Core i3 7350K Overclockable Dual Core Review

[NumLock21]

Is the benchmarks for the i3 7350K overclocked? Because Luke mentioned it being at 5GHz, when its stock clock is actually 4.2GHz and it seems there is no Turbo on the i3.

http://ark.intel.com/compare/97527,97147

 

[Drak3]

38 minutes ago, NumLock21 said:

Is the benchmarks for the i3 7350K overclocked? Because Luke mentioned it being at 5GHz, when its stock clock is actually 4.2GHz and it seems there is no Turbo on the i3.

http://ark.intel.com/compare/97527,97147

 

Yes, he overclocked it, because according to every big name in overclocking ever, there is no reason to not overclock a K sku processor.

[MageTank]

33 minutes ago, Drak3 said:

Yes, he overclocked it, because according to every big name in overclocking ever, there is no reason to not overclock a K sku processor.

Here I sit, with a stock clock, delidded, undervolted 6700k, lol. 

[Energycore]

22 minutes ago, MageTank said:

Here I sit, with a stock clock, delidded, undervolted 6700k, lol. 

Priorities, I could run my GTX 560 Ti at 1GHz but I like silence so I run it at stock clocks + undervolt

[Deli]

Why everyone is hating the video thumbnail? I think it's just fine.

[vanished]

4 minutes ago, Deli said:

Why everyone is hating the video thumbnail? I think it's just fine.

Yeah, if anything we should be giving them points for actually using a picture of the 7350k and not just any random CPU... it's not like most people would even notice

[Quad Damage]

Please learn how to do proper game performance testing.

If you're testing a CPU, you need to remove any possibility of a GPU bottleneck.

That means running at the lowest resolution available, not the highest - or at least something low like 720p.

 

Your testing methodology is completely flawed if your conclusion from testing Deus Ex: Mankind Divided was that it isn't a useful game to test with.

That game's performance is hugely affected by the CPU in many locations. (though possibly not the built-in benchmark)

It will put 100% load on all available CPU cores and you can swap out the GPU without seeing any performance change at all - you just see the GPU load drop when you use a faster GPU, which makes it a pure CPU test.

 

And it's demanding enough that even the fastest CPUs are all going to test well below 100 FPS in the demanding areas of the game, so it's not like testing older games where one CPU might run the game at 500 FPS and another at 550 FPS, which doesn't really tell us anything useful.

[Granular]

Yeah, considering the price, it's nothing special.

Figure in the price difference between H270 and Z270 boards and adequate coolers and it's more appropriate comparing it to the i5 7600, which will boost to 4.1 GHz (17% higher than the i5 7400).

[iPolymer]

My 4790K is at 4.36 ghz, so can my old tech not hold up to the 7350K? I convert a lot of videos to .mp4 format and it take 80% of my CPU usage, is it about speed or cores for this? 

[Issak]

Does the i3 actually make sense if the locked i5 can match its performance? You have to buy the Z270 motherboard and a cooler for the i3, while i5 comes with a stock cooler and you can stick it on a much cheaper motherboard.

 

i3 7350K: 195€

cheapest Z270 mobo: 120€

cooler: at least 30€

Total: 345€+

 

i5 7400 (3GHz): 187€

cheapest B250 mobo: 70€

cooler: included

total: 257€

You could get 7500 (3.4GHz, 212€; total: 282€

or 7600 (3.5GHz, 230€; total: 300€) and it would still be cheaper.

 

I know i3 was better in benchmarks, but in games it was a frame or two difference, really worth extra money? I thing a better build then this i3 would be the i5 and a 100€ better GPU.

 

Yeah, I agree the test should not be done (only) with 4K, and jay didn't compare it to locked i5. I think locked i5 is still a better deal then unlocked i3.

 

Edited January 15, 2017 by Issak
edit

[xAcid9]

Fun to see everyone complaining like LTT will give a damn. They still use average FPS as gaming performance metric even when lots of us complaint about that 2-3 years ago.

That video will sell a lot of 7350k and thats all they care. (or Intel care)

[Simon771]

They should make comparison between i3 7350k vs i5 7400 ... but this time also select motherboard that is rewuired for any of those CPUs.

H270 or B270 are not that expencive, so I'm pretty sure i5 7400 will win in price/performance ratio.

 

And that i3 also need AIO cooler or any other aftermarket. i5 7400 is just fine with stock

[NumLock21]

7 hours ago, Drak3 said:

Yes, he overclocked it, because according to every big name in overclocking ever, there is no reason to not overclock a K sku processor.

Then the benchmarks are flawed. If that i3 is at 5ghz, while i5 is at stock. It gives a false impression the i3 is faster, while in reality it's probably not as fast as the benchmarks says it is. The benchmark needs to show the i3, both at stock and oc with comparison to the core i5, running at stock speed.

 

E.g.

Cb15 score 

I3 stock

I3 oc

I5 stock

[Cakebring3r]

I wonder if they are even reading the comments about there dodgy reviews and click bait tactics. I dont know whats worse, them listening to our feedback and not caring or them not reading it at all. If they keep this up they might get some more casual views but will end up loosing a lot of there core hard core audience. I didnt watch any ces videos simply because they were click bait piles of trash and i watch their vlogs for god sake. I really hope they at least address this stuff.

Spoiler

(and secretly that all the videos they did this for do badly so they are forced to stop)

 

[turbo pascal]

For a review of a new i3, I would at least expect you to compare it to previous i3s (4370 and 6300, for example), not to mention actual i5s (come on, I know you have a 6400 or a 4440 at hand or could easily acquire one) and the likes of FX-8350 (almost exactly the price of this i3) and A10-7890K. Also, as other people are saying, benchmarking CPUs in GPU-heavy workloads isn't the way to go and you guys know it very well. And e-sports games? People probably wanna know if this CPU can get them over 200 fps in CS:GO, but nah, you can't test that while literally having a Global Elite (Ed) and piles of all sorts of GPUs in your office... I'm really disappointed.

[othertomperson]

"Can a dual core still game?" Considering I've heard Linus perpetuate the complete rubbish that some games won't launch on an i3 before and for this reason recommended the FX 6300 which does worse than the i3 in multithreaded uses, I don't put a whole lot of faith in these reviews. For a tech channel as prominent as this to not know the difference between the Pentium and i3, and apparently not even understand what hyperthreading is or the impact it has is frankly pathetic.

 

Bullshit like this is the reason people think you need a £3000 PC to game at all.

[cataria]

this was one of your worst videos ever and that means sth. do you not understand that a 7700k with forced clock and boost and disabled HYPER THREADING! is not a i5 7400. what is wrong with u??? why don't u ask intel for this shit chip, i'm sure they give u one. don't believe me? alright here are 2 websites with benchmarks of r15 single and multicore of an i5 7400 http://www.technikaffe.de/cpu_vergleich-intel_core_i5_2500k-5-vs-intel_core_i5_7400-670 http://www.cpu-monkey.com/en/compare_cpu-intel_core_i5_7400-670-vs-intel_core_i3_7350k-674 their scaling of the chip is: 3,785, which is what is to be expected of a non hyperthreaded quadcore that boosted in single a bit too. to compare my i5 2500k flat 4.4 ghz scales to: 3.811688 wanna know the scaling of their 7700 without hyperthreading? in the video their "7400": 2.887 ?????? what did u smoke? at ltt how can u make such a shit video, with fucking up the basics of it so hard????? holy shit. don't believe me check out the i5 6400 cinebenches to compare as the 7400 doesn't get benched much it has 3,69 multicore scaling on the random r15 benchmark i found. WHAT DA FUCK??? 7700k with hyperthreaded disabled =/= quadcore without ht!!!!!! unless u fucked sth. else up this is the reason why u got these absurd benchmarks, like wtf??? u can't even do a simple r15 benchmark that,s sth. i can do on my cpu it's not fucking hard, yet u screw this up. don't u do any math??? didn't u check the fucking multicore scaling and notice its 1: 2.887 ????????? while my old oced i5 2500 gets 1: 3811688. first the fucked up vulkan benchmarks now this, all hope is lost to ever get some real results from ltt ever again. please do the math yourself people. also the result of the video were it made right ffs!! is, that the 5 ghz ht dual core doesn't get to the stock quadcore. the quadcore is 6.8 percent faster than the dualcore at 5 ghz!! result: never ever get a dualcore!!!! and never ever get a fucking overpriced dualcore. i shit on this dualcore in cinbenech multicore by 11 percent on an ancient cpu. DO NOT GET THIS DUALCORE!!! look at real benchmarks see if gamernexus does some, they general do proper ones, wait for zen see what they get for cheap or if u really really need intel shit, get a quadcore or ht quadcore. and ltt please learn to do benchmarks!!

[Drak3]

11 hours ago, MageTank said:

Here I sit, with a stock clock, delidded, undervolted 6700k, lol. 

Shhh. Let the big name overclockers have their egos.

 

4 hours ago, NumLock21 said:

Then the benchmarks are flawed. If that i3 is at 5ghz, while i5 is at stock. It gives a false impression the i3 is faster, while in reality it's probably not as fast as the benchmarks says it is. The benchmark needs to show the i3, both at stock and oc with comparison to the core i5, running at stock speed.

 

E.g.

Cb15 score 

I3 stock

I3 oc

I5 stock

It's LinusTechTips, do you really expect them to be thorough with benchmarks and their testing methodologies? It's also why many of use pull benchmarks from multiple sources.

[MageTank]

1 minute ago, Drak3 said:

Shhh. Let the big name overclockers have their egos.

 

It's LinusTechTips, do you really expect them to be thorough with benchmarks and their testing methodologies? It's also why many of use pull benchmarks from multiple sources.

I've been breaking @Lays down in our HW telegram group. Finally he has seen the light, and is pulling some amazing flops on his CPU. Once we are done with him, he just might be the hero this world needs. 

[Lays]

1 minute ago, MageTank said:

I've been breaking @Lays down in our HW telegram group. Finally he has seen the light, and is pulling some amazing flops on his CPU. Once we are done with him, he just might be the hero this world needs. 

just call me king dingaling 

[MageTank]

1 minute ago, Lays said:

just call me king dingaling 

Only once your ram/CPU is stable, lol.

[Drak3]

4 minutes ago, MageTank said:

he just might be the hero this world needs

Unless @Lays' cable management looks like this:

 

He's not the hero this world deserves.

[kaiju_wars]

17 minutes ago, Drak3 said:

Unless @Lays' cable management looks like this:

 

He's not the hero this world deserves.

I.. its so beautiful 

[MageTank]

28 minutes ago, Drak3 said:

Unless @Lays' cable management looks like this:

 

 

He's not the hero this world deserves.

Actually, he has one stray cable at the top of his case for what I assume is the 8 pin EPS, and it bothers me, lol. 

[SteveGrabowski0]

15 hours ago, kaiju_wars said:

Honestly, with the way they've been handling thumbnails, and like others have said, their testing methodology... I'm seriously thinking of unsubscribing from Linus Tech Tips.

I'd probably stay subbed to Tech Quickie and Channel Super Fun.

I still love LTT for all the crazy shit Linus does like whole room water cooling or that beast of a system with the Xeons and all those gpus. I think Linus is first and foremost a showman, and doing those over the top ridiculous builds is where he really shines and separates LTT from all the other PC channels. And then the Scrapyard Wars competitions are really fun too. So I'll never unsubscribe because of that kind of content the channel does so well.