Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Something like this would have a good chance at getting you 5.3 GHz on Kaby Lake:
Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Dude the thing reaches 7z with nitrogen cooling. I'd say somewhere around 5.5 for liquid.
Do you want me to google my uneducated guess for you?
Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Dude the thing reaches 7z with nitrogen cooling. I'd say somewhere around 5.5 for liquid.
Do you want me to google my uneducated guess for you?
There's a big difference between "we got it to boot with all but one core disabled" and something stable enough that you can play games on it and have a normal, stable computer.
Nitrogen also has a boiling point around -196 C, or about 220 C below ambient temperature in a typical room. The difference between temperatures on a typical water cooling versus high end air cooling is what, several degrees, maybe?
Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Dude the thing reaches 7z with nitrogen cooling. I'd say somewhere around 5.5 for liquid.
Do you want me to google my uneducated guess for you?
Yes, i want you to goole your uneducated guess. In fact, you should have that do BEFORE you embarassed yourself once again lol
1/4 cant even hit 4,9 GHz
Almost half of 7700ks cant even hit 5 GHz lol
and 5% can hit 5,2 GHz
0 can hit 5,3 GHz with ANY conventional cooling without frying the chip
and for voltages > 1,3v you need water cooling unless you want it to hit 90c constantly lol
Running them at the same clock speed is not a good comparison. Unless the 7700 is just an overclocked 2600. But when they are running at their normal clock speed then the 7700 is about 50% faster. TBH if you can overclock the 2600 to 4.5hz then you can overclock the 7700 to 5.3hz.
This article kinda just kicks itself in the rear because the test subjects are being altered instead of using factory specifications.
You're ignoring that Sandy Bridge is a legendarily good overclocker. As a fraction of base clock speed, Sandy Bridge can typically be overclocked much, much further than Sky Lake, and almost certainly also Kaby Lake. If you get a 2600K and overclock it to 4.5 GHz, you're probably giving up a quite a bit of speed that you could have had. If you get a 7700K and overclock it to 5.3 GHz, good luck keeping that stable without exotic cooling.
You cant OC 7700k to 5,3 GHz. Even 5 GHz is big lottery lol and yeah thats WITH exotic cooling, on air all you can do if fry your chip (it will throttle anyway)
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
Dude the thing reaches 7z with nitrogen cooling. I'd say somewhere around 5.5 for liquid.
Do you want me to google my uneducated guess for you?
There's a big difference between "we got it to boot with all but one core disabled" and something stable enough that you can play games on it and have a normal, stable computer.
Nitrogen also has a boiling point around -196 C, or about 220 C below ambient temperature in a typical room. The difference between temperatures on a typical water cooling versus high end air cooling is what, several degrees, maybe?
They got 7z with two cores running. Using an older motherboard. And when they got ahold of newer motherboards it ran with all 4 cores at 7z.
By "exotic cooling", I didn't mean water. I meant sub-ambient. And I'd be very surprised if liquid nitrogen isn't able to reliably get a 7700K far past 5.3 GHz.
custom water loops are quite exotic
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
custom water cooling of any kind is not exotic, not even the one where you take a car radiator and put it outside in canada @ -20, and run water in it with a heavy-duty pump, ...
The first cooling that could be called exotic is phase-change cooling and then liquid nitrogen and liquid helium, ...
The real question to ask is does the 7700k have shitty TIM between the chip and IHS, moderately shitty TIM, almost not-shitty TIM, or indium solder, or different solder, ...
Oh, custom loops are still pretty exotic. AIOs are only now catching up.
it had shitty TIM, even silicon lottery recommends deliding and changing that insulator Intel uses as it has been proven even using some decent thermal paste reduces temperatures significantly. Its not what Intels engineers want, its what suits want even if it saves 0,05$/chip (like when they thiunned sunstrate and made CPUs much more prone to bending with Skylake)lol
Its clear they aren't running too well on the older motherboards. You are also talking about a chip that is barely on the market as well. From what I was able to gather it takes a spankin new motherboard to get it to overclock the way malabooga wants 5.3z without using water.
Still he loves to ignore the fact that they have had several people get the thing up to 7z on 2-4 cores. There are over 20 links to that information alone.
So we have this guy got 5z out of it very easily. He is using the z270 mobo though.
This guy got a solid 5z out of his z270 mobo as well.
They do need water cooling to run at those speeds.
So let's summarize this discussion.
You: Kaby Lake is a lot faster than Sandy Bridge.
Me: If you overclock both, Kaby Lake is still faster, but it's a lot closer.
You: If you overclock both, you should overclock Kaby Lake to still be 800 MHz faster than Sandy Bridge.
Me: Sandy Bridge tends to be able to overclock a lot further than Kaby Lake.
You: Some people can overclock Kaby Lake to about the same clock speed that they got with Sandy Bridge.
And that last line is apparently supposed to prove that an overclocked Kaby Lake will tend to have a clock speed about 1 GHz faster than an overclocked Sandy Bridge, as would be needed to maintain the performance difference that you argued is there.
So we have this guy got 5z out of it very easily. He is using the z270 mobo though.
This guy got a solid 5z out of his z270 mobo as well.
They do need water cooling to run at those speeds.
So let's summarize this discussion.
You: Kaby Lake is a lot faster than Sandy Bridge.
Me: If you overclock both, Kaby Lake is still faster, but it's a lot closer.
You: If you overclock both, you should overclock Kaby Lake to still be 800 MHz faster than Sandy Bridge.
Me: Sandy Bridge tends to be able to overclock a lot further than Kaby Lake.
You: Some people can overclock Kaby Lake to about the same clock speed that they got with Sandy Bridge.
And that last line is apparently supposed to prove that an overclocked Kaby Lake will tend to have a clock speed about 1 GHz faster than an overclocked Sandy Bridge, as would be needed to maintain the performance difference that you argued is there.
Actually I was trying to show malabooga that the 7700k wasn't designed the way he was trying to make it look. It was designed to run normally at a higher clock speed. It is also the one in the kaby lake series which they designed specifically to overclock.
Then you decided to tell us that they cannot overclock the 7700k to 5.3z without using liquid nitrogen.
So I decided to show you 3 different examples of it being overclocked to 7z using liquid nitrogen.
Then you decided to discredit the claim by saying they only had 1 core running which turned out to be false.
Then I decided it is way too early to get solid numbers from the 7700k when they barely have come out with motherboards which can handle all this overclocking. And half the chips being used for malabooga's claims are tech demo's that people are getting from second hand users.
It was designed like any other Intel chip. If you call 400 Mhz (12,25%) something "designed for overclocking", then Intel failed miserably. Because guess what: 2600k OCs from 3400 MHz all the way to average 5000 MHz, and thats tiny 47%
6400k, which stock clock is 2700 MHz OCs to ~4700 MHz. Tiny 74% OC.
Please do more, i guess you are trying to some serious trolling, but instead you manage to do such laughable pieces lol
the 7700k was designed to run at a higher clock speed. Which made it more stable then the other kaby lake processors when it came to overclocking. You again with miserable attempts at making intel or nvidia look bad. I can see right through it for what it is. They unfairly compared these processors in a state which the factory didn't intend.
When you look at the facts the kaby lakes are doing exactly what they are designed to do and that is run faster. It is when players decide to modify the intended use of the chips where you get all kinds of gray areas and lines that only make sense to certain groups of people.
If the sandy bridge is really that much better then why didn't intel just sell them at a higher clock speed? That would have eliminated the need for kaby lakes entirely. According to your ideas that is. You aren't looking at a whole picture and are instead looking at this little situation.
Look that drag car surely beat the poop out of that indie car. Well you only ran it for 1/2 mile. If you ran them at a 100 mile race the drag car would lose. Because that is how they are designed.
I'm hard core into flight sims. I run a two rig network, the main is for the sim and the secondary controls the real time weather and ATC plus a myriad of other less important apps.
I've just built a new main rig replacing my 4770K @ 4.6 with 16gigs G. skill Trident DDR3-2400 ram to a 7700k with 64gigs G. Skill Trident Z DDR4-3466 ram.
I clocked the 7700k to 4.94ghz with the press of a button. With HT deactivated (main sim makes no use of it) I easily clocked the 7700k to 5.15ghz with the highest temp using Real Bench = 78c using a Corsair 100i v2 for cooling. I got the chip running that fast with HT but I didn't like the temperatures.
For regular gaming the difference between the two CPU's wouldn't make a difference to me, the 4770k @ 4.6 runs games awesomely (more than enough). The difference in the flight sim is noticeable big time, I've gained about 30%-40% in performance. My cpu was my bottleneck, now my 1080 gpu is.
I clocked the 7700k to 4.94ghz with the press of a button. With HT deactivated (main sim makes no use of it) I easily clocked the 7700k to 5.15ghz with the highest temp using Real Bench = 78c using a Corsair 100i v2 for cooling. I got the chip running that fast with HT but I didn't like the temperatures.
Thank you Laserit for posting real life usage facts instead of assumptions based on dubious tests made in specific conditions in some lab.
So basically, the raw frequency gain between your 4770k @ 4.6 and your 7700k at 4.94 is 7%, and with disabling hyperthreading it's 11%. So the additional 20 to 30% performance you've gained come from the improvements of Kaby compared to Hasswell. That's 20 to 30% over 2 generations.
I'm hard core into flight sims. I run a two rig network, the main is for the sim and the secondary controls the real time weather and ATC plus a myriad of other less important apps.
I've just built a new main rig replacing my 4770K @ 4.6 with 16gigs G. skill Trident DDR3-2400 ram to a 7700k with 64gigs G. Skill Trident Z DDR4-3466 ram.
I clocked the 7700k to 4.94ghz with the press of a button. With HT deactivated (main sim makes no use of it) I easily clocked the 7700k to 5.15ghz with the highest temp using Real Bench = 78c using a Corsair 100i v2 for cooling. I got the chip running that fast with HT but I didn't like the temperatures.
For regular gaming the difference between the two CPU's wouldn't make a difference to me, the 4770k @ 4.6 runs games awesomely (more than enough). The difference in the flight sim is noticeable big time, I've gained about 30%-40% in performance. My cpu was my bottleneck, now my 1080 gpu is.
Very happy with the chip for my application.
We need more posts like this.
Originally posted by nethaniah
Seriously Farmville? Yeah I think it's great. In a World where half our population is dying of hunger the more fortunate half is spending their time harvesting food that doesn't exist.
Not that specific if you think about it. Such a simulator uses a wide range of CPU instructions. Other examples would be my approx. 15% gain from 4790k to 6700k AT THE SAME SPEED (4.4ghz, conservative overclock) for usages like compilation (C++, Java, etc...), 3D rendering and video encoding. Those aren't synthetic benchmarks but real world usages too.
For the average gamer, it doesn't make a big difference, but for really CPU intensive usages where the software is regularly updated to make use of the latest innovations, it definitely makes a quite noticeable difference.
I agree. Unfortunately, most of us don't use software that is updated regularly enough to make use of the latest innovations. Most PC users just use a web browser, maybe an email client, and MS Office if it's used at work.
I'd even go so far as to say that playing games that require even just moderate use of a graphics processor is more of a niche case than not. Simulators are an interesting use case, but a niche within a niche. Programmers that work on code bases large enough to realize a significant savings, again, a niche within a niche.
It's great that CPUs have these niche benefits that help these corner cases, particularly if it's a corner case that you happen to be able to use. But that doesn't reflect a typical case, and that's what this article is trying to show. Maybe software will evolve to turn some of these niche cases into more common occurrences, but until then I think it's perfectly fair to evaluate CPU performance based on typical use cases.
Maybe CPU performance ~has~ plateaued, and there are no more solutions to get increased performance, at least without some of these specialized cases becoming more common place or a radical breakthrough in technology. Or maybe it's just lack of competition has meant Intel doesn't have to put R&D into performance. Or maybe the performance is good enough and people just aren't willing to pay for more performance because there's no need.
I don't know, but I think this article confirms what I see on a day to day basis. Back in the 90's, a good computer may have lasted 3 years - by which time it was so outdated that it had to be upgraded or replaced because it just couldn't keep pace with current software. Every year and a half, performance seemed to double (or better). Once we hit about 2000-2005, that dropped dramatically. Today, I can't notice much real-world difference between my 4790K and my i920, or an Aarondale and Ivy Bridge Macbook Pros. Sure, one will bench faster than the other, and if I'm doing some specific things, I can tell, but just booting up the computer and loading up Chrome to hop on to troll these forums, I'd be hard pressed to tell one from the other, even though there are more than 5 years difference between them.
All the advances lately, I'd even say as far back as the past 10 years (that was the introduction of Intel Core lineup, and AMD Phenom lineup, and we haven't really pushed very far from those yet), haven't really been CPU-wise. SSD was the biggest jump in computer performance, and GPUs have continued to deliver 15+% generation over generation. RAM, USB, Thunderbolt, and other more peripheral technologies have come a good way, albeit at a slower pace, and we've seen display connectors mature as well. But CPUs, well... right now we are eliminating other bottlenecks, and once we get bottlenecked significantly by the CPU again, maybe we'll get back to it.
That is not what the OP is trying to show Ridelynn. He's trying to trash intel for not making better chips. When in fact about 90% of pc users can use intel chips that were made back in 2005 and never even notice the difference. He's trying to say they haven't improved and well for some people like gamers that is true. In fact gamers are better off with i5 processors that don't use hyperthreading.
That wasn't their goal on this chip it was something else entirely and it is a huge improvement for what their goal was.
I clocked the 7700k to 4.94ghz with the press of a button. With HT deactivated (main sim makes no use of it) I easily clocked the 7700k to 5.15ghz with the highest temp using Real Bench = 78c using a Corsair 100i v2 for cooling. I got the chip running that fast with HT but I didn't like the temperatures.
Thank you Laserit for posting real life usage facts instead of assumptions based on dubious tests made in specific conditions in some lab.
So basically, the raw frequency gain between your 4770k @ 4.6 and your 7700k at 4.94 is 7%, and with disabling hyperthreading it's 11%. So the additional 20 to 30% performance you've gained come from the improvements of Kaby compared to Hasswell. That's 20 to 30% over 2 generations.
Which is not that much.
30%-40% all things considered including the GPU.
I run 4k on a 48" Samsung TV. With the 4770K in demanding situations, I'd average about 33 fps. To reduce shimmering, I set my MSAA and SGSS to 4x. Dropping these settings does not increase my FPS as my cpu was my limiting factor.
With the same gpu in my 7700k in the same situation, I'm averaging 48 fps with MSAA and SGSS set to 2x. If I use the the 4x I'm dropping to about 40 fps and if I turn MSAA and SGSS off I'm getting over 60fps. The gpu is my limiting factor.
There is more at play here than just the cpu, I do have the DDR4-3466 ram and the simulator makes heavy use of the ram. I don't know the frames of running my 4770K without the MSAA and SGSS so I can't make that comparison.
The vanilla simulator will give me over 120fps, these frames are with all the candy including a study level Boeing 737 aircraft.
Comments
But thats just that filmoret showing his lack of education on the whole matter anyway, youre not getting 5 GHz on Kaby Lake anywhare near reliably even with beefiest water cooling lol
http://koolance.com/exc-800-portable-800W-recirculating-chiller
Assuming, of course, that you're willing to pay more for CPU cooling than most people do for an entire gaming computer.
Do you want me to google my uneducated guess for you?
Nitrogen also has a boiling point around -196 C, or about 220 C below ambient temperature in a typical room. The difference between temperatures on a typical water cooling versus high end air cooling is what, several degrees, maybe?
1/4 cant even hit 4,9 GHz
Almost half of 7700ks cant even hit 5 GHz lol
and 5% can hit 5,2 GHz
0 can hit 5,3 GHz with ANY conventional cooling without frying the chip
and for voltages > 1,3v you need water cooling unless you want it to hit 90c constantly lol
The first cooling that could be called exotic is phase-change cooling and then liquid nitrogen and liquid helium, ...
The real question to ask is does the 7700k have shitty TIM between the chip and IHS, moderately shitty TIM, almost not-shitty TIM, or indium solder, or different solder, ...
http://overclocking.guide/the-truth-about-cpu-soldering/
it had shitty TIM, even silicon lottery recommends deliding and changing that insulator Intel uses as it has been proven even using some decent thermal paste reduces temperatures significantly. Its not what Intels engineers want, its what suits want even if it saves 0,05$/chip (like when they thiunned sunstrate and made CPUs much more prone to bending with Skylake)lol
https://hardforum.com/threads/retail-7700k-not-up-to-5ghz-3600mhz.1922613/
Still he loves to ignore the fact that they have had several people get the thing up to 7z on 2-4 cores. There are over 20 links to that information alone.
Not necesssarily 20 rigs all hitting 7+
Also, done on an older Z170 motherboard, not a newer 270.
https://fossbytes.com/intel-core-i7-7700k-overclocked/
Doesn't say the person here but they used the z170 mobo and had all cores running. Also used a different motherboard and also had all cores running.
http://wccftech.com/intel-core-i7-7700k-7-ghz-7350k-5-ghz/
So we got 3 different situations where the it was clocked over 7z and not really possible for it to be fake news.
This guy got a solid 5z out of his z270 mobo as well.
They do need water cooling to run at those speeds.
You: Kaby Lake is a lot faster than Sandy Bridge.
Me: If you overclock both, Kaby Lake is still faster, but it's a lot closer.
You: If you overclock both, you should overclock Kaby Lake to still be 800 MHz faster than Sandy Bridge.
Me: Sandy Bridge tends to be able to overclock a lot further than Kaby Lake.
You: Some people can overclock Kaby Lake to about the same clock speed that they got with Sandy Bridge.
And that last line is apparently supposed to prove that an overclocked Kaby Lake will tend to have a clock speed about 1 GHz faster than an overclocked Sandy Bridge, as would be needed to maintain the performance difference that you argued is there.
Then you decided to tell us that they cannot overclock the 7700k to 5.3z without using liquid nitrogen.
So I decided to show you 3 different examples of it being overclocked to 7z using liquid nitrogen.
Then you decided to discredit the claim by saying they only had 1 core running which turned out to be false.
Then I decided it is way too early to get solid numbers from the 7700k when they barely have come out with motherboards which can handle all this overclocking. And half the chips being used for malabooga's claims are tech demo's that people are getting from second hand users.
6400k, which stock clock is 2700 MHz OCs to ~4700 MHz. Tiny 74% OC.
Please do more, i guess you are trying to some serious trolling, but instead you manage to do such laughable pieces lol
When you look at the facts the kaby lakes are doing exactly what they are designed to do and that is run faster. It is when players decide to modify the intended use of the chips where you get all kinds of gray areas and lines that only make sense to certain groups of people.
If the sandy bridge is really that much better then why didn't intel just sell them at a higher clock speed? That would have eliminated the need for kaby lakes entirely. According to your ideas that is. You aren't looking at a whole picture and are instead looking at this little situation.
Look that drag car surely beat the poop out of that indie car.
Well you only ran it for 1/2 mile.
If you ran them at a 100 mile race the drag car would lose.
Because that is how they are designed.
I've just built a new main rig replacing my 4770K @ 4.6 with 16gigs G. skill Trident DDR3-2400 ram to a 7700k with 64gigs G. Skill Trident Z DDR4-3466 ram.
I clocked the 7700k to 4.94ghz with the press of a button. With HT deactivated (main sim makes no use of it) I easily clocked the 7700k to 5.15ghz with the highest temp using Real Bench = 78c using a Corsair 100i v2 for cooling. I got the chip running that fast with HT but I didn't like the temperatures.
For regular gaming the difference between the two CPU's wouldn't make a difference to me, the 4770k @ 4.6 runs games awesomely (more than enough). The difference in the flight sim is noticeable big time, I've gained about 30%-40% in performance. My cpu was my bottleneck, now my 1080 gpu is.
Very happy with the chip for my application.
"Be water my friend" - Bruce Lee
Which is not that much.
I'd even go so far as to say that playing games that require even just moderate use of a graphics processor is more of a niche case than not. Simulators are an interesting use case, but a niche within a niche. Programmers that work on code bases large enough to realize a significant savings, again, a niche within a niche.
It's great that CPUs have these niche benefits that help these corner cases, particularly if it's a corner case that you happen to be able to use. But that doesn't reflect a typical case, and that's what this article is trying to show. Maybe software will evolve to turn some of these niche cases into more common occurrences, but until then I think it's perfectly fair to evaluate CPU performance based on typical use cases.
Maybe CPU performance ~has~ plateaued, and there are no more solutions to get increased performance, at least without some of these specialized cases becoming more common place or a radical breakthrough in technology. Or maybe it's just lack of competition has meant Intel doesn't have to put R&D into performance. Or maybe the performance is good enough and people just aren't willing to pay for more performance because there's no need.
I don't know, but I think this article confirms what I see on a day to day basis. Back in the 90's, a good computer may have lasted 3 years - by which time it was so outdated that it had to be upgraded or replaced because it just couldn't keep pace with current software. Every year and a half, performance seemed to double (or better). Once we hit about 2000-2005, that dropped dramatically. Today, I can't notice much real-world difference between my 4790K and my i920, or an Aarondale and Ivy Bridge Macbook Pros. Sure, one will bench faster than the other, and if I'm doing some specific things, I can tell, but just booting up the computer and loading up Chrome to hop on to troll these forums, I'd be hard pressed to tell one from the other, even though there are more than 5 years difference between them.
All the advances lately, I'd even say as far back as the past 10 years (that was the introduction of Intel Core lineup, and AMD Phenom lineup, and we haven't really pushed very far from those yet), haven't really been CPU-wise. SSD was the biggest jump in computer performance, and GPUs have continued to deliver 15+% generation over generation. RAM, USB, Thunderbolt, and other more peripheral technologies have come a good way, albeit at a slower pace, and we've seen display connectors mature as well. But CPUs, well... right now we are eliminating other bottlenecks, and once we get bottlenecked significantly by the CPU again, maybe we'll get back to it.
That wasn't their goal on this chip it was something else entirely and it is a huge improvement for what their goal was.
I run 4k on a 48" Samsung TV. With the 4770K in demanding situations, I'd average about 33 fps. To reduce shimmering, I set my MSAA and SGSS to 4x. Dropping these settings does not increase my FPS as my cpu was my limiting factor.
With the same gpu in my 7700k in the same situation, I'm averaging 48 fps with MSAA and SGSS set to 2x. If I use the the 4x I'm dropping to about 40 fps and if I turn MSAA and SGSS off I'm getting over 60fps. The gpu is my limiting factor.
There is more at play here than just the cpu, I do have the DDR4-3466 ram and the simulator makes heavy use of the ram. I don't know the frames of running my 4770K without the MSAA and SGSS so I can't make that comparison.
The vanilla simulator will give me over 120fps, these frames are with all the candy including a study level Boeing 737 aircraft.
"Be water my friend" - Bruce Lee