• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Intel Core i7-7700K "Kaby Lake" SANDRA Scores Surface

Its actually worse than that, the measured increase in IPC is 2,7% for GFLOPS(1,0785/1,05) and 2,2% for Mpix/s (1,0735/1,05).

Comparing the Xeon versions of them we see that the E3 1275 V5 at 3,6 Ghz with IGP is a 80W part, while the E3 1275 V6 at 3,8 with IGP is a 78 W part, add inn the anemic IPC increase at approximately 2,5 % and this seams like one massive let down performance wise.

You are both incorrect. You are using the term IPC synonymously with per core performance increase. That's not what the term means:

"instructions per cycle (IPC) is one aspect of a processor's performance: the average number of instructions executed for each clock cycle."

That means the only way to measure IPC of Kaby Lake against Skylake is at the same CPU frequency. Since Kaby Lake is simply made on a new 14nm+ node, but the underlying CPU architecture is 100% identical (GPU improvements are irrelevant here unless measuring 4K encoding/decoding acceleration), the IPC of Kaby Lake and Skylake will be identical at the same clocks. AnandTech confirms:

"We are being told that the underlying microarchitecture for Kaby Lake is the same as Skylake, and that the frequency adjustments from the new process, along with features such as Speed Shift v2 and the new fixed function media codecs, account for the performance improvements as well as battery life increases when dealing with 4K content.

For users that speak in pure IPC, this may/may not be a shock. Without further detail, Intel is implying that Kaby Lake will have the same IPC as Skylake, however it will operate with a better power efficiency (same frequency at lower power, or higher frequency at same power) and for media consumption there will be more idle CPU cycles with lower power drain. The latter makes sense for mobile devices such as tablets, 2-in-1s and notebooks, or for power conscious users, but paints a static picture for the future of the desktop platform in January if the user only gets another 200-400 MHz in base frequencies."


If you want to make any argument for an improved experience on the desktop against i7 6700K, it will either have to come from Kaby Lake's higher overclocking headroom, which we will not know about until January, or from system response time due to SpeedShift v2.

In my personal opinion, anyone on 2600K/3770K 4.5Ghz or faster who has not upgraded to i7 6700K starting August 2015 has little to no incentive to upgrade to Kaby Lake or even Cannon Lake. The most logical upgrade for those who skipped Skylake is the next true architecture with Ice Lake or at minimum Skylake-X on 2066.
 
Getting 200 to 400 MHz extra in the same thermal/power envelope is by no means bad ... Kaby Lake is what it is, an incentive for AMD to clock dat zen :rockout:
 
Is this really true??? Haven't anyone read recently that intel's kabby lake i7 7700k cpu is clocked at 3.6ghz with turbo boost up to 4.2ghz??? And that it'll be backward compatible with the existing z170 chipset and the 1151 socket???
 
...The most logical upgrade for those who skipped Skylake is the next true architecture with Ice Lake or at minimum Skylake-X on 2066.

With the current "progress" on CPU performance, that date might not even sound as irony ;)
 
Is this really true??? Haven't anyone read recently that intel's kabby lake i7 7700k cpu is clocked at 3.6ghz with turbo boost up to 4.2ghz??? And that it'll be backward compatible with the existing z170 chipset and the 1151 socket???

Why dont you post a link for us to see???????
 
You are both incorrect. You are using the term IPC synonymously with per core performance increase. That's not what the term means:

"instructions per cycle (IPC) is one aspect of a processor's performance: the average number of instructions executed for each clock cycle."

That means the only way to measure IPC of Kaby Lake against Skylake is at the same CPU frequency. Since Kaby Lake is simply made on a new 14nm+ node, but the underlying CPU architecture is 100% identical (GPU improvements are irrelevant here unless measuring 4K encoding/decoding acceleration), the IPC of Kaby Lake and Skylake will be identical at the same clocks. AnandTech confirms:

yeah, IPC is ideally measured at the same Hz, but since the two were reportedly NOT on the same frequency

Core i7-7700K "Kaby Lake"... The core ticks at 4.20 GHz, with a maximum Turbo Boost frequency of 4.50 GHz.
...
To put things in perspective, a Core i7-6700K "Skylake" processor running at its stock speed of 4.00 GHz core with 4.20 GHz Turbo Boost

the 7700k has a 5 % clock speed advantage(0,2 Ghz / 4 GHz = 0,05 = 5%, therefore dividing the performance advantage with 1.05 gives a (assumed linear scaling based on clock speed) approximation of the 7700k at 4 GHz.
 
See no reason to upgrade from my 4770k still. 4.6 ghz OC'd.. It's a good thing..
 
Given that some motherboard manufactures ( Asus *cough* ) can even override multicore speed for 1, 2, 3 or 4 cores usage patterns, I can't even see it as viable upgrades since the features is already there
Even the IPC is in the margin of error. Maybe they even test it with DDR4 4266Mhz to get the IPC there.
Who knows
I buy now because from what I see, the prices will increase once more for no reason
 
Back
Top