Some hyperthreading results

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
rainhaus
Posts: 143
Joined: Sun Feb 01, 2009 6:26 pm

Re: Some hyperthreading results

Post by rainhaus » Mon Dec 12, 2016 9:35 pm

kasinp wrote:
Laskos wrote:
kasinp wrote:
kasinp wrote:Kai,

By disabling turbo boost you are creating a more favorable setup for HT ON. It's as if the reserve CPU cycles (the "boost") are now available to the extra threads, making them more effective. However, leaving turbo boost ON, while disabling HT transfers that benefit to the physical CPUs.

Fritz benchmark i7-2600 (BIOS HT=ON, TBoost=OFF) 4 threads = 8442
Fritz benchmark i7-2600 (BIOS HT=OFF, TBoost=ON) 4 threads = 10453

I believe a conclusive test would have to use two identical machines one set up optimally for HT=ON, and the other for HT=OFF.

Regards,
PK
I did a quick test to get the corresponding benchmark my i7-4790:

Fritz benchmark (BIOS HT=ON, TBoost=OFF) 4 threads = 9747
Fritz benchmark (BIOS HT=OFF, TBoost=ON) 4 threads = 12155

PK
You are right if one takes into account overclocking. Without hyperthreading, CPU overclocks better. So, if thinking of the same temperature and stability with overclck, HT OFF is probably better performance-wise.
Right, and Turbo Boost is basically the chip's internal overclocking algorithm.
I ran an overnight test between two identical i7-2600 PCs.

Stockfish 8 (HT=OFF, TBoost on, 4 threads) +12 -12 =54
Stockfish 8 (HT=ON, TBoost off, 6 threads) +12 -12 =54

I know, a shorter series, but I still find the result pleasing. Modern engines seem to be able to squeeze the most of the available chip power one way or another!
PK
It really seems to be so. Running an i7-4790k engine-match on 4 CPU (Turbo on, HT off) my CPUID-HW Monitor shows permanently 4,2 Ghz, 200 Mhz more than the standard output. I had always thought the Turbo doesn't react under full CPU-load. However, I have to accept an increase of about 5 °C in the cores. But 58° C is not a problem at all.

Look rather similar your overnight scores, indeed : ), but forget it. You are handling 3 variables with 6 possible characteristic values: HT on/off, TBoost on/off, physical mode/logic mode. You need a little bit much more games to get into valid areas.
You suppose "by disabling turbo boost you are creating a more favorable setup for HT ON"
That calls for a basic research about the interaction of Hyperthreading, Turbo Boost and engine speed. I have something in mind resp. I'm just going to do something in this direction. Maybe there will be a nice table this week. Kai's empiric quickies are laudable and he shouldn't be the only one producing nice tables and diagrams ; )
Thread viewImage
flat view
is a bad view
without thread view

APassionForCriminalJustic
Posts: 415
Joined: Sat May 24, 2014 7:16 am

Re: Some hyperthreading results

Post by APassionForCriminalJustic » Tue Dec 13, 2016 1:45 am

kasinp wrote:
APassionForCriminalJustic wrote:
kasinp wrote:Kai,

By disabling turbo boost you are creating a more favorable setup for HT ON. It's as if the reserve CPU cycles (the "boost") are now available to the extra threads, making them more effective. However, leaving turbo boost ON, while disabling HT transfers that benefit to the physical CPUs.

Fritz benchmark i7-2600 (BIOS HT=ON, TBoost=OFF) 4 threads = 8442
Fritz benchmark i7-2600 (BIOS HT=OFF, TBoost=ON) 4 threads = 10453

I believe a conclusive test would have to use two identical machines one set up optimally for HT=ON, and the other for HT=OFF.

Regards,
PK

What difference does Turbo Boost make? I know that once all threads are running at 100 percent usage the all-core turbo is 2.28 gigahertz on my dual-Xeon socket rig. That frequency does not change whether hyperthreading is enabled or disabled. I can't see how having Turbo Boost disabled helps when hyperthreading is enabled.
"Intel Turbo Boost is a technology implemented by Intel in certain versions of its processors that enables the processor to run above its base operating frequency via dynamic control of the processor's clock rate."

In essence it is a built-in overclocking mechanism. Most newer processors have two speeds specified, like in this example:

Clockspeed: 3.6 GHz
Turbo Speed: 4.0 GHz
No of Cores: 4 (2 logical cores per physical)
Typical TDP: 84 W

PassMark.com list these specs (along withe benchmarks) for most CPUs, old and new.

PK
Yes but there is also something called all-core turbo. My CPUs run at 2.3 gigahertz with all 72 threads. It makes no difference whether or not I run 36 physical (HT OFF) - or all 72 threads with hyperthreading enabled. The frequency is always 2.3 or 2.28 gigahertz to be precise. And I DO not disable turbo boost to obtain this frequency. So why would disabling turbo boost help me?! Perhaps four-core CPUs are different. Disabling turbo boost with hyperthreading enabled would do nothing for me. Perhaps I am missing something here as I also use Process Lasso to do all of my optimizations for me automatically. Interesting stuff.

Milos
Posts: 3387
Joined: Wed Nov 25, 2009 12:47 am

Re: Some hyperthreading results

Post by Milos » Tue Dec 13, 2016 2:39 am

APassionForCriminalJustic wrote:Yes but there is also something called all-core turbo. My CPUs run at 2.3 gigahertz with all 72 threads. It makes no difference whether or not I run 36 physical (HT OFF) - or all 72 threads with hyperthreading enabled. The frequency is always 2.3 or 2.28 gigahertz to be precise. And I DO not disable turbo boost to obtain this frequency. So why would disabling turbo boost help me?! Perhaps four-core CPUs are different. Disabling turbo boost with hyperthreading enabled would do nothing for me. Perhaps I am missing something here as I also use Process Lasso to do all of my optimizations for me automatically. Interesting stuff.
Gee you don't even understand simplest possible architectural difference between Xeons and i7's, do you?
Xeon's (real ones not fake Skylake ones) have locked multiplier (on lower value than max) to preserve TDP since they have too many cores to sustain the spec. That's why their all core boost is lower than max boost.
Your CPU is engineering sample (in case you haven't already realized why it was so cheap) which means it's poor process technology far below spec that has multiplier locked to even lower value. E.g. nominal all core boost for E5-2698 v3 is 2.8GHz not 2.3GHz as yours, or in another words yours is more than 20% slower. That's why you see no difference in maximum multiplier when you turn off HT.

Btw. there is even an unofficial version (QS only) of your CPU with somewhat unlocked (higher) multiplier E5-2698A v3 and in order to allow it Intel raised TDP to 165W from 135W. This version has all core boost of 3.2GHz (equal to max boost).

Post Reply