Any Core i7 benchies with both HT and TurboBoost Disabled?

[Mechromancer]

I'm curious to see Core i7 benchmarks with both HT and TurboBoost disabled. Do any of you have any tests or links to some results?

[ben805]

I'm curious to see Core i7 benchmarks with both HT and TurboBoost disabled. Do any of you have any tests or links to some results?

can't take any screenshots as I'm at work now, but when I tested with the HT on and off, it made no difference in converting 1GB of RAW files. Winrar 3.80 bench shows higher score with HT on, Sandra 2009 bench shows higher cpu benchmarks with HT on, no difference in superpi 1m, cpumark and 3Dmark Vantage shows higher with it off, Photoshop CS4 runs better with HT on, gaming in COD World At War has slightly higher FPS with it off, well that's pretty much all i can remember for now.

[Ancient_1]

The Tech Report did on the 965 when they did their review of the new PII's

[ReverendMaynard]

can't take any screenshots as I'm at work now, but when I tested with the HT on and off, it made no difference in converting 1GB of RAW files. Winrar 3.80 bench shows higher score with HT on, Sandra 2009 bench shows higher cpu benchmarks with HT on, no difference in superpi 1m, cpumark and 3Dmark Vantage shows higher with it off, Photoshop CS4 runs better with HT on, gaming in COD World At War has slightly higher FPS with it off, well that's pretty much all i can remember for now.

I loose 3000 points in Vantage with HT off.

[ben805]

I loose 3000 points in Vantage with HT off.

is that just the CPU bench alone or overall score with the GPU and other features?

[jcool]

I'm curious to see Core i7 benchmarks with both HT and TurboBoost disabled. Do any of you have any tests or links to some results?

I can do any test you like at any frequency between 3 and 4,8Ghz on my 975.
Free multiplier, so ramspeed can be synced as well if you'd like to compare it to a PII
Unfortunately the only PII I have is a sucky 940 that can hardly do 3,6Ghz (needs 1,5V..) and also it sits on a NV 780a SLI mobo that uses DDR2, so not ideal for comparison.

[Mechromancer]

Thank you for all the info, but I'm asking for you to post benchies with both HT and TurboBoost Disabled together, not either or. TurboBoost overclocks and I want to see unskewed results. Sure, TurboBoost and HT are features ofthe Core i7, but my point is to see performance without these enhancements.

[jcool]

See edit... tell me which benches you'd like to see at what CPU/memspeed and I'll do a quick run.
Maybe 3Dmark, Cinebench and ???

[Mechromancer]

^I'd like to see anything you want to run at stock CPU speed/Core i7 920 speeds. I have 3DMark06, Cinebench R10, and POVRay 3.7 Beta. I'm actually just as interested in single-threaded performance as I am in multi-threaded. Also if you have any game benchmarks like Crysis or Fear, go ahead and toss that in there as well.

Do what you can. I really appreciate it!

[jcool]

Stock 920, that's 2666Mhz right? Oh boy....
I'll run CB10, 3D06 and Crysis at stock 975 speeds (3,33Ghz) so I won't have to wait that long, ok? Never done POVRay, but if you can explain to me how one does a benchmark there I'll do it.
How about ramspeed? I'd suggest around 1600 because that's what the i7 officially supports.

[Mechromancer]

http://www.povray.org/download/. Go there to download POVRay. Get version 3.6 first and have it installed before you install 3.7. 3.7 needs the base files from the 3.6 installation to run. Then just select the Render menu and choose a single or multi-threaded benchmark.

[Monstru]

Some tests with HT on and off - i965 ES

[Mechromancer]

Some tests with HT on and off - i965 ES

Ok, they have HT off, but is TURBO BOOST off too? I'm asking for both HT and TurboBoost off AT THE SAME TIME benches.

[jcool]

I'm working on it... halfway done, even tho I can't find Crysis any more, lol.

[ReverendMaynard]

What's the point in disabling Turboboost though? It's just another multiplier.

[jcool]

Ok then here we go. All stock setting on the 975 (133x25=3,33Ghz), DDR3-1600 C8-8-8-28 Tri-Channel Ram. Vcore 1,04/1,08V loaded (HT off/on).

GPU: HD4870X2 @ 820/930
OS: WIn Server 08 x64

Left: HT off, Right: HT on (if you have a big screen.. otherwise it's HT off top, HT on bottom)

POVRay 3.7 x64 single thread:



POVRay 3.7 x64multi thread:



Cinebench R10 x64:



3DMark 06:



--------------------------------------------------------------

Analysis of results: Enable HT for render jobs, disable for gaming.
However, 3D06 results are misleading. Once the CPU gets up to speed (>4Ghz), the score with HT enabled will be a lot better than the one w/o HT.

Lastly, here's what I got on a CB10 max run

[Monstru]

Ok, they have HT off, but is TURBO BOOST off too? I'm asking for both HT and TurboBoost off AT THE SAME TIME benches.

From what I remember Turbo-Boost was of too, but I will have to check to be sure.

[Mechromancer]

AWESOME! Thank you so much jcool.

[Ancient_1]

Here are my results from my 920 with HT and Turbo disabled (I have screen shots if needed).

XP32 920@2660

All 4 cores
CPU time used: kernel 0.41 seconds, user 416.31 seconds, total 416.72 seconds.
Elapsed time 105.67 seconds, CPU vs elapsed time ratio 3.94.
Render averaged 2480.73 PPS (629.07 PPS CPU time) over 262144 pixels

Single core
CPU time used: kernel 0.13 seconds, user 409.80 seconds, total 409.92 seconds.
Elapsed time 530.30 seconds.
Render averaged 494.33 PPS (639.50 PPS CPU time) over 262144 pixels

[Mechromancer]

My Phenom II 940 at 3.2Ghz Core/2.4Ghz NB, 8GB-DDR2@800mhz does-

Multi-threaded: 2775.36 PPS (712.73 PPS CPU time)
Single-threaded: 727.60 PPS (727.29 PPS CPU time)

I can now accurately judge the performance gains my overclocking has given me vs. Core i7 sans HT. That 965 spanks booty in R10.



Later today I will setup some Overdrive 3.0 profiles to clock a single core past 3.5Ghz for single-threaded benching. That is AMD's equivalent to Turbo Boost for now. Better than nothing I guess. Also notice how my fastest multi-threaded run in R10 is exactly 2.509918 times faster than my fastest Opteron 185 run. The upgrade was well worth it for the money.

[Ancient_1]

Here is my results from CB 10 and POV @ 3200 with HT off done in XP32 which should be a good comparison to your PII.

CS 10 920@3200 HT @ Turbo off


POV-Ray v3.7 920@3200 HT @ Turbo off




I also did one @ 3.2 with HT on.

[Mechromancer]

Those are amazingly interesting results at 3200mhz. You push OpenGL way faster than my system and I even have a 4870 vs your 4670. That part fascinates me as well.

[Ancient_1]

I see you are on a 64bit OS I would guess that is what makes uor results quite different. If I had a 64bit OS installed I would of done it all there, but I have driver issues with my capture cards (which is my main use) and XviD is aprox 20% slower in Vista I have no reason to install. Between this system and my secondary I have 7 HD tuners and record an average of 8hr and up to 12 of HDTV each night and edit and convert to XviD each day.

Here is a shot of my work folder.

[SoulsCollective]

It may be unknown to most (especially Intel folks) that if you disable both TurboBoost and Hyper Threading on a Core i7, you get the same or slightly less performance than a comparable Phenom II.

[jcool]

Yeah well, we did disprove that theory now I think