Coffeelake thread, benchmarks, reviews, input, everything.

[ozzy702]

Why didn't you respond to the person I was responding to as well, if you find these digressions irrelevant? Can't believe people have this much trouble finding relevant information on the internet in this day and age.

Because you're the biggest offender of derailing this thread so far and apparently also combatant when asked to keep to the topic at hand. Start a new thread about IPC where you can argue back and forth all you want. This isn't the place to do it.

 

[tamz_msc]

Because you're the biggest offender of derailing this thread so far and apparently also combatant when asked to keep to the topic at hand. Start a new thread about IPC where you can argue back and forth all you want. This isn't the place to do it.

Someone asked to clarify something, doing so means I'm derailing the thread. M'kay.

 

[mikk]

Coffee Lake vs KBL/SKL Pin Diagrams: 146 vs 128 VCC Pins
https://www.reddit.com/r/hardware/comments/73s6u3/coffee_lake_vs_kblskl_pin_diagrams_146_vs_128_vcc/

And more here: http://www.pcgameshardware.de/Coffee-Lake-Codename-266775/News/Z270-Inkompatibilitaet-Grund-1240154/

Intel had allocated some pins, which had previously supplied the integrated graphics unit with power, to the CPU cores. In addition, there have been unused pins, which are already used by the motherboard manufacturers for voltage supply. On the Z270 boards, Intel had assigned this a fixed function. The result: Coffee-Lake S-CPUs could literally burn on high-end Z270 models, because they got too high tensions over these pins.

(Google translation)

 

[TahoeDust]

4.0Ghz might be a bit hard for me to do, I've only been up to 3.8Ghz, I'm on stock air cooling on the rig I was planning to test with. I can give you a benchmark @ 3.8Ghz in a bit, maybe.

Edit: I've got one of my Ryzen rigs booted at 3.80Ghz, with 2x8GB DDR4-3000 @ 2667, with a (2+0) CCX configuration. They didn't offer a (1+0) config in the UEFI.

Edit: Ok, I managed to get my Ryzen 5 1600 to run at 4.000Ghz, @ 1.400V (UEFI), or 1.488V (CPU-Z).
It only boots once out of like 5-10 tries at these settings.



If you look carefully, that's with a 2+0 CCX configuration, and the CB R15 MT is almost double yours. Btw, my RAM is dual-channel, at 2667, 18-18-18-38.

Yet, your ST is quite a bit higher.

So, why isn't your MT quite a bit higher? Because the SMT on my Ryzen made up for the ST deficit, and is practically matching your score for MT, if you cut my score in half, because I was running two cores rather than one.

OK. I ran 2c/4t at 4.0GHz. Intel's multithread is a marginally higher ~2%, but single thread is ~9% higher.

Let's not forget, your result was at your max overclock. Coffeelake results at max overclock will look more like this...

~28% higher Multicore and 37% higher single core. That is a VERY significant difference.



This thread has gotten WAY off topic. Sorry for any contribution I had in that.

 

[Phynaz]

There is a possibility that AMD will bring 35W TDP CPUs into their lineup.

AMD is pricing their products at performance levels, not "brand levels". 65W six core, has lower price than 95W Six core, and similar situation we would get with 35W six core: lower price than 65W version.

If I remember correctly engineering samples of quad core/eight thread CPUs made on 14 nm process have had 3.0/3.3 GHz, and this clocks would be achievable even for 8 core CPU in this thermal envelope. So on 12LP process we can get barely higher clocks. If the 35W TDP CPU, with 6 cores/12T will have 3.2/3.5 GHz core clocks, or even 3.3/3.6 and price lower than 1600 - then unfortunately for Intel, I will have to consider this option, very much. Because Im pretty sure, that at this TDP, 6C/6T from Intel will have similar core clocks in Turbo state, but potentially higher price.

This isn't an AMD thread.

 

[Phynaz]

Apparently you cannot comprehend how turbo modes work, even when somebody posted them on this same page!


Now compare with Ryzen 3 specs.

Not an AMD thread.

 

[Phynaz]

That's funny because five years ago 4GHz Sandy Bridge Vs 5GHz Vishera was decided on IPC.

Now it's become irrelevant. M'kay.

You're posting about something from five years ago in the Coffee Lake thread.

 

[tamz_msc]

Not an AMD thread.

So? Just clarifying questions related to clock speed.

 

[Phynaz]

Superior in the sense that when workloads are ideally suited to SMT in general, which means excluding low latency or FPU intensive scenarios, AMD's SMT comes out ahead of Intel's HT more often than not.

This is still not an AMD thread.

 

[tamz_msc]

You're posting about something from five years ago in the Coffee Lake thread.

Does that disqualify it from being true?

 

[Phynaz]

Does that disqualify it from being true?

Not the place for it and you know it.

 

[tamz_msc]

This is still not an AMD thread.

Are you familiar with the protagonist of Gunther Grass's novel The Tin Drum? He looks like this.

Trolling is not allowed
Markfw
Anandtech Moderator

 

[Phynaz]

Are you familiar with the protagonist of Gunther Grass's novel The Tin Drum? He looks like this.

It's not the trolling thread either. Reported.

 

[Phynaz]

Coffeelake thread, benchmarks, reviews, input, everything.

 

[The Stilt]

ah ok, I guess that explains some video encodng bench Ive seen, where SK-X is clearly faster than Coffee, despite its lower freq clock

X264 has some preliminary AVX512 optimizations, however the speed improvements are rather diminishing.

This is rather interesting: http://x265.org/x265-receives-significant-boost-intel-xeon-scalable-processor-family/

 

[tamz_msc]

X264 has some preliminary AVX512 optimizations, however the speed improvements are rather diminishing.

This is rather interesting: http://x265.org/x265-receives-significant-boost-intel-xeon-scalable-processor-family/

Is it true that the disproportionate uplift in x264 compared to x265 due to AVX512 is due to the former being less FP heavy, and thus reflects the scenario as reported by Sandra that AVX512 FP improvements>>AVX512 integer improvements?

 

[coercitiv]

Coffeelake thread, benchmarks, reviews, input, everything

We definitely need to be more specific.

 

[PhonakV30]

Coffeelake thread, benchmarks, reviews, input, everything.

Use Quote Not reply.

 

[MarkPost]

X264 has some preliminary AVX512 optimizations, however the speed improvements are rather diminishing.

This is rather interesting: http://x265.org/x265-receives-significant-boost-intel-xeon-scalable-processor-family/

Dont know about x264, but I saw a Handbrake bench (supposedly with x265) in one of those romanian sites, showing 8700K (oced at 4.7) slower than 7800X (at 4.0). I assume the difference is due to AVX full support in 7800X and not in 8700K. Dont know the accuracy of the test though

 

[Zucker2k]

Dont know about x264, but I saw a Handbrake bench (supposedly with x265) in one of those romanian sites, showing 8700K (oced at 4.7) slower than 7800X (at 4.0). I assume the difference is due to AVX full support in 7800X and not in 8700K. Dont know the accuracy of the test though

I believe the 7800x's AVX512 ports can do 2xAVX2. The 8700k can only do 1xAVX2 per instruction cycle.

 

[The Stilt]

Is it true that the disproportionate uplift in x264 compared to x265 due to AVX512 is due to the former being less FP heavy, and thus reflects the scenario as reported by Sandra that AVX512 FP improvements>>AVX512 integer improvements?

X265 doesn't use AVX512 at all at the moment.
It's assembler (YASM) doesn't even support AVX512 as it is heavily outdated.

In order to support AVX512 the assembler used for X265 must be changed to NASM (which requires work).

 

[DrMrLordX]

For those of you complaining about derails:

https://forums.anandtech.com/thread...input-everything.2519949/page-5#post-39095978

It was derailed that far back. Set out the bait and the flies come swarming.

It is generally best not to mention such-and-such other CPU being rendered obsolete or sent back to the bargain bin lest ye prompt others to come and defend said CPU or whoever makes it. Just a waste of everyone's time. Well, in a thread dedicated to one particular CPU line anyway.

 

[MarkPost]

For those of you complaining about derails:

https://forums.anandtech.com/thread...input-everything.2519949/page-5#post-39095978

It was derailed that far back. Set out the bait and the flies come swarming.

It is generally best not to mention such-and-such other CPU being rendered obsolete or sent back to the bargain bin lest ye prompt others to come and defend said CPU or whoever makes it. Just a waste of everyone's time. Well, in a thread dedicated to one particular CPU line anyway.

yeah, its just a matter that some folks, if they read something that doesnt fit with their agenda, then its derail; if it fits with their likes, then its just fine.

 

[MarkPost]

I believe the 7800x's AVX512 ports can do 2xAVX2. The 8700k can only do 1xAVX2 per instruction cycle.

yeah I suppose its due to something like that

 

[Glo.]

I will repost from locked thread since there I have not got answer to my question:

In my country Asrock H270 mITX mobo costs less than some B250 MoBos. I don't believe we will see price changes between MoBos from X2XX chipset to X3XX chipset for Intel platform. So here is a question, actually.

If I have a MoBo that has 2666 MHz memory compatibility, and CPU that is locked and capable of running 2400 MHz, can I still buy 2666 Mhz and run it full speed?

Also. So now we know that only K CPUs are getting price hike, vs Kaby Lake? All other SKUs have the same prices?