Intel Core i5-10400F

[ncrs]

why do you need 2666 results,
3200- CL16 and 2666- CL14, are the same performance. If you include 3200 CL14 that should be reflected in the price, it is not the same price at all.

Because the processor is limited to 2666 on non-Z motherboards (which are the most likely target for people wanting this CPU) and testing above it is overclocking. Changing this will affect results and in some cases might even change the review conclusions.

 

[OneMoar]

so on a z extreme board with the limits released and a blkclock boost
its beating the 3300x by 1 fps at 1440 and maby 5 fps at 1080p ..
totally going to buy a 160 dollar processor and put it on a 500 dollar board
huzzah more Intel marketing shenanigans

 

[Dirtdog]

why do you need 2666 results,
3200- CL16 and 2666- CL14, are the same performance. If you include 3200 CL14 that should be reflected in the price, it is not the same price at all.

Not always. About the same real latency but that's not the only thing that matters otherwise DDR3 or even DDR2 would be superior.

 

[cucker tarlson]

great performance for the price (gaming,don't care about the rest,would be fine with a sandy i3 for what I do),good temperatures.

I was thinking about getting a z490,a 4400mhz viper kit and this as a placeholder for whatever lake is coming next.please talk me out of it.

 

[moob]

I didn't see it mentioned anywhere but are these benchmarks done with the built-in benchmarks in these games? I ask because Techspot has very different results in Shadow of the Tomb Raider for example but they're using a different section of the game as they found the built-in benchmark wasn't indicative of actual in-game performance.

Edit: I realize it's a different CPU, but taking the test systems together those are very very different results.

 

[cucker tarlson]

I long suspected TPU's gaming cpu tests are just gpu tests ran at 720p.

 

[mahoney]

That's not the point... Potential 10400F buyers are not going to be using the memory, motherboard (that allows to overclock RAM above 2666) and cooler used in this review in their builds.
Anyway the issues (conclusion in title and lack of 2666 RAM results) are fixed/will be fixed soon. Yet again @W1zzard proves that he can take warranted criticisms and act upon it.

The intel box cooler is more than enough for this cpu.

 

[Daven]

Hi there sir, 2017 was 3 years ago... things have progressed a little since.

A quad core for gaming was not the go-to part since 2018. It became clear already with the 7700K, it would fall short against higher thread parts despite a much higher clock. And that was not a quad, but a quad with HT, so 8T CPU. 4T CPUs have not been in fashion for gaming for a loooong time.

According to the TR performance curves in this review, all quad cores in the latest generations come within 15% of the performance of the fastest CPU tested at 1080p and higher. Can't really say that about GPUs where the fastest GPU is many times faster than budget GPUs and iGPUs. CPUs just don't have that kind of effect in games. Again, any latest gen quad core or higher will work if you are ONLY playing games and not multitasking with other applications simultaneously. All the performance graphs bear this out whether someone thinks its in fashion or not. If you are ONLY playing games, buy a $100-$150 CPU versus anything higher and put the savings towards the GPU.

 

[cucker tarlson]

According to the TR performance curves in this review, all quad cores in the latest generations come within 15% of the performance of the fastest CPU tested at 1080p and higher. Can't really say that about GPUs where the fastest GPU is many times faster than budget GPUs and iGPUs. CPUs just don't have that kind of effect in games. Again, any latest gen quad core or higher will work if you are ONLY playing games and not multitasking with other applications simultaneously. All the performance graphs bear this out whether someone thinks its in fashion or not. If you are ONLY playing games, buy a $100-$150 CPU versus anything higher and put the savings towards the GPU.

as much as I defend 6c/6t is still very fine for gaming (cuz it's true), 4c/4t are shieeeeeeeeeeet.even for gaming alone.

 

[W1zzard]

but are these benchmarks done with the built-in benchmarks in these games?

Of course not, not in any of the tested games. Test scenes do vary though

 

[moob]

Of course not, not in any of the tested games. Test scenes do vary though

Thanks. Good to know. It must just be the different scenes.

 

[cucker tarlson]

Thanks. Good to know. It must just be the different scenes.

scenes shoud vary.
if you wanna get into analyzing cpu benchamrks you better have time cause it's gonna take days.
if you think there's one magical chart you're not serious.

 

[W1zzard]

Tests at 2666 MHz have been added

 

[moob]

scenes shoud vary.
if you wanna get into analyzing cpu benchamrks you better have time cause it's gonna take days.
if you think there's one magical chart you're not serious.

That wasn't really the point. If you look at Techspot's chart for SotTR, it's clear they're using a scene that's hammering the CPU so you can spot the differences in performance much easier. For example there's a stark difference between the 10600K and the R5 1600. Compare that to TPU's chart, and there isn't much differentiation. That's why I asked if the built-in benchmark was used since it can be inaccurate. But knowing those aren't used for any of the tests is good to know going forward.

 

[Vayra86]

According to the TR performance curves in this review, all quad cores in the latest generations come within 15% of the performance of the fastest CPU tested at 1080p and higher. Can't really say that about GPUs where the fastest GPU is many times faster than budget GPUs and iGPUs. CPUs just don't have that kind of effect in games. Again, any latest gen quad core or higher will work if you are ONLY playing games and not multitasking with other applications simultaneously. All the performance graphs bear this out whether someone thinks its in fashion or not. If you are ONLY playing games, buy a $100-$150 CPU versus anything higher and put the savings towards the GPU.

Then I guess all you play is benchmarks because its not my experience. My 3570k became a stutterfest a few years ago.

 

[GeorgeMan]

No concrete plan for that yet. Today I've ordered 10100, 10300, 10320, 10500, 10700 and 10700K, which will keep me busy for a while

Aren't they supplying review samples? One cpu of each for TPU shouldn't hurt their inventory that much...

 

[W1zzard]

Aren't they supplying review samples? One cpu of each for TPU shouldn't hurt their inventory that much...

Nope, Intel sent out 10900k and 10600k, and no plans for further sampling, I asked

 

[GeorgeMan]

Nope, Intel sent out 10900k and 10600k, and no plans for further sampling, I asked

Oh that's unfortunate. Thanks for ordering and thoroughly testing them for us then. I'm personally very interested to see exactly how even low end parts perform, so thank you for your offer to the community.

 

[Dave65]

Pair this with the obvious value board and the cooler it comes with and it is just crap.

 

[OneMoar]

Tests at 2666 MHz have been added

Yikes took a considerable hit
unusual for intels there imcs have always done better at lower clocks them amd

 

[Nihilus]

Yikes took a considerable hit
unusual for intels there imcs have always done better at lower clocks them amd

Did you even look at the charts. It lost <7% at 720p using a 2080ti which is the absolute worst case scenario.

CPU performance was down a whopping 2%.

 

[heflys20]

It's "noticeably" good for gaming...If you pair it with something equivalent to a 2080ti. Other than that, it loses to the similarly priced 3600 in most cpu related tests. Obviously. At least it comes with a cooler, I guess.

 

[Gmr_Chick]

Lovely. This should clear up a lot of confusion going forward for many people. Because you guys have many awards. I had a discussion with my wife and it's the first time in the last 3 months we had a agreement. No one will buy an expensive Z490 board to pair up with this 10400F. Get cheaper board.
Looking forward to the updated review on the new chipsets. I heard the lower cost chipset for the 10th gen Intel will limit the DDR4 to 2666. That will, for sure, reduce the performance of the 10400F.

And thank you for the hard work. No matter the opinion is, I value what you guys did to inform us about new products. #respect.

"Expensive" is a subjective term -- what Person A considers expensive, Person B may not. While I certainly agree that pairing this 10400F with a $300-$500 Z490 board is absolutely insane, I CAN see the purpose of doing so if the intention is to upgrade to an i7 or i9 in a year or so. People forget that it's much easier to pop in a new CPU, GPU or RAM than it is to switch out a motherboard. So while such scenarios probably don't make much sense to most people here, I can see the sense in it somewhat.

 

[cellar door]

In which universe this is better value then a cheaper Ryzen 3600??? - that is because its not. You need a z490 to even touch the BLCK for that tiny OC, not to mention the memory restriction.

Sorry but this is a FAIL

 

[AddSub]

Excellent review! Pair this with a bitcoin miner surplus $120 GeForce GTX 1070 8GB off of ebay and a $60 16GB kit and you have something that will play anything in your gaming catalog. Play it well in fact. An Intel CPU of the last 2-3 generational itterations in the 4GHz+ range is pretty beastly for gaming. Productivity you say? If your job, you know the thing that pays the bills, depends on productivity, then pass GO and do not collect $200. That is, skip all made-for-average consumer/gamer CPUs and get yourself a Xeon (I did) or Threadripper chug-machine.

...
..
.