I would be curious to see the benchmarks of a macbook with a new battery versus one with an old battery. Ha. If I buy a new battery maybe I'll do it.

At least we have the option of crippling our macbooks, unlike the majority of other Apple products which don't even allow the user to take a battery out.

This has been the modus operandus for a long time now. I'm surprised you're only now noticing this, after years of Intel macs being made.

This functionality was originally implemented on PC laptops in the original Pentium era, and most PC laptops have gotten over it. But apple is still stuck in the past when it comes to this "misfeature"

[.. En la evaluación comparativa, la desaceleración también es empinada, cayendo un CINEBENCH R10..]

my first-gen macbook (2.0 core duo) and santa rosa macbook pro (2.2 core 2 duo) do this as well. this is not news.

This issue wasfully documented on 2006 by me, reported as a bug to Apple on the first generation MacBook Pro, circa february 2006. Apple answer was to include into the docunentation the issue.

Could you check whether this occurs under Windows/Linux as well, that would me it is integrated into the hardware

MacBook and MacBook Pro: Mac reduces processor speed when battery is removed while operating from an A/C adaptor. If the battery is removed from a MacBook or MacBook Pro, the computer will automatically reduce the processor speed. This prevents the computer from shutting down if it demands more power than the A/C adaptor alone can provide.

http://support.apple.com/kb/HT2332

This isn't too uncommon in general - actually pretty common not only in computers but in electronics in general. As an EE working in the field I have visibility into any medical devices used in the hospital. A striking number of them function like this too - except instead of slewing performance down so the power supply can always meet demand, they assume a battery will always be there for unexpected surges in demand. Now this is a safety important device that is allowing this type of wreckless functioning - Scary indeed. There are times when the battery has failed, or has low charge or just isn't there and people are placed into unsafe situations because everything is running as if a battery was there. You can look at the recalls and notices on the FDA website for a sobering view onto some of the devices that have actually harmed people. For us on the computer side all we loose is some data/work if we shut down so preventing brown/black outs is more of a favor to the users than a necessary feature

I'm surprised that more manufacturers do not do slew performance as this is mandated by a engineering tradeoff seen in many cases when designing the power supply for a dynamic draining electronic device. Either create a power supply that can meet the power demands at all times (much more expensive than 99.99% of the time meeting) or allow a device that is great at providing surge currents (a battery especially lion and NiMh) do it when its necessary and save some money and weight on the power brick/supply. It turns out this is a very costly part of any power electronics so saving money here really can have an impact on the final system cost.

I have a hard time believing no one else in the industry has computers doing this (which the comment above suggests). I would need little convincing that every laptop will do this at some point when the power management software notices it will likely "brown/black" itself out understanding that any low cost laptop will be designed to save money in places it can.

Now apple might have not as an aggressive power usage estimation as others to predict needs thus taking a cleaver approach when slewing performance to guarantee never reaching a brown out situation, but using a less sophisticated power prediction algorithm/system removes overhead every time you actually do have a battery in the system. From my experience in airports, coffee shops, on the train, etc.. I would say most users do usually have a battery installed in the system. From this regard I think it's a good thing to be aware of the battery and a good thing to slew down performance if its not there.

I actually like this design choice as well to save on the overhead the majority of time people do have batteries in there as well as to save on the cost and weight of the power supply. This is a far better situation than a 2lb power supply that is the size of 2 bricks that won't have a performance hit in the times you choose to not install your battery. (unless of course you never take your laptop anywhere....)

Just tested my my MacBook (2GHz Intel Core Duo with 1GB RAM running Leopard) with and without a dead battery and noticed no significant change in performance. I think thats even more of a reason to buy myself a new battery now!

With a dead battery inserted
Rendering (Single CPU): 1008 CB-CPU
Rendering (Multiple CPU): 1860 CB-CPU
Multiprocessor Speedup: 1.85
Shading (OpenGL Standard) : 373 CB-GFX

Without a battery
Rendering (Single CPU): 1007 CB-CPU
Rendering (Multiple CPU): 1857 CB-CPU
Multiprocessor Speedup: 1.84
Shading (OpenGL Standard) : 389 CB-GFX

^^^^ Thanks for posting that. I currently have a bad battery and i was going to run this test, now i don't have to! But i do have to buy a new battery!

So basically Apple underspecified the AC adapter to power the computer when it's running at full speed? I've designed power systems for portable computers in the past and this makes no sense at all. The wall adapter should be able to run the system without battery assist under all possible conditions - including fully loading all the slots to max spec, plugging in USB peripherals that draw max load, etc. Add it all up in a spreadsheet and that's the number you give to the wall adapter vendors for a quote. If you want to cheap out on the adapter, it would make more sense to throttle the CPU if the slots and ports are taking too much power. Designing the system so the adapter can't even run the core system at full speed for an indefinite time is a bad tradeoff. The customer pays a lot of extra money, hundreds of dollars sometimes, to upgrade the CPU - not running it at full speed because you're too cheap to put a good AC adapter in the box is an an outrage. Especially coming from Apple, where they charge a 30% premium for prettier plastic.

It implies that you cannot ever run the CPU at full speed for longer than the battery can help to power it. If you were doing, say, a huge batch of video transcoding, or some 2-week scientific simulation, the battery would run dead over a few hours or days even though it was plugged in. Then once the battery ran dead the CPU would run at a reduced speed (the no-battery speed) until you swapped out or stopped to recharge the battery. Granted, most users would never discover that problem. But anyone using their computer for several hours of intense computation would be stymied as to why their battery was half dead afterward even though the computer was plugged in. Just the cost of tech support calls and the bad PR ought to be enough monetary incentive to go ahead and pay up for a better AC adapter in the first place.

These are my test scores for a C2D 2.4Ghz Penryn Macbook white.

Without battery:
OpenGL standard : 1773 CB
Single Core : 1372 CB
Multiple Core : 2620 CB
Multiprocessor speedup: 1.91x

With battery:
OpenGL standard : 2005 CB
Single Core : 2716 CB
Multiple Core : 5249 CB
Multiprocessor speedup: 1.93x

With a performance increase of more than 50% its a wonder why i've never noticed this before given my heavy use of photoshop.

Absolutely Disgusting.

My wife went to an Apple Store to replace a lost power adapter for our MacBook. Someone in a forum had suggested buying a MacBook Pro adapter at a higher wattage when replacing a MacBook brick. That made sense to me as my last two windows laptops (Toshiba and Lenovo) both sold different bricks for the same machines.

My wife was told by the techs at the Store that she should NOT buy the higher wattage adapter. Seems like they should have applauded the idea. If not, why not?

Toshiba and Lenovo suggested buying the higher capacity for those that wanted faster recharging or had more objects plugged in to USB ports, etc. Both companies made 65- and 90-Watt bricks. I never tested performance differences between the two different bricks.

What does the EE-smart crowd here think of the alternate bricks?

Jonathan7007

I might be delusional but I thought that the AC brick was supposed to be powerful enough to power the computer AND recharge the batteries at the same time. With this approach by Apple it cant do any of those things. By this logic you should recharge your batteries with your computer powered down in order to enjoy full performance .. for the duration of the battery charge afterwards.. insane..

Fantastic... thanks for running the tests on your own, Az and Shane. Az -- if you do buy a new battery, would you mind running the test again? I'm curious to see if it makes a huge difference.

Thanks!
Zach Honig
Product Reviews Coordinator
PCMag Digital Network

In my experience (since the very first Apple notebooks) it is better for a battery to be used (and cycled all the way down regularly) rather than set aside. That's true of many types of batteries (you don't pull the battery out of your car when you're cruising down the highway). I would need to see some reliable empirical evidence to the contrary before I'd bother running my MacBook Pro without battery inserted. Hey, I don't even want the environmental contamination to get on my battery terminals inevitable from things being exposed.

Regardless, Apple is making an environmentally and financially responsible design decision by not providing a dumbvee of a power brick -- it's supposed to be the river, not the reservoir. The power adapter provides more than adequately for overall average operation and charging of the battery. If you're constantly maxing out the processor your fans are no doubt running and that's just going to take more power.

And to the person who claimed to have designed power adapters in the past: I wonder what kind of horrifyingly-costly power vampires your equipment turned out to be. Can you say CO2? Anybody?

I have a 12" Dell laptop that's from around 2004 (1GB RAM, probably low 1.xx GHz processor). I had 2 A/C adapter bricks with different wattages (65 and 90 -ish). The smaller brick could not support the laptop when it was in its docking station with a USB mouse/keyboard and 19" dell monitor. I was trying to extend the desktop to use both the external monitor and the laptop's screen and the graphics card (I think Intel integrated) wouldn't handle it with the limited power. It needed the higher-watt adapter.

Hey Zach,

I have a question for you, apple says that performacne drops incase that the computer uses more power then the adaptor can provide right?
a Macbook uses a 65 W Power Adaptor and a MBP uses a 85W power adaptor

What would happen if you used the MBP's 85W power adaptor on the new Aluminum Macbook? Would you be able to keep performance or would it drop as well?

Please let me know how the tests go!

Hi Kenneth,

This seems to be embedded in the firmware, so I don't believe that using a higher wattage adapter will make a difference. I'll test if time allows, but it won't happen until after the holiday at the earliest.

Thanks!
Zach

Apple notebook probably come with 65W power adapters. If Apple shipped 90 W power adapters like other companies, this problem could be avoided.

I tested my 17 inch MacBook Pro 2.5 Ghz Duo with 4 GB ram, and saw a 50 % drop in performance with the battery ejected.

Open GL 5357/2847
Single CPU 2890/1378
Multiple CP 5479/2598

Extending the battery life and reducing heat is fair reasons for leaving the battery out, but I would rather have performance...