Intel Core i5-8400 Review: Six Cores On A Budget

[Guest]

Intel's Core i5-8400 is the cheapest modern six-core CPU out there. Can Coffee Lake on a budget hold its own against the mighty Ryzen 5?

Intel Core i5-8400 Review: Six Cores On A Budget : Read more

 

[velocityg4]

I'd be interested in knowing. Under what circumstances can you get it to hit 3.8Ghz reliably under a 6-core load? That is quite a spread.

If one was to delid the CPU and use a decent CPU cooler. Would it reliably maintain the max turbo boost when the CPU usage demands it? Is the stock heatsink and decent case cooling plenty?

On another note. It is time for the return of the Turbo button. That would be pretty sweet to click the button and manually have the CPU jump between 1, 2, 4 and 6 cores at their respective boost frequencies or down to standard. I know it isn't necessary as it is all automated and that wasn't the purpose of the Turbo button. Some people just like manual control. Plus old time computer geeks would get a kick out of it.

 

[AgentLozen]

This is a very neat chip at it's price point. Very little has changed between Kaby Lake and Coffee Lake except for the core count. Luckily for Intel, those added cores make all the difference.

This CPU would sit nicely on a budget system. It's a shame that there are no inexpensive motherboards that it could fit into like the conclusion of this article states.

If you didn't plan to overclock, this is the best CPU on the market for gaming and general productivity.

 

[The_King]

We have AMD to thank for all these budget cpu's from Intel, and you all know why! So thank you AMD.

 

[rwinches]

No, you can't have charts based on FPS or seconds or related to MSRP. Drivers update, Prices change, Memory Speeds Increase and price lowers, MB prices are wide range.

It is known that Some games work better for AMD when AMD GPUs are used. Game FPS can be dramatically improved by changing just one parameter. Test results that are milliseconds or a few seconds or frames faster are irrelevant and subject to variations in real world use on systems that are not clean installs and have other SW installed and running.

 

[10tacle]

No, you can't have charts based on FPS or seconds or related to MSRP. Drivers update, Prices change, Memory Speeds Increase and price lowers, MB prices are wide range.

I don't understand your point. This is a review of the 8400 and comparing it to other CPUs only. Memory, motherboards, and all the other variables are you talking about in a full PC build are irrelevant to this chart comparison. They have to establish a constant standard across the spectrum, and they did so.

It is known that Some games work better for AMD when AMD GPUs are used. Game FPS can be dramatically improved by changing just one parameter. Test results that are milliseconds or a few seconds or frames faster are irrelevant and subject to variations in real world use on systems that are not clean installs and have other SW installed and running.

Again, they are using a single standard across the spectrum comparison. Of course there are infinite combinations of hardware that can game change a little. The bottom line here is that among every major tech review website, all of Intel's chips are better for gaming than Ryzen. The only exception is when dealing with beyond 1080p gaming like QHD or UHD where it's mostly on the GPU. People who buy this chip are the perfect candidate for a 144Hz 1080p G-sync or Freesync monitor.

 

[lotusmotors]

Glad I didn't waste money on a 1600x.

 

[elbert]

Compared to the Ryzen 1600 with a B350 its $50 higher. That is a higher video card level. IE get the 8400 with a 1050ti the Ryzen budget would get a 1060 3GB. Intel needs their B360 motherboards for the 8100 and 8400 to ever be a budget winner.

 

[hardwarefox1234]

Has the author even bothered to look at the street prices for the CPUs??

Core i5 8400:

https://www.newegg.com/Product/Product.aspx?Item=9SIA7HN6HF3442&cm_re=i5_8400-_-9SIA7HN6HF3442-_-Product
http://www.microcenter.com/product/486090/core_i5-8400_coffee_lake_28_ghz_lga_1151_boxed_processor
http://www.ncixus.com/products/?sku=142465&vpn=BX80684I58400&manufacture=Intel

The lowest price is $249.99!

The Ryzen 5 1600 is much cheaper:

https://www.newegg.com/Product/Product.aspx?Item=N82E16819113435&cm_re=ryzen_5_1600-_-19-113-435-_-Product
http://www.microcenter.com/product/478826/Ryzen_5_1600_32GHz_6_Core_AM4_Boxed_Processor_with_Wraith_Spire_Cooler
http://www.ncixus.com/products/?sku=139481&vpn=YD1600BBAEBOX&manufacture=AMD
https://www.amazon.com/AMD-Processor-Wraith-Cooler-YD1600BBAEBOX/dp/B06XNRQHG4/ref=sr_1_3?s=pc&ie=UTF8&qid=1508952175&sr=1-3&keywords=ryzen+5+1600

Price is $199.99~$219.99!

Then if you add the price of the B350 motherboards,they start at a lower level than the Z370 ones.

 

[hixbot]

Why did you label section 10, overclocking, cooling and temperature? There's no overlocking on this chip? I would have like to see at least an attempt at Bclk overlcock with a mobo that has a clock generator.

 

[TJ Hooker]

I really wish more reviewers would look at what sort of boost clocks the i5 8400 can sustain under load, with the stock cooler as well as aftermarket. From what I know you could typically assume past Intel CPUs would operate at or near max turbo almost indefinitely, but with a 50% increase in core count and such an unusually wide gap between base and boost clocks (not unlike their mobile CPUs), I'm curious how Coffee Lake will behave in that regard.

 

[ATI9800Pro]

There's a typo in the temperature charts ; it reads "Core i7-8400".

 

[John__Titor]

Seems like a very solid CPU, but no way would I list the included cooler as a pro. It's cheap garbage. At the very best it's neutral. AMD has set a new standard for stock coolers that should be encouraged for budget processors.

 

[PaulAlcorn]

<snip>

On another note. It is time for the return of the Turbo button. That would be pretty sweet to click the button and manually have the CPU jump between 1, 2, 4 and 6 cores at their respective boost frequencies or down to standard. I know it isn't necessary as it is all automated and that wasn't the purpose of the Turbo button. Some people just like manual control. Plus old time computer geeks would get a kick out of it.

I second the vote for the return of the Turbo button. Good times!

 

[PaulAlcorn]

Has the author even bothered to look at the street prices for the CPUs??

Core i5 8400:

https://www.newegg.com/Product/Product.aspx?Item=9SIA7HN6HF3442&cm_re=i5_8400-_-9SIA7HN6HF3442-_-Product
http://www.microcenter.com/product/486090/core_i5-8400_coffee_lake_28_ghz_lga_1151_boxed_processor
http://www.ncixus.com/products/?sku=142465&vpn=BX80684I58400&manufacture=Intel

The lowest price is $249.99!

The Ryzen 5 1600 is much cheaper:

https://www.newegg.com/Product/Product.aspx?Item=N82E16819113435&cm_re=ryzen_5_1600-_-19-113-435-_-Product
http://www.microcenter.com/product/478826/Ryzen_5_1600_32GHz_6_Core_AM4_Boxed_Processor_with_Wraith_Spire_Cooler
http://www.ncixus.com/products/?sku=139481&vpn=YD1600BBAEBOX&manufacture=AMD
https://www.amazon.com/AMD-Processor-Wraith-Cooler-YD1600BBAEBOX/dp/B06XNRQHG4/ref=sr_1_3?s=pc&ie=UTF8&qid=1508952175&sr=1-3&keywords=ryzen+5+1600

Price is $199.99~$219.99!

Then if you add the price of the B350 motherboards,they start at a lower level than the Z370 ones.

We covered the price gouging in the introduction, specifically calling out the $260 price point. Last paragraph, page 1.

 

[PaulAlcorn]

Why did you label section 10, overclocking, cooling and temperature? There's no overlocking on this chip? I would have like to see at least an attempt at Bclk overlcock with a mobo that has a clock generator.

Good catch, fixed. Intel has been pretty good at locking out BCLK overclocking even on boards with an external clock generator. We'll keep an eye out for any further attempts at this, though.

 

[Nintendork]

VELOCITYG4
Why even bother having to delid a cpu and change tim to actually have a working cpu instead of simply buying the cpu king Ryzen 5 1600-1600X and forget about having to patch intel's mess?

 

[HERETIC-1]

Hi Paul,
When you mention-comes with a free cooler-would have been nice if you could
have tested it-perhaps running Handbrake,and also then checked what turbo
speeds it could maintain with stock cooler.
As a budget 6 core I feel that is typical use,

I feel for gaming-those that can't afford 8600K are likely to go for the cheaper
4 core 8350K,which will probably be faster than 8400 in most games............
If we get a non K 8600 next year-that could be interesting...........

 

[Papatom]

I have looked at Tom's coverage of both Ryzen and Coffee Lake.
From what I could see, there is a strong preference to show intel products in a better light.

1. Pricing. THG chooses platform/processor alone costs as they see fit; any sane choice would include at least the cooling solution and, perhaps, motherboard into this. Not at THG though; I expect this to change when cheap intel mobos arrive.
2. Cooling. You guys here seem to use good cooling (as in: expensive and performing), yet you do not inculde the costs into the equation. Combined with better thermals of competitive AMD offerings, this is smelly.
3. Hand-picked benchmarks - blender and lux-what? included, POV-ray and Cinebench excluded. You guessed it, AMD hardware tends to have upper hand in the latter (excluded ones).
4. Looking thru fingers at contest rules. RX 480 was over the TDP, you named and shamed it. i9s were over their respective TDPs, you failed to clearly state so.

I no longer find this website serving just information for me to make educated purchasing decision.
I am fed up. From a regular reader back in '90 thru early this decade, I decided to leave this site for good.

 

[mapesdhs]

"... the 8400's multi-core Turbo Boost bins get a lot more aggressive than Core i5-7400's."

Just curious, if the CPU can handle all six cores at 3.8 anyway, why is the base clock so much lower? Why not just set the base clock to 3.8 anyway? Or is it because it could only sustain 3.6 if the cooler was up to the task? If so, then Turbo has morphed into a sort of pseudo-Intel-sanctioned auto oc'ing.

Also, since the 6-core turbo level starts off from a higher frequency, the percentage gain of the higher levels vs. the 6-core turbo level is lower than the inter-level gains for the old 4-core, which is if anything IMO somewhat less imppressive.

Ian.

 

[TJ Hooker]

I have looked at Tom's coverage of both Ryzen and Coffee Lake.
From what I could see, there is a strong preference to show intel products in a better light.

1. Pricing. THG chooses platform/processor alone costs as they see fit; any sane choice would include at least the cooling solution and, perhaps, motherboard into this. Not at THG though; I expect this to change when cheap intel mobos arrive.
2. Cooling. You guys here seem to use good cooling (as in: expensive and performing), yet you do not inculde the costs into the equation. Combined with better thermals of competitive AMD offerings, this is smelly.
3. Hand-picked benchmarks - blender and lux-what? included, POV-ray and Cinebench excluded. You guessed it, AMD hardware tends to have upper hand in the latter (excluded ones).
4. Looking thru fingers at contest rules. RX 480 was over the TDP, you named and shamed it. i9s were over their respective TDPs, you failed to clearly state so.

I no longer find this website serving just information for me to make educated purchasing decision.
I am fed up. From a regular reader back in '90 thru early this decade, I decided to leave this site for good.

1. They addressed the fact that you'll need to pair the 8400 with a more expensive Z370 mobo for the time being in the first page, as well as in the conclusion.
2. Ok this I kind of agree on. In the review they use a massive 420mm rad AIO cooler or a compressor cooler water loop, but then conclude that a basic air cooler is fine for cooling, and don't address stock cooler performance at all.
3. AMD performs just fine in the benchmarks they did pick though. The 1600X outperforms the 8400 (and usually the 8600K) in both Luxrender and Blender.
4. They called out the RX 480 for violating the PCIe spec on the max current/power that could be drawn from the motherboard slot (and even then they said this probably wouldn't be an issue 99% of the time, IIRC). That's not the same as going over TDP. TDP was never supposed to be an accurate measure of power consumption anyway. Apples to oranges.

 

[TMTOWTSAC]

I have looked at Tom's coverage of both Ryzen and Coffee Lake.
From what I could see, there is a strong preference to show intel products in a better light.

1. Pricing. THG chooses platform/processor alone costs as they see fit; any sane choice would include at least the cooling solution and, perhaps, motherboard into this. Not at THG though; I expect this to change when cheap intel mobos arrive.
2. Cooling. You guys here seem to use good cooling (as in: expensive and performing), yet you do not inculde the costs into the equation. Combined with better thermals of competitive AMD offerings, this is smelly.
3. Hand-picked benchmarks - blender and lux-what? included, POV-ray and Cinebench excluded. You guessed it, AMD hardware tends to have upper hand in the latter (excluded ones).
4. Looking thru fingers at contest rules. RX 480 was over the TDP, you named and shamed it. i9s were over their respective TDPs, you failed to clearly state so.

I no longer find this website serving just information for me to make educated purchasing decision.
I am fed up. From a regular reader back in '90 thru early this decade, I decided to leave this site for good.

Core i5-8400 drops into a LGA 1551 interface, but as we know from our previous Coffee Lake reviews, it isn't backward compatible with 200-series motherboards. That means you'll have to step up to a 300-series motherboard, even though that means getting no new features in the process. Worse, a rushed launch means the Z370 chipset is your only option for now. Cheaper B- and H-series motherboards, which will cost a lot less, arrive at the beginning of next year. The Core i5-8400 is a locked processor anyway, so the reason most enthusiasts would have had for spending extra (overclocking) doesn't apply here.

Back when they were still referring to Ryzen CPU's by the architecture name, Zen, AMD relentlessly promoted Blender benchmarks. Everywhere, during IDC, Computex, during launch, and live onstage and in posted videos. At the Threadripper announcement, even during gaming demos.

And the big issue with the 480 was that it violated the PCIe spec by drawing more than 75 watts avg from the slot, not that it exceeded TDP. It wasn't the first card to do it. Other cards were also found to be out of spec including cards from Nvidia, and while not to the same degree, it was the 480 findings that brought the issue to light so it caught the most flak.

Not everything is some sort of grand conspiracy.

Edit: Mr Shatner types (or thinks) way way, faster than I do.

 

[mapesdhs]

1. They addressed the fact that you'll need to pair the 8400 with a more expensive Z370 mobo for the time being in the first page, as well as in the conclusion.

And yet the article states six times that the 1600X is more expensive, adding it as a caveat almost every time the CPU is mentioned ("pricier" seems to be the favoured word; ok, we get the point!), but this difference is completely wiped out by the more expensive mbd cost for the i5. As long as this is the case, it's a bit misleading to allow the reader to conclude the Ryzen platform will cost more overall just because the CPU costs more. I was also rather surprised at how little power the 1600X uses compared to the 8400, even when overclocked - another notch in AMD's favour.

I'm also tired of seeing gaming conclusions based on 1080p results where frame rates are already well over 100fps, it's meaningless. Reviews already use 1080p in order to more fully emphasise CPU differences, using powerful GPUs to remove any bottleneck, GPUs of a type that are unlikely to be paired in reality with this class of CPU. This allows the tests to show differences which are not representative of what real users will experience. Try the test again with a P600 or, at most, a P4000, see what happens. Ditto a more realistic GPU.

Have a look at older reviews from back in the X58 days, articles would wow if frame rates could reach 60Hz. I know high frequency monitors are a thing these days, but that really is a minority audience, especially for those buying into this class of hardware in the first place.

Ian.

 

[TMTOWTSAC]

I think of 1080p results a little like car specs. Every manufacturer lists the top speed, even if there's no realistic chance you will ever be able, or allowed to go that fast. But if they replaced it with a simple 70mph yes/no checkbox it would be leaving rather a lot of information out. In a perfect world sure they'd test every GPU with every CPU and you'd know exactly what configuration will suit your needs. Short of that, showing where each component tops out and becomes a limiting factor seems like the best compromise.

 

[PaulAlcorn]

I have looked at Tom's coverage of both Ryzen and Coffee Lake.
From what I could see, there is a strong preference to show intel products in a better light.

1. Pricing. THG chooses platform/processor alone costs as they see fit; any sane choice would include at least the cooling solution and, perhaps, motherboard into this. Not at THG though; I expect this to change when cheap intel mobos arrive.
2. Cooling. You guys here seem to use good cooling (as in: expensive and performing), yet you do not inculde the costs into the equation. Combined with better thermals of competitive AMD offerings, this is smelly.
3. Hand-picked benchmarks - blender and lux-what? included, POV-ray and Cinebench excluded. You guessed it, AMD hardware tends to have upper hand in the latter (excluded ones).
4. Looking thru fingers at contest rules. RX 480 was over the TDP, you named and shamed it. i9s were over their respective TDPs, you failed to clearly state so.

I no longer find this website serving just information for me to make educated purchasing decision.
I am fed up. From a regular reader back in '90 thru early this decade, I decided to leave this site for good.

1. We mention the cost of motherboards and coolers in the text, In addition, we have two dedicated charts in the conclusion that specifically list performance and platform costs combined. That includes price adjustments for motherboards and coolers.

2. see response 1.

3. we have plenty of benchmarks that are in that same class. Luxrender and Blender (Multiple blender tests) that almost always go in AMD's favor. And we also include those benchmarks in the price v. performance breakouts at the end of the article.