Discussion i7-11700K preliminary results

Page 19 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Mopetar

Diamond Member
Jan 31, 2011
7,848
6,015
136
You don't need to run it at high speeds because if it's a more efficient way to do the work it will be still be more efficient at lower clock speeds. If running the AVX code gets the task finished more quickly that's time the CPU doesn't need to spend using the cores as hard.
 
  • Like
Reactions: Tlh97 and coercitiv

BFG10K

Lifer
Aug 14, 2000
22,709
2,973
126
"It's hard to overstate how badly AMD is shaming Intel in the data center lately."


Someone give Dr. Lisa Su a "CEO of the year" if she doesn't already have one, because the turnaround under her leadership has been nothing less than a miracle.
 

moonbogg

Lifer
Jan 8, 2011
10,635
3,095
136
"It's hard to overstate how badly AMD is shaming Intel in the data center lately."


Someone give Dr. Lisa Su a "CEO of the year" if she doesn't already have one, because the turnaround under her leadership has been nothing less than a miracle.

I fully agree. It does need to be noted that Intel has been stuck in the mud for the past 5 years though. They've added cores and nothing else while AMD has created a killer architecture, one that we didn't even realize it's true potential until they moved past the whole CCX latency thing.
Had Intel's 10nm plans worked out, we would have had the choice between 6 fast Intel cores for $600 and 12 pretty fast AMD cores for $500. Good thing Intel screwed themselves, right? I mean haha, that was a close one for sure. Now I'm stuck with 8 slow Intel cores for $280 'cause I didn't want to pay $450 for 8 fast AMD cores. Wait, is this better?
 

coercitiv

Diamond Member
Jan 24, 2014
6,214
11,962
136
You don't need to run it at high speeds because if it's a more efficient way to do the work it will be still be more efficient at lower clock speeds. If running the AVX code gets the task finished more quickly that's time the CPU doesn't need to spend using the cores as hard.
The irony is many professionals are interested in ultra-portable performance for their workloads, and that includes designers, photographers and other visual oriented professionals who are heavy PS users. Some may choose a powerful "medium" portable machine and try to use it for most of their work, others rely on their workstations and seek an ultra-portable for increased mobility that is still able to handle most of their projects if need be.

PS has been my tool for almost 2 decades and over this span of time I tried out every combination of hardware: desktop only, ultra-portable only, ultra portable + medium laptop, ultra-portable + desktop (currently my preferred setup). As you can see the ultra-portable was almost always a choice, ever since the Core 2 Duo days.
 

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
They've added cores and nothing else while AMD has created a killer architecture, one that we didn't even realize it's true potential until they moved past the whole CCX latency thing.

AMD also has additional ACE in their sleeve: their current IO chip is burning ~100w of power on 14nm. If they can save half of that by moving it to TSMC 7nm and optimizing it, each core will receive 40-50% extra power.

In servers, extra power = higher clocks when more cores are busy = almost linear performance scaling in MT tasks. Basically they have 30-40% of extra performance ready for 2022 even if they don't touch actual cores.
 

scannall

Golden Member
Jan 1, 2012
1,946
1,638
136
AMD also has additional ACE in their sleeve: their current IO chip is burning ~100w of power on 14nm. If they can save half of that by moving it to TSMC 7nm and optimizing it, each core will receive 40-50% extra power.

In servers, extra power = higher clocks when more cores are busy = almost linear performance scaling in MT tasks. Basically they have 30-40% of extra performance ready for 2022 even if they don't touch actual cores.
Even moving to TSMC 10nm would improve power a lot, and not take up leading edge fab space.
 

jpiniero

Lifer
Oct 1, 2010
14,630
5,249
136

Videocardz has the official pricing - $399 for the 11700K and $374 for the 11700KF. That's 25 dollars more than the Comet Lake version. The 11900K is $539 which seems crazy.
 

Gideon

Golden Member
Nov 27, 2007
1,646
3,712
136

Videocardz has the official pricing - $399 for the 11700K and $374 for the 11700KF. That's 25 dollars more than the Comet Lake version. The 11900K is $539 which seems crazy.
True. Then again 11600KF at $237 and particularily the 11400F at $157 (now with overclocked memory!) look like very good deals as power consumption of these is managable.

AMD will hopefully release the 5600 and also lower 5600X price somewhat to offer the same bang/buck. Unfortunately until supply improves there's probably fat chance for that :(
 

CP5670

Diamond Member
Jun 24, 2004
5,514
592
126
The irony is many professionals are interested in ultra-portable performance for their workloads, and that includes designers, photographers and other visual oriented professionals who are heavy PS users. Some may choose a powerful "medium" portable machine and try to use it for most of their work, others rely on their workstations and seek an ultra-portable for increased mobility that is still able to handle most of their projects if need be.

PS has been my tool for almost 2 decades and over this span of time I tried out every combination of hardware: desktop only, ultra-portable only, ultra portable + medium laptop, ultra-portable + desktop (currently my preferred setup). As you can see the ultra-portable was almost always a choice, ever since the Core 2 Duo days.

I do the same thing, although I don't use PS. I use a Surface Pro for office work or software development, and a desktop for games and music. It's much slower than the desktop but fast enough, and the portability is much more valuable than more speed beyond a point.
 

coercitiv

Diamond Member
Jan 24, 2014
6,214
11,962
136
Well, this should provide the final pieces the "review" left out.
The original review was updated with results from new microcodes, and Anandtech followed the subject closely.

See updates in the original review.
Update 1: This review was originally posted on March 5th using 0x2C microcode, and has been updated on March 14th with data from 0x34 microcode. The difference between the two is about +1.8% on CPU tests, +3% on Gaming tests, including performance regression in some areas. This review will showcase both sets of results.
See more details on microcodes and memory behavior, emphasis mine.
Some commentary has appeared to suggest that the default setting for these ratios change between Core i9 and Core i7 – specifically that the Core i7 should default to 2:1 mode when run in DDR4-3200, effectively halving the memory controller which has historically been a limiting factor in DRAM overclocking. We cannot confirm if those are the official specifications at this point. However we can confirm that the motherboards we are testing do offer the user the choice of selecting a 1:1 ratio or a 2:1 ratio.

It should be noted that on all of the motherboards we have tested, all BIOS versions, the actual default operation for a Core i7 running at DDR4-3200 does appear to be the 1:1 mode. For the avoidance of doubt, in our testing on every microcode to date, all of our motherboards were running at a 1:1 ratio.