i7 Hyper-threading

[mwyoung]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores. This is true with NPS and timed test positions. As this is my first i7 cpu, does Hyper-threading with the i7 help instead of hurt chess engines performance. This was not true in the past. In my testing I am clearly faster with Hyper-threading turned on running my chess engines.

Example Fritz 12

Fritz benchmark 4 cores. 5380 Nps 11.21

Fritz benchmark with HT 8 logical cores. 6995 Nps 14.57

[LaurenceChen]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores. This is true with NPS and timed test positions. As this is my first i7 cpu, does Hyper-threading with the i7 help instead of hurt chess engines performance. This was not true in the past. In my testing I am clearly faster with Hyper-threading turned on running my chess engines.

Example Fritz 12

Fritz benchmark 4 cores. 5380 Nps 11.21

Fritz benchmark with HT 8 logical cores. 6995 Nps 14.57

What happens when you run engine vs engine with hyperthreading on?
Which engine gets the true CPU and which one gets the logical one?
Assuming that you are not going to run engine vs engine in your laptop, then hyperthreading should not hurt the performance of the chess engine.

[mwyoung]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores. This is true with NPS and timed test positions. As this is my first i7 cpu, does Hyper-threading with the i7 help instead of hurt chess engines performance. This was not true in the past. In my testing I am clearly faster with Hyper-threading turned on running my chess engines.

Example Fritz 12

Fritz benchmark 4 cores. 5380 Nps 11.21

Fritz benchmark with HT 8 logical cores. 6995 Nps 14.57

What happens when you run engine vs engine with hyperthreading on?
Which engine gets the true CPU and which one gets the logical one?
Assuming that you are not going to run engine vs engine in your laptop, then hyperthreading should not hurt the performance of the chess engine.

I am running Ponder ON with HT ON. Rybka 4 with 4CPU. Houdini with 4 CPU. I am running each engine with 2GB of hash.

The engines are working flawlessly with HT On with the i7. This is great, I am getting better NPS with ponder on using 4 cores for each engine. Then I can get on my desktop PC Q6600 with ponder off and 4 cores. Looks like Intel has got HT working much better with their i7 Chip.

[PauloSoare]

I read in Rybka Forum that Rybka has no
advantage with Hyper-threading.

[stevenaaus]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores.

It's interesting you've found such a large apparent speed-up by enabling HT. In the P4, hyperthreading was quite a non-feature (ie, it made very little difference). But i don't know much about the latest i7 architecture, so maybe it's more than another Intel gimmick.

Have you tried running super pi. It's a fairly decent CPU benchmark.

[mwyoung]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores.

It's interesting you've found such a large apparent speed-up by enabling HT. In the P4, hyperthreading was quite a non-feature (ie, it made very little difference). But i don't know much about the latest i7 architecture, so maybe it's more than another Intel gimmick.

Have you tried running super pi. It's a fairly decent CPU benchmark.

Must be more then a gimmick, I have had HT CPU's in the past. None worked like this i7 Mobile chip HT.

I am running Rybka-4 4 cores, and Houdini 4 cores with ponder on. Rybka is hitting 230 Knps, Houdini is hitting 3700Knps. The i7 has clocked itself to 2Ghz (132x15.0).

[stevenaaus]

Yeah, since after the P4, Intel's CPUs have been pretty good.

And thinking about it, Super Pi probably isn't a good test for multithreading anyway.

[hgm]

What happens when you run engine vs engine with hyperthreading on?
Which engine gets the true CPU and which one gets the logical one?

There is no 'true CPU' with HT on. Just two logical ones (per core).

[Houdini]

After loading my new i7 laptop with a few chess engines. I am finding Rybka 4, Fritz 12, and Houdini are running faster using 8 logical cores. This is true with NPS and timed test positions. As this is my first i7 cpu, does Hyper-threading with the i7 help instead of hurt chess engines performance. This was not true in the past. In my testing I am clearly faster with Hyper-threading turned on running my chess engines.

Example Fritz 12

Fritz benchmark 4 cores. 5380 Nps 11.21

Fritz benchmark with HT 8 logical cores. 6995 Nps 14.57

Mark,

I've never experienced any useful improvement from hyper-threading for Houdini. Even if the nps is slightly higher, the overhead of the additional threads could very well reduce the actual playing strength.

Could you try the "autotune" feature of Houdini? This feature was intended for picking the best Split_Depth parameter, but it also functions as an accurate benchmark for multi-thread node speeds.

Double-click on the Houdini executable to open it in a console window.
Then enter the following commands to "autotune" for 4 threads.

Code: Select all

setoption name threads value 4
autotune

Houdini will run for about 10 minutes analyzing a number of positions for different values of split_depth. Make sure you're doing nothing else on your computer, just leave it running for 10 minutes. At the end you'll get a summary of the node speeds.

Then repeat the procedure for 8 threads.

Code: Select all

setoption name threads value 8
autotune

What results do you get?

[yanquis1972]

fwiw, with 8 cores my best result came w/ a split depth of 12 (by about ~3-4% over the default of 10). my guess is the best result for four cores is 10 but i'll let someone else confirm this.