[RUMOUR] AMD's Zen CPU rumoured to compete with high end intel for $300

[Space Reptile]

23 minutes ago, patrickjp93 said:

Define still shit. Not only is that 1.05^5 = 1.27628 or 27.628% higher IPC than Sandy Bridge (5% per generation on average), but the clocks are also 20% higher. Then there's the fact you get DDR4, AVX2, FMA3, a much more powerful iGPU media engine, and many other capabilities that older chips just plain lack.

for a matter of fact I own a 2600 here , and well lets have a look shall we ?

 

http://cpu.userbenchmark.com/Compare/Intel-Core-i7-2600-vs-Intel-Core-i7-6700/620vs3515

 

apples to apples , same clock speeds same everything and 20~25% improvement across the board

 

that's neat for 5ish generations if you add things like

"DDR4, AVX2, FMA3, a much more powerful iGPU media engine, and many other capabilities "

 

but from skylake to kaby you get LESS than 1% and NOTHING NEW

totally worth the ~400$ admission price for a new chip amiright?

[patrickjp93]

21 minutes ago, MageTank said:

So now you are making an assumption of their testing methodology, with no way to prove it, and are throwing out a random number that in and of itself, cannot be confirmed? Why do you insist on using this strategy against me every single time we speak? Your entire defense is "You can't prove me wrong", which is sadly true because I lack the ability to travel to this fairy tale realm to obtain the same information you have. 

 

Let's humor your logic for a second though. That they ran it 12 times before submitting a score. Why did they submit two scores? 

https://browser.primatelabs.com/v4/cpu/583275

https://browser.primatelabs.com/v4/cpu/583064

 

24 minute difference between the submission of these two scores. Perfectly capable of running the test 12 times in 24 minutes, I won't argue that, but what purpose does it serve? They are basically the exact same scores, with nothing changed. If they had the information gathered from 12 offline runs, why even bother posting online again, after running another 12 runs, if the score is so similar? 

Memory tweak, windows setting changed. You know how benchmark people can be.

[patrickjp93]

2 minutes ago, Space Reptile said:

for a matter of fact I own a 2600 here , and well lets have a look shall we ?

 

http://cpu.userbenchmark.com/Compare/Intel-Core-i7-2600-vs-Intel-Core-i7-6700/620vs3515

 

apples to apples , same clock speeds same everything and 20~25% improvement across the board

 

that's neat for 5ish generations if you add things like

"DDR4, AVX2, FMA3, a much more powerful iGPU media engine, and many other capabilities "

 

but from skylake to kaby you get LESS than 1% and NOTHING NEW

totally worth the ~400$ admission price for a new chip amiright?

No one said you have to, nor should, upgrade every generation. That's just dumb as far behind as code optimization has gotten in the consumer world.

[Dackzy]

@MageTank Oh god I love it when you two argue. 

[Space Reptile]

7 minutes ago, patrickjp93 said:

Memory tweak, windows setting changed. You know how benchmark people can be.

so they would make the new hot chip look shit on purpose?

 

0_o

[patrickjp93]

5 minutes ago, Space Reptile said:

so they would make the new hot chip look shit on purpose?

 

0_o

No, just squeeze as much out of it as possible.

[Space Reptile]

Just now, patrickjp93 said:

No, just squeeze as much out of it as possible.

so the 0.7~0.9% is a perfect score?

[patrickjp93]

Just now, Space Reptile said:

so the 0.7~0.9% is a perfect score?

There's still more to do. Let the professional overclockers work their magic after release. overclock.net and luck always produce top quality results.

[Space Reptile]

2 minutes ago, patrickjp93 said:

There's still more to do. Let the professional overclockers work their magic after release. overclock.net and luck always produce top quality results.

so don't judge what we have rumored until we actually have it on the shelves ,

goes for kaby as much as for zen

 

thank you

[patrickjp93]

1 hour ago, Space Reptile said:

so don't judge what we have rumored until we actually have it on the shelves ,

goes for kaby as much as for zen

 

thank you

Nope, because KBL has verified benchmarks. Zen only has the Dresdenboy AOTS tests which showed Zen fairly far behind Haswell at the same clocks and core count.

[Space Reptile]

1 hour ago, patrickjp93 said:

Nope, because KBL has verified benchmarks. Zen only has the Dresdenboy AOTS tests which showed Zen fairly far behind Haswell at the same clocks and core count.

oh so VERTIFIED BENCHMARKS vs some guys test who MAYBE has early silicone is a FAIR comparison

 

ಠ_ಠ

[patrickjp93]

24 minutes ago, Space Reptile said:

oh so VERTIFIED BENCHMARKS vs some guys test who MAYBE has early silicone is a FAIR comparison

 

ಠ_ಠ

Dresdenboy has never once been inaccurate in his 6 years of reporting AMD leaks. Not, once.

[TotaLInsanity]

9 hours ago, djdwosk97 said:

There's literally zero chance that AMD will launch something that competes with a 6900k for $300. AMD has always priced their CPUs competitively. So if it performs like a 6900k, then it will be priced like a 6900k (maybe 10% less).

its still fun to dream about it even if it will probobly never happen.

[Prysin]

bullshit rumor is total bullshit

[Prysin]

3 hours ago, patrickjp93 said:

Memory tweak, windows setting changed. You know how benchmark people can be.

weak argument patrick. Try again?

[patrickjp93]

1 hour ago, Prysin said:

weak argument patrick. Try again?

I laid out the likely possibilities. Overclocking enthusiasts tweak everything and run again and again until they're blue in the face. It's not a weak argument. It's just a short, valid one.

[Space Reptile]

4 minutes ago, patrickjp93 said:

I laid out the likely possibilities. Overclocking enthusiasts tweak everything and run again and again until they're blue in the face. It's not a weak argument. It's just a short, valid one.

can you admit that you are wrong once and move on? I know that nobody wants to do that but for the sake of everyone and your reputation please do so

[Space Reptile]

-accidental double post-

[That Norwegian Guy]

4 minutes ago, Space Reptile said:

can you admit

There's an ignore function... For your sanity...

Any AMD thread tends to bring out everyone on mine.

[patrickjp93]

9 minutes ago, Space Reptile said:

can you admit that you are wrong once and move on? I know that nobody wants to do that but for the sake of everyone and your reputation please do so

Except I wasn't, and no one has even come close to proving otherwise.

[Space Reptile]

Just now, patrickjp93 said:

Except I wasn't, and no one has even come close to proving otherwise.

can you read?

[Space Reptile]

6 minutes ago, That Norwegian Guy said:

There's an ignore function... For your sanity...

I tempt to not ignore/block/ban people until they get hostile or are just plain assholes , give everyone a fair chance

[Prysin]

17 minutes ago, Space Reptile said:

can you admit that you are wrong once and move on? I know that nobody wants to do that but for the sake of everyone and your reputation please do so

ive proven him totally wrong 4 times this year, and partially 2 more... he claims he is never wrong, yet he is at 4-6 completely provable faults already

[Prysin]

8 minutes ago, patrickjp93 said:

Except I wasn't, and no one has even come close to proving otherwise.

you were vague enough, and basing your argument purely on circumstances well outside your own power of influence. You cannot prove or disprove even your own thesis. Thus it is that, a subjective opinion without any more merit then anyone else.

The only logical answer here is to admit neutrality, rather then claim to be positive in ones claim.

[patrickjp93]

1 hour ago, Space Reptile said:

can you read?

I can. Care to explain what your standard of proof is, because it certainly wouldn't hold up to any decent logician.