AMD fanboys.

[KingFox]

Just now, Benji_w said:

Actually no it wasn't, the OP isn't a rant against AMD fanboys so yeah...

I was just about to say the same thing...i definately didnt see the OP as a rant at all lol

[Bhav]

13 minutes ago, Dabombinable said:

Lesson 1 for overclocking Haswell CPU. Never run stress testers that use AVX 2.0 instructions on overclocked Haswell CPU. That's what gave Prime95 its reputation for killing CPU. And you should know that no matter which method gets used to delid a CPU, your still fucking with it and risk destroying it in the process. 4.4GHz BTW at 1.26V stock with the Seidon 120V (which can just handle my Xeon X5450 at 1.45V 4.4GHz, and at 3GHz it consumes more than my 4790K when its boosting all cores), running Prime 95 disables turbo and has the voltage at 1.19. 24/7 stable BTW is a weak of non stop F@H with the graphics cards in use for folding as well.

Can I get more info on this please? I was using Prime95 to test mine, and maybe that's why I wasn't getting good results. What should I use instead?

[HugeMadKitten]

I had legendary AMD Phenom II x4 965 Black Edition and i think its one of the best AMD CPU's, but it couldn't hold up with modern titels so i changed to intel platform, and im now running i5 4690k

[KingFox]

Can I get more info on this please? I was using Prime95 to test mine, and maybe that's why I wasn't getting good results. What should I use instead?

This issue was with Skylake as well. Prime95 would freeze the CPU during testing. Because of the AVX instruction set in Haswell, it would apply extra voltage on top of your overclock, causing overvoltage to your CPU, which would lead to damage and overheating. The voltage boost because of AVX was 0.1 to 0.2, so adding extra voltage via bios on top of this would cause some serious issues.

[Benji_w]

2 minutes ago, KingFox said:

This issue was with Skylake as well. Prime95 would freeze the CPU during testing. Because of the AVX instruction set in Haswell, it would apply extra voltage on top of your overclock, causing overvoltage to your CPU, which would lead to damage and overheating. The voltage boost because of AVX was 0.1 to 0.2, so adding extra voltage via bios on top of this would cause some serious issues.

Prime would freeze up skylake because of a code error or something, and it added too much voltage to the chips, it's soon been patched.

[Dabombinable]

20 minutes ago, Benji_w said:

Lesson 1 - Incorrect after updates

Lesson 2 - Delid tools are 100% safe

Lesson 3 - Changing your data to try an prove someone wrong like a child makes you look ignorant

Lesson 4 - Always test your cooling, setup and so on before overclocking, if it isn't stable stock, something is wrong

 

Go back to school kid.

 

Lol, changing what data?

[ybriK]

On 5/3/2016 at 9:23 AM, Prysin said:

Single core performance isnt alpha omega anymore. More and more software is multi-threaded. Perhaps not optimized for multiple threads, but it USES multiple threads. This means FX will suck much less then synthetic benchmarks suggest.

How come a 6700K is beating a 5960K? Shouldn't it be the other way around? Seems to me single core performance is still very relevant for modern games. Unless DX12 can truly change the Bulldozer architecture in the future.

 

 

 

[Bhav]

23 minutes ago, KingFox said:

This issue was with Skylake as well. Prime95 would freeze the CPU during testing. Because of the AVX instruction set in Haswell, it would apply extra voltage on top of your overclock, causing overvoltage to your CPU, which would lead to damage and overheating. The voltage boost because of AVX was 0.1 to 0.2, so adding extra voltage via bios on top of this would cause some serious issues.

OOOOOOOOOOPPPPPPPPPPPPPPPPPPPSSSSSSSSSSSSSSSSSSS

 

So my first CPU likely wasnt bad, however I was having lock ups and freezing in Windows at stock frequencies which is also why I sent it back. Doh. But my new one is already more stable in Prime95, so it should be better. I'm downloading ROG Realbench instead now.

[Benji_w]

6 minutes ago, Bhav said:

OOOOOOOOOOPPPPPPPPPPPPPPPPPPPSSSSSSSSSSSSSSSSSSS

 

So my first CPU likely wasnt bad, however I was having lock ups and freezing in Windows at stock frequencies which is also why I sent it back. Doh. But my new one is already more stable in Prime95, so it should be better. I'm downloading ROG Realbench instead now.

Realbench isn't worth the time, XTU is the same. OCCT, Prime, IBT AVX or AIDA any of those will help find stability.

[Kinda Bottlenecked]

2 hours ago, Dabombinable said:

Your FX8350 isn't an i5. And your obviously doing something wrong because you should easily be running that i5 4670K that you supposedly have at 4.8GHz, and watching as your 5GHz FX8350 shows that it isn't that much faster multi threaded with the i5 at 4.8GHz, and slower single threaded than if your i5 4670K was at 3.3GHz.

 

Wow, I would love to get 4.8Ghz out of my 4670K. Fucker can't even stay stable at 4Ghz 1.3v. The 4670k isn't a god sent. At least not anymore. Benji's 5Ghz FX will out perform my CPu in certain benchmarks that's for sure. And if he isn't impressed with the 4670K. That's pretty understandable. 

[harrynowl]

9 minutes ago, Pohernori said:

 

Wow, I would love to get 4.8Ghz out of my 4670K. Fucker can't even stay stable at 4Ghz 1.3v. The 4670k isn't a god sent. At least not anymore. Benji's 5Ghz FX will out perform my CPu in certain benchmarks that's for sure. And if he isn't impressed with the 4670K. That's pretty understandable. 

That's the kind of chip people blow up and send back

 

@Benji_wI'm lazy and just run a pass of the XTU benchmark.. that thing is brutal (and is also based on prime, the stress is a weaker version of the linpack test). If it passes that odds are it's pretty much solid

[Benji_w]

2 minutes ago, harrynowl said:

That's the kind of chip people blow up and send back

 

@Benji_wI'm lazy and just run a pass of the XTU benchmark.. that thing is brutal (and is also based on prime, the stress is a weaker version of the linpack test). If it passes that odds are it's pretty much solid

I know someone who swears by XTU but his pc constantly blue screens, also know a few intel clockers who say XTU isn't worth the energy, I dunno, I'll give it a try but don't see the point when IBT AVX is plenty quick and taxing.

[harrynowl]

1 minute ago, Benji_w said:

I know someone who swears by XTU but his pc constantly blue screens, also know a few intel clockers who say XTU isn't worth the energy, I dunno, I'll give it a try but don't see the point when IBT AVX is plenty quick and taxing.

I just run 5ghz/1.375v daily without problems which is where it will pass a full run of xtu benchmark

 

the xtu benchmark isn't worth the energy as a benchmark but it's good for lazy people to test cpu bins

[Bhav]

31 minutes ago, Benji_w said:

Realbench isn't worth the time, XTU is the same. OCCT, Prime, IBT AVX or AIDA any of those will help find stability.

You mean XTU also isnt worth the time, or it is a suitable equivilent?

 

I'll try OCCT for now.

[Benji_w]

4 minutes ago, harrynowl said:

I just run 5ghz/1.375v daily without problems which is where it will pass a full run of xtu benchmark

 

the xtu benchmark isn't worth the energy as a benchmark but it's good for lazy people to test cpu bins

Lol fair enough, I'll give it a whirl regardless. I hope this 4670k hits 5ghz and is stable.

[harrynowl]

Just now, Benji_w said:

Lol fair enough, I'll give it a whirl regardless. I hope this 4670k hits 5ghz and is stable.

only 1 46X0K i tested did 5ghz on ambient

[Benji_w]

Just now, harrynowl said:

only 1 46X0K i tested did 5ghz on ambient

[harrynowl]

Just now, Benji_w said:

I approve

 

Delidded an e5200 and ran direct die single stage.. -42c 1.65v

[Benji_w]

Just now, harrynowl said:

I approve

 

Delidded an e5200 and ran direct die single stage.. -42c 1.65v

I love it when you talk dirty.

[Dabombinable]

28 minutes ago, Pohernori said:

 

Wow, I would love to get 4.8Ghz out of my 4670K. Fucker can't even stay stable at 4Ghz 1.3v. The 4670k isn't a god sent. At least not anymore. Benji's 5Ghz FX will out perform my CPu in certain benchmarks that's for sure. And if he isn't impressed with the 4670K. That's pretty understandable. 

Ok. Now I understand. The 4670K is shit compared to the refreshed versions. Now what @Benji_w was saying makes more sense. I can definitely understand why its a disappointment-them being that bad at overclocking would kind of make me feel ripped off.

5 minutes ago, Benji_w said:

Lol fair enough, I'll give it a whirl regardless. I hope this 4670k hits 5ghz and is stable.

It should as long as you don't meet the under-water voltage wall. Getting my 4790K above 4.8GHz requires a lot more voltage than I'm comfortable with pumping through it-I don't want to be forced to fall back to my i5 4440.

[harrynowl]

3 minutes ago, Benji_w said:

I love it when you talk dirty.

 

my 8320 makes me feel a bit sad cause it gets bagged by my core2s at a lower speed but it was good fun for frequency

 

[Benji_w]

2 minutes ago, harrynowl said:

 

my 8320 makes me feel a bit sad cause it gets bagged by my core2s at a lower speed but it was good fun for frequency

 

Marry me? 

[AcidicStylez]

Titling this topic "AMD Fanboys" probably was not the best way to go about this.

[Kinda Bottlenecked]

25 minutes ago, harrynowl said:

That's the kind of chip people blow up and send back

 

My 4670K?

[harrynowl]

Just now, Pohernori said:

 

My 4670K?

Yes. Poo! I know people that blow up chips that do 4.6GHz at 1.325V...

 

4 minutes ago, Benji_w said:

Marry me? 

Do I get to borrow a board?