• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Money no object: which CPU do you choose?

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

5960X + X99 or 6700K + Z170 given to you for free?

  • 5960X + X99, no question.

  • 6700K + Z170, without a shadow of a doubt.

  • I would flip a coin.

Results are only viewable after voting.
I own a rock solid (RealBench solid) 4.4Ghz 5960x and it's fast, both in gaming and in multi-tasking.

For pure gaming the 6700k is probably faster but not by much.
 
Last edited:
guskline said:
I own a rock solid (RealBench solid) 4.4Ghz 5960x and it's fast, both in gaming and in multi-tasking.

For pure gaming the 6700k is probably faster but not by much.
Click to expand...
If faster at all. Some games benefit from cache, and 5960x has lots of it. 6700K is a good CPU but doesn't offer a class of its own perfomance compared to older tech such as 4790K & Broadwell i7's. In fact, from a cpu perspective, desktop Skylake may have been the least exciting mainstream top processor in recent years... In my humble opinion of course. Personally, I value desktop mainstream Broadwell much more, purely because of its superb power efficiency and the inclusion of L4 cache that can be used in some non-gaming tasks.
 
Last edited:
theAnimal said:
My primary use would be Folding @ Home, but really any software that is capable of multi-threading, which will only increase in future.
Click to expand...

Fair enough, for your specific and quite niche use case 144 cores might be better. For my use case, even the 6 I have now would be faster than the system you're proposing. Slower doesn't sound like much better to me. 😛
 
Magic Carpet said:
You call that stability?
Click to expand...

At least his use of RealBench is a lot better than the synthetics like Prime95/OCCT/IBT.
http://www.overclock.net/t/1510388/haswell-e-overclock-leaderboard-owners-club/2390#post_22900116

I used to believe the old timers and veterans when I first joined this forum but my view is now 180*. A stable system for me is the maximum operating speed it retains in all real world applications I run. It makes no difference to me if some synthetic power virus loads my CPU/GPU to 100.00% because I will never run any program that will use my components like that.

Also, the newer versions of these programs place a completely unrealistic load on CPU/GPU. FurMark = useless junk.

What's more important to you, having your system running 200-300mhz slower but stable in synthetic tests that have no association with real world programs? Or, 100% rock solid stable in every game, every distributed computing, rendering, encoding, etc. application you use?

I am not suggesting that RealBench is the only valid way to stress a CPU but I'll prefer real world apps over synthetic apps for stability testing.
 
Realbench is perfect for stability testing. Prime95 CPU heavy test is not recommended for use on X99 because it puts an unrealistic load on the CPU, as well as the power delivery. 2011-v3 draws a ton of power when you overclock it and Prime95 can blow out your motherboard's VRMs when you're running a 5960x at 4.6ghz with close to 1.4V 😀

I'm a bit lazy to dig up the numbers right now, but a highly overclocked 5960x can draw close to 300W fully loaded if my memory serves.

edit: 4.4ghz 5960x is pulling about 350w http://www.corsair.com/en-us/blog/2014/september/haswell_e_overclocking_and_power

Skimmed it quickly, they said they used corsair link and the software for the psu to measure draw. I'm actually not sure if that was full system load or not. Have to account for MB consumption and whatever a 780ti idling consumes as well then.
 
Last edited:
RussianSensation said:
At least his use of RealBench is a lot better than the synthetics like Prime95/OCCT/IBT.
http://www.overclock.net/t/1510388/haswell-e-overclock-leaderboard-owners-club/2390#post_22900116

I used to believe the old timers and veterans when I first joined this forum but my view is now 180*. A stable system for me is the maximum operating speed it retains in all real world applications I run. It makes no difference to me if some synthetic power virus loads my CPU/GPU to 100.00% because I will never run any program that will use my components like that.

Also, the newer versions of these programs place a completely unrealistic load on CPU/GPU. FurMark = useless junk.

What's more important to you, having your system running 200-300mhz slower but stable in synthetic tests that have no association with real world programs? Or, 100% rock solid stable in every game, every distributed computing, rendering, encoding, etc. application you use?

I am not suggesting that RealBench is the only valid way to stress a CPU but I'll prefer real world apps over synthetic apps for stability testing.
Click to expand...

I would sacrifice the 200-300mhz for the peace of mind.
 
crashtech said:
Better for what, though? Very little in practical reality.
Click to expand...

Having 144 cores (with enough ram) would be better for a lot:

- Very quick video encoding time
- Very quick rending time
- Very quick compressing time (7zip, which I use a lot)
- Very quick chess analysis (depth of 30 with MultiPV of 10 would take seconds)
- Run many virtual images, each with 2 or more cores
- Random other things: folding at home
- Coding: I sometimes mess around with very large numbers in simulations and calculations in c#, id love to have access to 144 cores
 
Given a choice between a free quad core and a free Octacore....

I would def overclock.

Plus I already have a 6700K, does that count?
 
Last edited:
6700K. I need the ST performance and the IPC that Skylake offers. I don't need more than 4c/8t for a good while.

The current HEDT alternatives has nothing to offer for my use. If there was a 6c/12t alternative based on Skylake, I'd have been tempted because of the novelty but I probably wouldn't have had use for it. The current older HEDT platform is not interesting at all.
 
Shmee said:
What about the broadwell-E upgrade?
Click to expand...

Sure, if you're willing to wait, you can have Broadwell-E 6950X in this scenario.

The point was to try to determine what is more important to people: fewer, higher perf/clock cores, or more, slightly lower perf/clock cores.
 
This scenario doesn't really seem controlled enough for that. If you look at the answers people are factoring overclocks (greatly negating the clockspeed difference), cache size differences (negates single thread performance differences), platform feature differences, and various other factors.

I wonder how much the results would change if it were phrased as:
CPU A: 4c/8t CPU
vs
CPU B: 8c/16t CPU

CPU A is clocked 33% higher. CPU B has 50% higher power consumption at full load (so both are at parity factoring core counts and clockspeeds). Everything else is identical. Rest of the system is identical. You cannot tweak either setup further.

What would be even more interesting is you set a baseline performance so the differences mattered more. What if we set CPU A's single threaded performance back to stock Sandybridge 2600k levels as the baseline.
 
Last edited:
In that case higher single thread IPC is more important for me. Hence why I decided to go Skylake instead of X99.
 
arandomguy said:
This scenario doesn't really seem controlled enough for that. If you look at the answers people are factoring overclocks (greatly negating the clockspeed difference), cache size differences (negates single thread performance differences), platform feature differences, and various other factors.

I wonder how much the results would change if it were phrased as:
CPU A: 4c/8t CPU
vs
CPU B: 8c/16t CPU

CPU A is clocked 33% higher. CPU B has 50% higher power consumption at full load (so both are at parity factoring core counts and clockspeeds). Everything else is identical. Rest of the system is identical. You cannot tweak either setup further.

What would be even more interesting is you set a baseline performance so the differences mattered more. What if we set CPU A's single threaded performance back to stock Sandybridge 2600k levels as the baseline.
Click to expand...
But you are not limited to stock speeds in real life, why should this be different?
 
bystander36 said:
But you are not limited to stock speeds in real life, why should this be different?
Click to expand...

If we are trying to determine this -
Arachnotronic said:
The point was to try to determine what is more important to people: fewer, higher perf/clock cores, or more, slightly lower perf/clock cores.
Click to expand...

Then we need controls in place.

Otherwise are you really preferring "more, slightly lower perf/clock cores" or are you just preferring more cores. With basically no trade off how many people will want less cores?
 
arandomguy said:
If we are trying to determine this -


Then we need controls in place.

Otherwise are you really preferring "more, slightly lower perf/clock cores" or are you just preferring more cores. With basically no trade off how many people will want less cores?
Click to expand...

The way I see it, we have 2 things we could do:
1) What we have now if both OC. The i7 6700K is slightly faster, the i7 5960x has more cores.
2) Or you could ask, how much faster does a 4c/8t CPU have to be than an 8c/16t for you to choose it?

There really doesn't need to be more controls than that. Either you look at what we do have as a choice, or what the tipping point is. Obviously if the gap is large enough, speed wins, otherwise people would pick AMD's 8 core FX processors.

IMO #1 is the most valid, because this is what we would have as a difference. Or what we actually do have.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top