Benchmarks of the Skylake i7 6700K have emerged... Only marginally faster than 4790K...

[Curufinwe_wins]

Something sure does.

It's the way broadwell is on those low power chips. They also allowed a lot higher voltage than previous chipsets. You can read the maximum pc review. They talk about it quite a bit.

[patrickjp93]

Something sure does. 

that's about par for the 5775C. It's not a great overclocking chip, but to be fair all we've seen are the OEM chips since it's not a mass-market product. Rumor has it Intel intends to launch to e-tailers alongside the Skylake launch at Gamescon. we'll have to see if it's true or not.

[Curufinwe_wins]

that's about par for the 5775C. It's not a great overclocking chip, but to be fair all we've seen are the OEM chips since it's not a mass-market product. Rumor has it Intel intends to launch to e-tailers alongside the Skylake launch at Gamescon. we'll have to see if it's true or not.

That and for whatever reason those two chips use and allow for quite a bit higher voltages than we have been used to. That review mentioned they were told that anything below 1.4 was very much safe. (Not something we have seen in recent memory).

[patrickjp93]

That and for whatever reason those two chips use and allow for quite a bit higher voltages than we have been used to. That review mentioned they were told that anything below 1.4 was very much safe. (Not something we have seen in recent memory).

It likely means Intel found ways to reduce resistance and allow higher drive currents with the 4-FET designs.

[Curufinwe_wins]

It likely means Intel found ways to reduce resistance and allow higher drive currents with the 4-FET designs.

That was my thought as well, I mentioned on the original broadwell pages that it really isn't relevant what voltage the cpu is running at, just what that voltage compares to the safe limits.

In fact, my single biggest question is if skylake will run at high voltages like broadwell or more like previous iterations. (Actually, the "stock" voltage was the first thing that made me go look at the broadwell review because it seemed way too low)

[Master Disaster]

Uh, 0.4V?! No one should be believing this. Even if you take dynamic voltage into account, the decimal precision is ALWAYS 2-3 places, even if there has to be a 0. I've never seen a CPU-Z or GPU-Z give only 1 decimal place of precision, and it's not like I tweak the software.

EIST

[patrickjp93]

EIST

That would not explain the software displaying less decimals than usual. 0.4 should be 0.40.

[DildorTheDecent]

That would not explain the software displaying less decimals than usual. 0.4 should be 0.40.

that looks pretty accurate for a decimal readout.

 

It'll never be as good as a digital multi-meter for readout but it looks pretty normal. 

[Homicidium]

These benches are in all likelihood fake. This projects worse IPC than broadwell. Hold your damn horses for 5 weeks until objective reviews come out. Seriously people, learn to think for yourselves and verify before trusting!

 

Rumor or not, but do you really expect some huge gains with Broadwell? It has been the same story ever since Ivy Bridge (?) came out.

[Curufinwe_wins]

Rumor or not, but do you really expect some huge gains with Broadwell? It has been the same story ever since Ivy Bridge (?) came out.

page 2 has actual benchmarks from haswell and broadwell at the same clock speed that this supposed skylake is running. In order for this to be true, skylake would need to be straight up worse than broadwell is. That has never happened before. It will not happen this time.

This is a bs rumor page.

[patrickjp93]

Rumor or not, but do you really expect some huge gains with Broadwell? It has been the same story ever since Ivy Bridge (?) came out.

The difference in computational power between Sandy Bridge and Haswell is on the order of 85% if you have the software to leverage the new instructions. That requires the coding be done well and a compiler capable of doing vector-wise optimization on non-trivial be used. The former is already rare when the third tier programmers get into game studios and consumer computing, the second tier end up at big firms like Nielsen Group and Facebook, and the top tier end up working for IBM, Intel, Nvidia, and the supercomputing centers around the world. The latter is basically not possible because most people stick to Microsoft's Visual C/C++ compiler bundled in Microsoft Visual Studio which is pathetic at optimization compared to GCC/Clang/ICC. Furthermore, you have the business teams who want to minimize the executable size and hit the widest audience possible. That requires legacy support and not allowing alternate code paths for newer instructions on newer architectures.

 

In other words, your software today is extremely crippled on 3 sides, and you're blaming the hardware producer for slow gains on the oldest instructions which were optimized within inches of theoretical limits back with Sandy Bridge? Go look at the instruction latency tables for x86. Haswell has all the SISD computation instructions down to 1-3 cycles a piece, and the ones higher than 1 cycle (you can't go lower) are algorithmically very complex even in hardware, to the point it isn't theoretically possible to reduce them further. Don't blame Intel for greedy software businesses' choices.

[patrickjp93]

page 2 has actual benchmarks from haswell and broadwell at the same clock speed that this supposed skylake is running. In order for this to be true, skylake would need to be straight up worse than broadwell is. That has never happened before. It will not happen this time.

This is a bs rumor page.

It happened once, the Pentium 4 vs the Pentium 3.

[Agosto]

Well, I still have to upgrade from my 1366 rig. Since X99 has actually grown more expensive than some months ago ( and I want a real new platform ) I'm probably going to get a 6700K by the end of this year. 

Hope it's a good overclocker.

[Master Disaster]

That would not explain the software displaying less decimals than usual. 0.4 should be 0.40.

It says 0.400v?

[Curufinwe_wins]

It happened once, the Pentium 4 vs the Pentium 3.

Ahh yes those good old days. I was going to say in recent memory but I couldn't remember one. That was quite a transitional period itself.... Still eventually the pentium 4 showed to be significantly better than the Pentium 3 (even if only through brute clock speed).

Remember when Intel was talking about 7-10 GHz with netburst?

[Curufinwe_wins]

How i make the single core button appear ?

Anyway just got 1282 while ago

Honestly I'm not sure. I don't use cinebench, I've just been watching its scores since lots of others like it.

[patrickjp93]

Ahh yes those good old days. I was going to say in recent memory but I couldn't remember one. That was quite a transitional period itself.... Still eventually the pentium 4 showed to be significantly better than the Pentium 3 (even if only through brute clock speed).

Remember when Intel was talking about 7-10 GHz with netburst?

Nope, I was too young at the time to pay attention to Netburst. At that point I was playing soccer still and my eyesight hadn't gone to crap yet.

[Curufinwe_wins]

Nope, I was too young at the time to pay attention to Netburst. At that point I was playing soccer still and my eyesight hadn't gone to crap yet.

I was a kid that at the time believed the core speed mythos. I was super hyped up on the massive speeds I thought we were going to get.

[patrickjp93]

I was a kid that at the time believed the core speed mythos. I was super hyped up on the massive speeds I thought we were going to get.

It's possible with graphene to get up to 10GHz, but we're a long way off from making that happen unfortunately.

[Lays]

One segment of tech where I don't feel the upgrade itch.

 

But seriously, 0.4V??? as in a bit over a THIRD of the 4790k voltage? Can that be right?

 

That would mean the power consumption would be under 15% of Haswell from voltage alone and yet the TDP is 95W. Are they planning to run at 25GHz? I might be oversimplifying but I don't believe that it runs at .4V AND has a 95W TDP

 

EDIT: nvm I am dumb... 15% power consumption at idle makes sense... I thought that was somehow a spec sheet, excuse my ignorance about CPU-Z

 

Intel's adaptive voltage goes down to like .1v on haswell as well, it depends on the clock speed it's running. 

 

The operating voltage at full speed of this cpu will probably be like the rest, anywhere from like 0.9-1.2v

[Curufinwe_wins]

But seriously can we get grains of salt poured on this title and description. I mean most people are not going to read the actual refutation on page 2.

Calling mods and/or op.

[Norman_the_Owl]

inb4 "Wahh Moores law is dying!"

Literally nobody thinks that. We all know intel is just being lazy

[Nineshadow]

Uh, 0.4V?! No one should be believing this. Even if you take dynamic voltage into account, the decimal precision is ALWAYS 2-3 places, even if there has to be a 0. I've never seen a CPU-Z or GPU-Z give only 1 decimal place of precision, and it's not like I tweak the software.

My 4710HQ is at ~0.8V w/ Chrome open, and that's a laptop CPU D: .

[patrickjp93]

Literally nobody thinks that. We all know intel is just being lazy

Intel's not being lazy at all. Software developers and studios are the primary problem, with Microsoft doing nothing to help. When your target OS supports the Pentium III processors, your marketing team will say you have to as well to make sales, and then people don't upgrade as long as their computers keep working. Windows 10 still supports the Pentium IV. Look at the swathe of new instruction extensions which have come out since then. We haven't remotely begun to tap all the potential of the 2600K, and the 4790K is twice as powerful on its own.

 

Stop blaming Intel when it's the one doing all the work. The great thing about open source applications is you can compile them to use the best instructions you have on your PC with the newest compiler of your choice. On the Windows platform we have a giant stagnation caused by greedy studios with no regard for their own futures.

[Norman_the_Owl]

Intel's not being lazy at all. Software developers and studios are the primary problem, with Microsoft doing nothing to help. When your target OS supports the Pentium III processors, your marketing team will say you have to as well to make sales, and then people don't upgrade as long as their computers keep working. Windows 10 still supports the Pentium IV. Look at the swathe of new instruction extensions which have come out since then. We haven't remotely begun to tap all the potential of the 2600K, and the 4790K is twice as powerful on its own.

 

Stop blaming Intel when it's the one doing all the work. The great thing about open source applications is you can compile them to use the best instructions you have on your PC with the newest compiler of your choice. On the Windows platform we have a giant stagnation caused by greedy studios with no regard for their own futures.

If you honestly think that, i don't even know what to say. Intel's really not doing any work, seeing as AMD isn't any competition at all. And not to mention that the argument about past support is completely irrelevant. Microsoft doesn't need to build support in for a pentium 4, it's in there naturally...seeing as CPUs are still the same, just stronger/with more cores. It also has nothing to do with instruction sets, seeing as they are all included...Didn't you ever wonder why windows was 60GB? it's not all those neat UI elements, that's for sure