the last gasp and send off of the current architecture for Intel. There doesn't seem to be much justification to buy this for someone who is already on an 8700k from my personal perspective. Especially not given the requirements of what a build would need to get the most from this CPU. As said over at Anandtech:

"At $488 MSRP, plus add $80-$120 for a decent cooler or $200 for a custom loop, it’s going to be out of the range for almost all builds south of $1500 where GPU matters the most. When Intel’s own i5-9600K is under half the cost with only two fewer cores, or AMD’s R7 2700X is very competitive in almost every test, while they might not be the best, they’re more cost-effective."

Double the cost of a 2700X and double the power just doesn't make any sense and especially not when gaming at resolutions at or above 1440p.
 


Paul is correct, increased energy (temps) leads to higher electron leakage and reduced efficiency.
 


AMD sent over its official price list the night before the embargo for this review lifted. That price is directly from AMD.
 
@Tamalero

Hey Tamalero! Thanks for posting on Tom's Hardware.

I have to defend jimmysmitty on this one. Torture loops are fringe cases and they shouldn't be used to represent the general performance of a CPU. This was Jimmy's point when he was replying to s1mon7 who said that the 9900k consumes twice the power of the 2700x. There's no debate that the 9900k eats more power than the 2700x, but the difference is exaggerated in a torture loop.



You might see 100% CPU usage in rendering situations but its not happening literally all the time. If you compared the power consumption of a 2700x and a 9900k over several years, the numbers wouldn't be 2x. Anyone who can prove otherwise is an anecdote that doesn't account for the vast majority of users.



Intel's 9900k is the very best CPU that isn't in workstation or server territory. They're allowed to charge a premium for it as long as there is no immediate competition. Nvidia is doing the same thing (even more so) with the 2080 Ti. The cost of a 9900k build is going to be more expensive than a 2700x build. Redgarl claimed that a $1000 2700x build is roughly equal to a $2000 9900k build, but when you run the numbers, that's not the case at all. Our friend Jimmy explained that this price difference would only exist if you bought a high end Intel motherboard and compared it to a cheap AMD motherboard.

I agree with you that you DO need a nice cooler for the 9900k. A Noctua u14 cooler is $63 at Amazon. A be quiet! Dark Rock Pro 4 is $90. Neither cooler is cheap but nether account for the price discrepancy.

If jimmysmitty's post defended Intel, it was because Intel deserved to be defended in that situation.
 


Ahhhh CMON Paul...?!! At these prices, I would buy a TR 2950x over that piece of garbage.

Review should say, faster at 1080p with a minimum of a 1080 GTX. Anandtech review is quite accurate that this chip is not that much better at anything above 1080p.

https://www.newegg.com/Product/Product.aspx?Item=N82E16819117957&cm_re=9900k-_-19-117-957-_-Product

https://www.newegg.com/Product/Product.aspx?Item=N82E16819113499&cm_re=2700x-_-19-113-499-_-Product

n.b.: ROFL, 2700x - Tom’s Hardware - Best Overall Gaming CPU of 2018
 
A little disappointing I would say. Not a real compelling performance argument for a new 9th generation i7 or i9 CPU over a previous gen i7-8700K, and it seems what little insignificant performance advantage they DO have comes with allot of power headaches and overhead, especially if you aren't using a $600 mobo (See Tom's review of the new Taichi and other Z390 mobo's published today).

Might have been worth mentioning that the i9-9900K performance and value proposition could change in the future if (1) more multi-threaded games start to get released and (2) improved motherboard BIOS and firmware versions get released, better tuned for the new CPU's.

But TODAY, well, it seems like your your power headaches get worse, your cooling options get fewer, and all in exchange for....nothing, really.
 
I see Tom's Hardware is continuing their bullshit shilling for corporations. First the "NVIDIA Just Buy It" article, now they are spreading false information "Fastest Gaming Processor Ever" and shilling for Intel
Stop discrediting yourself even TOM
 


Actually the empirical data contained within Tom's Hardware review of the new CPU's is far LESS flattering for Intel then what I'm seeing from many other bench-marking sites, which mostly seem to show a larger performance gain then Tom's.

The varying delta's also seem to correlate strongly with mobo's and coolers used, so it may be awhile before a clear picture emerges.

I'm guessing that you have become personally invested in a decision to buy an AMD CPU?
 


Right!?!! To be fair though, Tom's has also simultaneously published reviews using cheaper mobo's today..
 
i started off applauding Intel for finally coming back to solder, only to find out they really had no other logical choice for this CPU; it runs so hot w/solder, that it would have lost to their own 8700k w/o the upgrade. actually, it may not have been market-viable at all. the 7700k OC'd was a bit toasty, 8700k's 50% added cores was hot enough to throttle w/o liquid, so i can only imagine yet another 2 Skylake cores with only TIM... they'd have lost all their clock advantage over the 2700X. which, given there's only 10-15% perf delta between the two, may have lost to the 2700X almost entirely (IPC advantage still).

i understand not everyone will run fully-threaded demanding workloads, but Meltdown security fixes aren't going to help with actual power-draw meltdowns for those who do stress their systems -heh- ...insanely high consumption! no wonder the mobo vendors are struggling to feed this thing.

the 9900k is certainly the best performing desktop platform CPU avail, & i'm glad (for competition & just as a fan of tech) to see they've kinda made it work ...but sweet-baby-science look at those added system costs on an already VERY expensive mainstream chip! $150 for an above-avg AIO cooler, & mobo prices that make HEDT blush.

it's like Intel wasn't even aiming near a "value" target on this one. which will be fine for a few people, but not recommended (i understand Tom's "Choice" tag).

---thx for the review, Paul
 


When I was making my original post I perfectly understood that the CPU obviously does not use double the power 100% of the time. I'm quite sure everyone understands that, as they can also see the idle or gaming power drain results.
The fact that the CPU lowers its clocks most of the time and enters lower power states unless needed or processes fewer threads than it's capable of will obviously drop it to low power drains, but any processor does that, and it is not a representation of CPU power needs. During a mixed usage, it will swing between 2700x-level power drains and double that depending on the usage, sure, but you can say that it won't be in the state drawing max power all the time about almost any power-hungry component. That doesn't change the fact that it does drain twice the power in the top performance state it was designed for.

Saying that if you observe them over several years the numbers wouldn't be 2X is true with any high power drain device - even your microwave, since it's not cooking your food 100% of the time and can actually use less power than your CPU if you observe it over the years and you don't use it for what it was designed for too often. That is not a good metric for a CPU and completely obscures the fact that it is a power hungry CPU. You need a PSU, cooling and case ventilation all able to handle such requirements because the CPU needs to be treated like a >200W CPU, as opposed to a ~100W Ryzen 2700X CPU, because it will work like one, even if it's not 100% of the time, that doesn't matter.

The fact remains that when you use this chip as it was designed - a 8c16t CPU at close to 5ghz, it will use twice the power of a fully loaded 2700X at its full capacity to deliver a ~10% performance gain. It will extend even further when overclocked. That's just really bad efficiency. What an average user does, especially with a CPU rarely targeted at casual CPU buyers, we don't know. Since this is a heavy duty chip meant for heavy workloads and overclocking, it is likely to spend plenty of its time in states that in fact make it use double or close to double the power of a 2700X, as that's what it does when fully loaded. That's a perfectly valid argument.
 
Back when the 7700K was the "fastest game" CPU no one saw any practical reason there would be to gain with more Cores for the overall bulk of games... Compare the 7700K to the 9900K and tell me where the justification is for buying a 9900K based on FPS rates using top of the line graphic cards at 1020P rendering? As I've requested before, show the CPU utilization levels for these FPS tests and something important will be revealed about the overall efficiency of these "game" benchmarks that might add some value to the conversation.

I find it hard to justify where moving beyond a 7700K or Ryzen 7 2700 delivers any perceptible differences on the screen. What the eye can perceive and what delta there is between say 60 fps and 90 is very difficult to quantify. At the level these benchmarks are being played, even at 1080P if it takes a 7700K to provide playable frame rates and the top of the line graphic card that means the overwhelming bulk of the gaming market doesn't have the hardware for it at any given point in time. I'm sure the 9900K properly set up and running something that actually can benefit directly from its $500 cost or 60% increase in cost over a 7700K will find a market but from where I set (with both a non overclocked watercooled 4790K and 8086K) using gaming FPS benchmarks to sell this as the "fastest gaming CPU" is like saying buying a ZR1 Vet will get you to work faster than a Ugo if you ignore the penalties for exceeding the speed limit to do that else the Ugo and the ZR1 provide the same experience at the legal speed limit and no value is added by the $100,000 ZR1 for that use. Look at the FPS carts and plot against cost. The 9900K is a net loser there. It just doesn't add any perceptible and meaningful value to "games". Like the Threadrippers and even the Ryzen 7 2700 series, said multi-core things aren't really taxed by "games" beyond even 4 Cores. Games are all about IPC rate gains not Cores. The law of diminishing returns sets in immediately with SMP multi-threaded tasking in the real world where a computer system isn't solely tasked for a single benchmark like task. The 9900K will find a market but it won't be for "games" just the Threadrippers aren't.
 
"AMD sent over its official price list the night before the embargo for this review lifted. That price is directly from AMD."

Paul, the launch price is $329 and this is reflected correctly on page 1 of your review. If you are saying that AMD has since officially increased its official price to $378, then please reflect that on page 1.

Additionally, you should at least reach out to AMD for clarification, given that the price increase ($378 for 2700x) does not make any sense. And especially since you used this new price to compare the 2700x to "The $263 Core i5-9600K" making the price disparity way bigger than actual market prices.

Otherwise, that particular price comparison is unfair.
 


I wish I could wear those AMD goggles everyone else is wearing. I have the R7 2700X and an i7-7700K and I just don't see it. For the most part they are similar, but when things get rough in a game my 2700X turns belly up while my i7 chugs along with 20-30% higher frames. Maybe that's an anomaly, but Ryzen's architecture just doesn't seem like it's designed for high-performance gaming.

See
https://www.youtube.com/watch?v=eIsK58nUcE0
https://www.youtube.com/watch?v=eS37fjP_FWw

*Update: To clarify, from where I'm sitting the 6-core i5 is possibly the best gaming CPU price/performance.
 


Regarding the motherboard scenario, I think its because for the Ryzen system, one can basically bring out 95% of the potential of a r7 2700x with a cheaper B series board while the i9 9900k requires a decent Z series board and decent cooler to bring out its true potential. Just because one motherboard is cheaper than another doesn't inherently make it worse. In this case its even more true because even using the best AMD motherboard, most Ryzen systems won't overclock or boost much more than your average or even cheaper B series board, while Intel's CPU's can take more advantage of better motherboard and cooling. Thats just my two cents.
 


That's really quite an eye opening result. My feelings tell me its just an anomaly, but perhaps this is normal? Is there anything in the background eating up core usage? I guess the story is that although the r7-2700x is 10% slower than even a i7-7700k in gaming sometimes, it means that an i7-7700k compared to an i9-9900k in gaming seems closer, which doesn't paint Intel in a good light either.

 
After Intel's tainted benchmarks crapping on AMD, I have lost all respect for that company and will never buy Intel.
 
After Intel's tainted benchmark results claiming ridiculous performance over AMD, as if we wouldn't find out, I can't trust Intel any more.
 
There was nothing in the background. I've ran the benchmark about 15 times on two different fresh Windows installs and it's pretty much the same everytime. I was wondering if the game was an anomaly. I'll have to install more games to test. But i gave my i7 to a friend because I also bought a z390 Mobo, just have to buy a CPU. I'm kind of leaving towards an i5-9600K since my internet is too slow to stream. I wanted the 9700K but not for over $400.
 


I think they were struggling with how to market this. You can get much better and more powerful rendering processors at the price. Heck, you can get a threadripper that will run circles around it for the price, and they'd have to compete against that. Instead, they targeted the mainstream. AMD has the 2700x Ryzen, which multi-threads about as well for half the price, so the comparisons wouldn't look good at all. They went with gaming, as that's where the larger difference is.

The problem is that for games, such core counts stiil don't matter much, with the Pentium G4560/G5400 pulling 70% of the frames the 9900k does for 10% of the price (and stutters in some games being hard to articulate in marketing materials), and the 6700k/7700k/8600k/8700k are still faster than AMD chips in this regard, even if they (when overclocked) perform almost as well as the 9900k does.

But they went with that, since in the worst case they are just competing with themselves, and at best they can claim that they have the fastest processor at SOMETHING, even if its advantage over the nearest, much cheaper competition is almost irrelevant. For "money is no object" gamers, this will be the go-to processor, no matter how wasteful of a purchase that is, but considering the tiny market for >$500 processors in general, I guess that's a better fit than any other.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom