Monday, May 18th 2020

Comprehensive Core i9-10900K Review Leaked: Suggests Intel Option Formidable

A comprehensive review of the Intel Core i9-10900K 10-core/20-thread processor by Chinese tech publication TecLab leaked to the web on video sharing site bilibili. Its testing data reveals that Intel has a fighting chance against the Ryzen 9 3900X both in gaming- and non-gaming tasks despite a deficit of 2 cores; whereas the much pricier Ryzen 9 3950X only enjoys leads in multi-threaded synthetic- or productivity benchmarks.

Much of Intel's performance leads are attributed to a fairly high core-count, significantly higher clock speeds than the AMD chips, and improved boosting algorithms, such as Thermal Velocity Boost helping the chip out in gaming tests. Where Intel loses hard to AMD is power-draw and energy-efficiency. TecLab tested the three chips with comparable memory- and identical graphics setups.

More charts follow.

The games above are Assassin's Creed Odyssey, Metro: Exodus, and Tomb Raider.

Find the video presentation (in Chinese language) here.
Source: TecLab (bilibili)
Add your own comment

159 Comments on Comprehensive Core i9-10900K Review Leaked: Suggests Intel Option Formidable

#1
XL-R8R
Summery: Intel finally gives users a viable alternative to AMD's 6 month old 3900x while using 10-20% more power.



Nothing much is news-worthy in this news.





Also @btarunr there is a spelling mistake; "change" should be "chance".



Edit - arent the games shown in the screenshots, the ones that usually favour clock speed/Intel CPU's any way?
Posted on Reply
#2
XiGMAKiD
Competition is (figuratively and literally) heating up
Posted on Reply
#3
Dredi
Their puget bech results are off by about 20% for the AMD systems when compared to benches run by puget itself. Seems a bit iffy.
Posted on Reply
#4
Darmok N Jalad
XL-R8R
Summery: Intel finally gives users a viable alternative to AMD's 6 month old 3900x while using 10-20% more power.



Nothing much is news-worthy in this news.





Also @btarunr there is a spelling mistake; "change" should be "chance".



Edit - arent the games shown in the screenshots, the ones that usually favour clock speed/Intel CPU's any way?
Isn’t it more than 10-20%? I’ve read board makers are prepared to give the chip 320-350W of headroom for max sustained TVB. That‘a more like 100% more than 3950X at 145W.
Posted on Reply
#5
Dave65
Dredi
Their puget bech results are off by about 20% for the AMD systems when compared to benches run by puget itself. Seems a bit iffy.
Very IFFY!
Posted on Reply
#6
Firedrops
the AMD figures for single core/gaming could easily be substituted within margin of error for the 3700x, then suddenly intel only has slight performance increase over the much cheaper AMD, and is comparatively overpriced.
Posted on Reply
#7
Mats
Darmok N Jalad
Isn’t it more than 10-20%?
It's 29 % more than the 3900X in that test, probably not the CPU alone.
Darmok N Jalad
I’ve read board makers are prepared to give the chip 320-350W of headroom for max sustained TVB. That‘a more like 100% more than 3950X at 145W.
I don't think it's a good idea to grab numbers from different sources.
Posted on Reply
#8
willgart
wow... 28% more power than the 3900X.... crazy hot !
Posted on Reply
#9
Object55
Great news, Intel Top end chip almost beats AMD's midrange.
Posted on Reply
#10
theoneandonlymrk
Kinda funny, now we're getting posts suggesting Intel are not too bad at gaming and are a viable option.

Where were the news posts declaring Ryzens as the same ,when intel were not formidable.
Posted on Reply
#11
Darmok N Jalad
Mats
It's 29 % more than the 3900X in that test, probably not the CPU alone.


I don't think it's a good idea to grab numbers from different sources.
My source was Anandtech, so I don’t have an issue with those numbers.
Users wanting the 10-core 5.3 GHz will need to purchase the new top Core i9-10900K processor, which has a unit price of $488, and keep it under 70 ºC to enable Intel’s new Thermal Velocity Boost. Not only that, despite the 125 W TDP listed on the box, Intel states that the turbo power recommendation is 250 W – the motherboard manufacturers we’ve spoken to have prepared for 320-350 W from their own testing, in order to maintain that top turbo for as long as possible.
www.anandtech.com/show/15758/intels-10th-gen-comet-lake-desktop
Posted on Reply
#12
Decryptor009
So not really anything new then.

The intel chip is faster for my needs, but not fast enough to warrent the price hike and power consumption / heat hike either.

Generally, intel would need to be out pacing AMD something like 50% to even get me to consider it because AMD gives us such a great package at a great price, this is the difference intel, i think it would be nice if you know.. could learn?
Posted on Reply
#13
Turmania
They really needed the die shrink this launch.
Posted on Reply
#14
ZoneDymo
yeah if you push clock speeds and add more cores, im sure it is "formiddable".

cant wait for the 11900K or 20900K that has 12 cores and boosts to 5.5ghz that we can call once again "formiddable" while ignoring the rediculous power consumption and heat.
Posted on Reply
#15
Mats
Darmok N Jalad
My source was Anandtech, so I don’t have an issue with those numbers.
Those are from documentation. Nothing wrong with Anandtech as a source, but pulling numbers from different sources is a tricky one.
I want side by side comparisons from benchmarks and measured power draw, like the OP. Not sure how legit this one is tho.
ZoneDymo
cant wait for the 11900K or 20900K that has 12 cores and boosts to 5.5ghz that we can call once again "formiddable" while ignoring the rediculous power consumption and heat.
The next one is supposed to be the biggest improvement since 2015, so maybe it has a chance to be not as hot as Comet Lake. :D Still 14 nm tho..
Posted on Reply
#16
john_
The average consumer will buy this CPU, put a mid range water/air cooler on it and never see these numbers.
On the other hand, the average consumer who will buy an AMD CPU, will put a mid range water/air cooler on it and probably see results close to these numbers.
Posted on Reply
#17
Mats
john_
The average consumer will buy this CPU, put a mid range water/air cooler on it and never see these numbers.
The average consumer doesn't buy a $500 CPU.
Posted on Reply
#19
john_
Mats
The average consumer doesn't buy a $500 CPU.
I think you are wrong here. There are plenty who will buy the best CPU and then try to cut costs from other parts (mobo, PSU, etc....).

Also let's be a little logical here and don't expect the other person to explain the obvious. When I am saying about average consumer, I am not talking about the whole market, from Celerons to Threadrippers, I am talking about those who will buy that CPU. Maybe I should say "the average buyer of that CPU", but even then you could argue that the average buyer of that CPU has a masters degree in CPU cooling, overclocking, memory optimizations....
Posted on Reply
#20
Decryptor009
john_
I think you are wrong here. There are plenty who will buy the best CPU and then try to cut costs from other parts (mobo, PSU, etc....).

Also let's be a little logical here and don't expect the other person to explain the obvious. When I am saying about average consumer, I am not talking about the whole market, from Celerons to Threadrippers, I am talking about those who will buy that CPU. Maybe I should say "the average buyer of that CPU", but even then you could argue that the average buyer of that CPU has a masters degree in CPU cooling, overclocking, memory optimizations....
Average consumers never pay $500 for a CPU, it is just plain common sense.

You also moved the goal posts making your whole statement very illogical in the end.
Posted on Reply
#21
john_
Decryptor009
Average consumers never pay $500 for a CPU, it is just plain common sense.

You also moved the goal posts making your whole statement very illogical in the end.
Whatever makes you happy.
I have learned not to argue with people who post half a line of "true facts".
Posted on Reply
#22
E-curbi
XiGMAKiD
Competition is (figuratively and literally) heating up
Heat output decreases as silicon quality increases. Inverse relationship.

If your work applications can truly benefit from 20threads to get work completed quickly and you wish to take advantage of Intel's lower memory ddr4 latency values - there's always the option to pick up a 10900K selectively binned by one of the aftermarket sellers. Yea, it costs more for the higher performing bins, but if it's your work computer - there's an obvious Return on Investment if the CPU increases your ability to produce quickly.

Siliconlottery.com in Texas
Der8auer in Germany
8-Pack in the UK

I'd wait for Rocket Lake's improved ST performance later this year, but then only 8cores 16threads to work with. :oops:

So the decision is better made by your work applications need for clock speed vs number of threads, which is always the case when selecting a work CPU from Intel or AMD.
Posted on Reply
#23
SIGSEGV
power hungry cpu.
I don't know why I am so happy to see Intel having a hard time against AMD Ryzen to take the performance crown.
Posted on Reply
#24
Mayclore
Since I have no interest in trying to install a fusion reactor into my PC case, I'll stick with AMD.
Posted on Reply
#25
Aldain
Formidable?? Really??
Posted on Reply
Add your own comment