No Turbo Boost on the new Sandy bridges?? 13" i7

[aznguyen316]

So I came across this article...
stating throttling and no turbo boosting. Well the downclocking was there, but it didn't throttle when under load as I ran Intel Burn Test at maximum and it stayed clocked at 2.7Ghz (27x multiplier at 99.8MHz bus speed (??).

When just idling it's at 798Mhz 8x multiplier, that's fine with me since it will go up when it needs. But I don't see the any sort of turbo boosting at maximum load. I'm missing out on the 3.2Ghz at dual core boosting and 3.4 Single? It stays solid at 2.7Ghz though but I'm kind of bummed there's no boosting to 3.2Ghz as that's pretty good jump and with Turbo 2.0 that could help a bit. It was one of those things I looked at before purchasing - was turbo capabilities.

I'm just really curious if other new machines are turbo boosting? The i5 in the 13" as well as any of the i7" quads. Please post results, thanks!

Tests were done in Windows 7 Ult 64bit.

Intel Burn Test 2.5

CPU-z

Intel Turbo monitor

I checked Apple.com to see if maybe they don't advertise turbo boosting b/c this may have been disabled. But it's advertised. Pic linked. Also I ran a intel turbo check program and it shows TURBO is NOT DISABLED. So it should be kicking it...

 

[computertechno]

I will test on my 15"

 

[cparktd]

Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

Yep. Up to 3.4.

http://www.apple.com/macbookpro/features.html#processor

 

[AdamRock]

It says 3.4Ghz

 

[Gav2k]

Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

Turbo boost only boosts the core speed when all 4 cores are not active. If it's using all the cores it won't up the speed!

 

[aznguyen316]

Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

Turbo boost only boosts the core speed when all 4 cores are not active. If it's using all the cores it won't up the speed!

But Turbo 2.0 shows it can boost EVEN QUADs and DUALs? So even if say my dual core, both are being used, it should boost up to 3.2Ghz.. and single to 3.4Ghz. I'll test single threaded burn test then.

http://www.intel.com/technology/product/demos/turboboost/demo.htm?iid=tech_tb+demo

as show there.. it will turbo even when all cores are active. At least in Turbo 2.0.

 

[gnasher729]

But Turbo 2.0 shows it can boost EVEN QUADs and DUALs? So even if say my dual core, both are being used, it should boost up to 3.2Ghz.. and single to 3.4Ghz. I'll test single threaded burn test then.

The advertised speed (dual core 2.3 or 2.7, quad core 2.0, 2.2, 2.3) is the speed that you get when all cores (two or four) are running all the time. If you run a "burn test" that presumably uses all cores to the max, you won't get more than the advertised speed.

You'll get higher speed if you don't use all cores, or if you use all cores for a short time only.

 

[brentsg]

For turbo to work it's got to meet certain power and thermal characteristics. I don't think burn tests are the best choice.

 

[Gav2k]

Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

You have to remember the boost is also based on the current operating temp and available power to the processor. Generally u will see the boost occur under single core use not dual. More cores more heat less chance of boost.

 

[aznguyen316]

For turbo to work it's got to meet certain power and thermal characteristics. I don't think burn tests are the best choice.

Yet it seems to work for desktop users to get their turbo boost to kick in?

here's another example of Turbo 2.0 (it's different than the Arrandale Turbo!) at work. Even if ALL FOUR cores are at work, there is still turbo boosting from 2.0 quad to 2.6 on all quad cores which is what is supposed to happen on 4 cores for the i7-2630QM, 2.8GHz on dual and 2.9Ghz on Single.

http://www.youtube.com/watch?v=4gMKumOA9I8

Mine is not boosting to 3.2Ghz on both cores. I ran Cinebench in Windows, still just at 2.7Ghz No dual core boosting. And YES, it should boost on two cores to 3.2Ghz still as seen in the i7-2630qm 2.6 on four cores test in the youtube video.

Anybody test theirs?

 

[brentsg]

Yet it seems to work for desktop users to get their turbo boost to kick in??

I can stress test my desktop, do video encodes, whatever and it never exceeds 53C. You simply can't say the same about a mobile CPU. So the desktop CPU will be operating in turbo-friendly conditions and the mobile CPU will not.

 

[aznguyen316]

I can stress test my desktop, do video encodes, whatever and it never exceeds 53C. You simply can't say the same about a mobile CPU. So the desktop CPU will be operating in turbo-friendly conditions and the mobile CPU will not.

Alrighty then, well the youtube video above with a Sandy bridge quad mobile is boosting using cinebench. Mine isn't. So that's why I'm wondering if other users are turbo boosting or not.

 

[DustinT]

Generally speaking, I understand that Sandy Bridge should allow boosting for mobile applications. Apple may have chosen to limit that feature but their own advertising mentions it at a specific speed, 3.4 ghz.

So, the real question is, does it work!

Note as well, both the CPU and GPU have turbo boost enabled by Intel so you should get maximum performance possible under a given thermal environment.

To the OP, did you test on battery or on A\C? That could easily make a difference.

 

[aznguyen316]

A/C definitely. Yeah and the i7 turbo clocks the graphics up a bit higher than the i5 too so whatever little difference that makes would've been nice.

But like I said, it's a simple test, just waiting for someone to try it out.

 

[LunaC]

.

To the OP, did you test on battery or on A\C? That could easily make a difference.

Since throttling has been known to be restricted with smaller power supplies, I think I need to compare with the stock PS and the one from my MBP17. Anyone done this yet?

-C

EDIT: well at least on Geekbench 32, A/C and battery made no difference...and AC was with my larger PS

 

[Evil Spoonman]

I have had difficulty reading turbo levels in the past on multiple Intel machines. This Mac Pro included. One of the CPUs will show turbo up to 2.93GHz, the other shows stock bins all the time no matter what. Yet if I turn off Turbo on that CPU all of my benchmarks drop considerable. So the CPU is clearly turboing properly, it just isn't displaying properly.

Anecdotally, I sort of feel that the software that reads this stuff isn't really that reliable.

 

[dusk007]

Did you check the temps.
Turbo 2.0 should work the best only for short boost on more or less idle system that is comparably cold.
Maybe you should try some test that gives you a displays constant frame rate like an encode session. If Turbo 2.0 works. The framerate should start high and later after the machine runs hot drop by 15%. If that fps drop happens it means Turbo works and the clock speed read out is simply wrong.