Competitor rules

DarrenM343 · 6 Jan 2017 at 12:36

Disco_P said:
Every card and every game. Clock for clock the 6700K and 7700K are identical.

The max overclock is a couple of hundred mhz more and it's about £25 more. Doing the math and it's about an 8% performance improvement for a 7% cost increase. (assuming 4.6 sky vs 5 kaby)

Rubbish

. There will be some variation. Nobody has actually tested all the games or cards. Don't make assumptions and don't believe too much what you read/ hear. It may all be right but I suspect it may not all be. FarCry 4 for example might be 30% faster on 6700 than 3770 at same clock speed so I reckon with a top end card in some games there will be a decent difference. I said might as it may be wrong but we dont know.

But point is I'd rather pick up the 7700k right now if I was in the market for a 6700k level processor.

Legend · 6 Jan 2017 at 12:38

DarrenM343 said:
Rubbish . There will be some variation. Nobody has actually tested all the games or cards. Don't make assumptions and don't believe too much what you read/ hear. It may all be right but I suspect it may not all be. FarCry 4 for example might be 30% faster on 6700 than 3770 at same clock speed so I reckon with a top end card in some games there will be a decent difference. I said might as it may be wrong but we dont know.

Are you referring to the 6700k vs 7700k? There is no assumption, it's a scientific fact that there is no IPC gain between them, performance in real world is virtually identical. Every benchmark proves this. Variation yes, in some cases, but it's so small as to be within the margin of error.

Martini1991 · 6 Jan 2017 at 12:53

Like I say, Darren is the other end
Kaby lake is more akin to a stepping revision than a new CPU.

Orangey · 6 Jan 2017 at 12:56

Acme said:
I think we need a technological breakthrough in the way the chips are manufactured before we see any massive gains like years ago...

They will switch from silicon in 2019/2020 or so. One big upshot is that the new materials are capable of much faster frequencies. I'm sure Intel and their **** shareholders will keep putting it off and only use it to provide a meagre bump at first though.

DarrenM343 · 6 Jan 2017 at 13:09

Martini1991 said:
Like I say, Darren is the other end
Kaby lake is more akin to a stepping revision than a new CPU.

.
No, as I said in above I may be wrong but not so willing to make assumptions based on a few synthetic benchmarks I've read about on the internet.

Martini1991 · 6 Jan 2017 at 13:10

Synthetic benchmarks are going to be the things that show the performance differences

DarrenM343 · 6 Jan 2017 at 13:21

Martini1991 said:
Synthetic benchmarks are going to be the things that show the performance differences

No they're not. I work in hardware and software and see all sorts of surprising events in real usage.

And Nvidia and AMD both could for example optimise their cards for GPU benchmarks. GPU benchmarks are scripted events that happen in sequence. They also dont run for long either.
CPU only benchmarks kind of fine as it's testing a single component but the overall integration of the hardware, ie, hardware and GPU working together is more difficult as there's so much variation. I'd like to see a benchmark of a game or two that is known to be affected by the processor and see what the result is.

I'm not denying I'm not wrong though but I'd only believe in my own testing of products personally, knowing I've done it properly. Some of the review sites have their own agenda's and probably don't even do it that scientifically in some cases.

ultimately a new product even with small changes is a good evolution of something that was designed 1.5-2 years ago now will simply be better all round with all the small improvements

Yaayuh! · 6 Jan 2017 at 13:38

Still unsure on what the "

" is for.

Legend · 6 Jan 2017 at 13:43

DarrenM343 said:
.
No, as I said in above I may be wrong but not so willing to make assumptions based on a few synthetic benchmarks I've read about on the internet.

What are you talking about? The 6700k and 7700k have been put up against each other in real world game tests by loads of YouTubers and tech/gaming sites/blogs. Nothing synthetic or presumptive about that! They offer virtually identical performance.

Jumper118 · 6 Jan 2017 at 13:52

DarrenM343 said:
No they're not. I work in hardware and software and see all sorts of surprising events in real usage.

And Nvidia and AMD both could for example optimise their cards for GPU benchmarks. GPU benchmarks are scripted events that happen in sequence. They also dont run for long either.
CPU only benchmarks kind of fine as it's testing a single component but the overall integration of the hardware, ie, hardware and GPU working together is more difficult as there's so much variation. I'd like to see a benchmark of a game or two that is known to be affected by the processor and see what the result is.

I'm not denying I'm not wrong though but I'd only believe in my own testing of products personally, knowing I've done it properly. Some of the review sites have their own agenda's and probably don't even do it that scientifically in some cases.

ultimately a new product even with small changes is a good evolution of something that was designed 1.5-2 years ago now will simply be better all round with all the small improvements

the whole point of a benchmark being the same preset stuff is so its consistent. real world usage is never consistent. you cant have both real world usage and a benchmark. you just have to run a decent set of benchmarks which test a wide range of features.

Nutella33 · 6 Jan 2017 at 14:08

Rossi~ said:
Still unsure on what the "" is for.

it was for kaby lake.

5000 units and then a product launch a few days later.

Yaayuh! · 6 Jan 2017 at 14:09

Nutella33 said:
it was for kaby lake.

5000 units and then a product launch a few days later.

Severe disappointment then.

Nutella33 · 6 Jan 2017 at 14:12

Rossi~ said:
Severe disappointment then.

Theme of the week it seems

Uriel · 6 Jan 2017 at 14:15

I'll take what the

was for when Gibbo tell us.

I want to hear about the Ryzen PC he's presumably been demonstrating at CES.

Aa83 · 6 Jan 2017 at 14:27

Nutella33 said:
Theme of the week it seems

Even if that's the theme of 2017, I'll take that over 2016

Disco_P · 6 Jan 2017 at 14:48

Martini1991 said:
Like I say, Darren is the other end

Seems about right.

humbug · 6 Jan 2017 at 14:53

Do people still think the majority of these chips will do 5Ghz and with an average cost cooler?

DarrenM343 · 6 Jan 2017 at 14:55

Jumper118 said:
the whole point of a benchmark being the same preset stuff is so its consistent. real world usage is never consistent. you cant have both real world usage and a benchmark. you just have to run a decent set of benchmarks which test a wide range of features.

And thats why sonetimes the real feel of a game doesn't always represent what a bench shows.

Agree with doing wide set of benchmarks

opethdisciple · 6 Jan 2017 at 15:04

Rossi~ said:
Are there 7700K vs 6700K benchmarks (non-gaming) at same clocks?

Anandtech.

Hambut_Bulge · 6 Jan 2017 at 16:01

Uriel said:
I'll take what the was for when Gibbo tell us.

I want to hear about the Ryzen PC he's presumably been demonstrating at CES.

Second that!

Competitor rules

;)