• 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.

Coffeelake thread, benchmarks, reviews, input, everything.

Page 69 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Markfw said:
So you can get an AMD 8 core that does 4.3 ghz stock, and 90% of the IPC of the Intel for $319, but $500 for 5 ghz makes them "poor" in IPC and clockspeed ? And Intel and AMD both gain on fast ram, but AMD can run just fine on 2400. For over 50% more cost I think the AMD is a better value than the Intel, but yes, that is a nice chip IF it will do 5 ghz at stock vcore.

We will see.
Click to expand...

What AMD does 4.3ghz all core stock? My 2700x requires 1.5v to get 4.15ghz all core stable and 1.525 to get 4.2ghz. The system will not boot with all core set to 4.3ghz.
I tested this on two 2700x chips(I bought two off amazon and returned the one which couldn't do 4.1ghz all core at any voltage). I would agree with the statement above regarding price performance if the 2700x were clocked at 4.5+ghz but, when\if this Intel processor arrives it could have a 1ghz advantage all core and more in single\dual thread scenarios.
 
slashy16 said:
What AMD does 4.3ghz all core stock? My 2700x requires 1.5v to get 4.15ghz all core stable and 1.525 to get 4.2ghz. The system will not boot with all core set to 4.3ghz.
I tested this on two 2700x chips(I bought two off amazon and returned the one which couldn't do 4.1ghz all core at any voltage). I would agree with the statement above regarding price performance if the 2700x were clocked at 4.5+ghz but, when\if this Intel processor arrives it could have a 1ghz advantage all core and more in single\dual thread scenarios.
Click to expand...
The 2700x does that stock turbo boost on one core. Are you saying you think this new $500 chip will do all cores stock at 5 ghz ? I doubt that.

But lets see before we argue.
 
Markfw said:
So you can get an AMD 8 core that does 4.3 ghz stock, and 90% of the IPC of the Intel for $319, but $500 for 5 ghz makes them "poor" in IPC and clockspeed ? And Intel and AMD both gain on fast ram, but AMD can run just fine on 2400. For over 50% more cost I think the AMD is a better value than the Intel, but yes, that is a nice chip IF it will do 5 ghz at stock vcore.

We will see.
Click to expand...

In relative terms, he is correct on both counts, AMD has lower IPC and clockspeed, hence 'poor'.

As strong as AMD is at MT performance, you're always giving up significant ST performance because it's up to 1GHz lower in fmax plus lower IPC.

WR to value I'm sure I've mentioned this before because you don't buy a CPU in isolation, and the cost of a motherboard and 16GB of high speed DDR4 really dilutes the price differences at CPU level alone.

Again, if the difference between a 9900K platform and 2700X platform was $200 ($1000 vs $800, for arguments sake) I think a lot of people would prefer the 9900K even if it has poorer price/performance at the CPU level, on a total platform cost/performance basis it's not really any worse, in fact it's probably better if it turns out to be >25% faster overall, which it should be considering the mooted clockspeeds
 
Markfw said:
But lets see before we argue.
Click to expand...

Agreed. I will say though that every recent Intel CPU I have owned has run all cores at the highest single core turbo frequency at stock voltage. I am wondering if this CPU is running some new tweaked 14nm+ plus one plus in addition to whatever plus they are on 🙂
 
slashy16 said:
Agreed. I will say though that every recent Intel CPU I have owned has run all cores at the highest single core turbo frequency at stock voltage. I am wondering if this CPU is running some new tweaked 14nm+ plus one plus in addition to whatever plus they are on 🙂
Click to expand...
AFAIK, all recent Intel CPUs will run all their cores at the single core turbo speed with no problem.*

If the 9700K and 9900K aren't able to do that, it would be unusual.

*Locked Haswell chips, including Xeons, will also do this on an overclocking mobo. My E3-1231V3 runs all it's cores at 3.8 on my Z97 board, for example. 4790S runs all it's cores at 4.0, etc.
 
LTC8K6 said:
AFAIK, all recent Intel CPUs will run all their cores at the single core turbo speed with no problem.*

If the 9700K and 9900K aren't able to do that, it would be unusual.

*Locked Haswell chips, including Xeons, will also do this on an overclocking mobo. My E3-1231V3 runs all it's cores at 3.8 on my Z97 board, for example. 4790S runs all it's cores at 4.0, etc.
Click to expand...

There have been discussions about this before. IIRC, none of the locked *Lake chips work this way. You are limited to the boost table speeds.
 
PeterScott said:
There have been discussions about this before. IIRC, none of the locked *Lake chips work this way. You are limited to the boost table speeds.
Click to expand...
I believe that is correct, but I don't recall ever seeing a demonstration of the behavior of a locked *lake chip on a Z board.
 
Unconfirmed reports are that Whiskey Lake and Amber Lake have the hardware fixes for Smeltdown. I wonder if that also now includes the Coffee Lake Refresh 8 core. Probably have to wait for a review to confirm if this is true or not.
 
jpiniero said:
Unconfirmed reports are that Whiskey Lake and Amber Lake have the hardware fixes for Smeltdown. I wonder if that also now includes the Coffee Lake Refresh 8 core. Probably have to wait for a review to confirm if this is true or not.
Click to expand...
That seems very unlikely. The fixes are going to go into server chips first, I'd think.
 
ehume said:
Got a question. Will the i7 9700k (8c/8t) have trouble with the meltdonbug and its descedants?
Click to expand...

sure it will but, its a none issue. I'm surprised most desktop users even care about it. If you were to run VM's on your machine and you made them accessible to the public I could see an issue.
 
slashy16 said:
sure it will but, its a none issue. I'm surprised most desktop users even care about it. If you were to run VM's on your machine and you made them accessible to the public I could see an issue.
Click to expand...

People care because the patches are forced on us and degrade performance regardless if I run public VMs or not. In fact intel should have "manned up" and said the patches are unneeded for consumers and the issue overblown outside of cloud providers.
 
beginner99 said:
People care because the patches are forced on us and degrade performance regardless if I run public VMs or not. In fact intel should have "manned up" and said the patches are unneeded for consumers and the issue overblown outside of cloud providers.
Click to expand...
If I see or detect a performance decrease, I'll complain.

If I have to run benchmarks to notice it, well, I don't run benchmarks much.
 
I have opted to ignore windows updates anyway. I usually install windows and disbale everything related to updates. The same goes for bios patching. I never update a bios unless there is a microcode update for a new cpu or whatever. Note for professional use(workplace) we patch everything obviously for CYA purposes.
 
Amberlake's only improvement is moving to 5W TDP. It uses the last generation PCH without the changes(integrated WiFi MAC, quad core DSP). The CPU core doesn't have any changes either.
 
Last edited:
Thank you all for replying. Now I ask, How will the meltdown/spectre bugs affect a CPU like the i7 9700k that has no hyperthreading?
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top