4 Years Later... Same Processor???

[oo7ml]

Hi,

I have a Mid 2011 MBA. It has a 1.7 GHz Core i5 processor.

I just bought a new / refurbished Early 2014 MBA from the Apple Store. It has a 1.4 GHz Core i5 processor.

Does this mean that my 4 year old laptop has a faster processor OR am i missing something here

 

[ryanmillercg]

Your suspicion is correct.. you're missing something. CPU performance is dictated by far more than just clock speed.

Take a glance a this performance chart, you'll notice your new computer is much faster in single and multicore performance. http://www.primatelabs.com/blog/2014/05/macbook-air-performance-may-2014/

I'm sure you've also noticed that the battery life is a great deal better.

The integrated graphics in the 2014 mba are a huge step forward compared to the 2011.

The newer chipset also allows for things like faster RAM, faster SSD, and much more.

 

[motrek]

Hi,

I have a Mid 2011 MBA. It has a 1.7 GHz Core i5 processor.

I just bought a new / refurbished Early 2014 MBA from the Apple Store. It has a 1.4 GHz Core i5 processor.

Does this mean that my 4 year old laptop has a faster processor OR am i missing something here

I don't even know why they publish those GHz numbers because they have very little to do with anything.

Both of these chips will run anywhere between 800MHz and ~3GHz depending on thermal headroom, i.e., how hot they are.

The newer chips run cooler so they can run at max speed longer (basically indefinitely) whereas your older chip will heat up and run slower after not too long.

 

[2IS]

I don't even know why they publish those GHz numbers because they have very little to do with anything.

Both of these chips will run anywhere between 800MHz and ~3GHz depending on thermal headroom, i.e., how hot they are.

The newer chips run cooler so they can run at max speed longer (basically indefinitely) whereas your older chip will heat up and run slower after not too long.

The GHz rating means a whole lot when comparing it to processors of the same family and generation. It's in fact pretty much the only way to know which is the faster processor. It's when you start comparing clock speeds of processors several generations apart where it becomes a rather useless metric.

 

[oo7ml]

Thanks for clarifying.

Would it not make more sense for Intel to change the name of the processor, or at least add some variation?

 

[Significant1]

Thanks for clarifying.

Would it not make more sense for Intel to change the name of the processor, or at least add some variation?

They have. If you click on any family here:
http://ark.intel.com/
You will see that the full-names includes a 4 digit number. And here is a description of how to decode it:
http://www.intel.com/content/www/us/en/processors/processor-numbers.html

but Apple is dumping it down, to make things Simple for the customer and have them focus on the technicalities that Apple's believe they have an edge on, compared to competitors.

 

[keysofanxiety]

Thanks for clarifying.

Would it not make more sense for Intel to change the name of the processor, or at least add some variation?

Intel do have a lot of changes and their naming scheme is okay to understand. For instance you'll have i3-2xxx, or i3-3xxx to note which generation it is (2nd or 3rd). The issue is that Apple don't publish the exact model - they just say i5 1.4GHz without hinting which generation or model it is. Apple have occasionally underclocked some CPUs as well, so the clockspeed may cause some ambiguity.

The best thing to do is search for a model of Mac at www.everymac.com, as that will give you the exact CPU model for that machine. Also you can get an idea of how powerful the processor is by searching for the model at www.cpubenchmark.net (as a rule of thumb, higher number is better performance).

Hope this helps for future!

 

[oo7ml]

Thanks guys, this is really helpful information.

 

[motrek]

The GHz rating means a whole lot when comparing it to processors of the same family and generation. It's in fact pretty much the only way to know which is the faster processor. It's when you start comparing clock speeds of processors several generations apart where it becomes a rather useless metric.

You would be right if the processors in question actually ran at the published speed (GHz).

But they don't. As I said in my post, their speed is constantly changing, from 800MHz to ~3GHz and almost anywhere in-between.

So saying that one chip is 1.3GHz and the other one is 1.7GHz is basically irrelevant since they won't be running at those speeds.

My guess is that the rated clock speed is the minimum speed that a chip will run when it's under sustained load and has the amount of cooling that Intel specifies. So this would be relevant when comparing two 2011 MBAs, because those laptops "overheat" pretty quickly and a higher-rated processor would run faster, all things being equal. But the 2013+ MBAs have excess cooling so their rated clock speeds are irrelevant.

 

[2IS]

You would be right if the processors in question actually ran at the published speed (GHz).

But they don't. As I said in my post, their speed is constantly changing, from 800MHz to ~3GHz and almost anywhere in-between.

So saying that one chip is 1.3GHz and the other one is 1.7GHz is basically irrelevant since they won't be running at those speeds.

My guess is that the rated clock speed is the minimum speed that a chip will run when it's under sustained load and has the amount of cooling that Intel specifies. So this would be relevant when comparing two 2011 MBAs, because those laptops "overheat" pretty quickly and a higher-rated processor would run faster, all things being equal. But the 2013+ MBAs have excess cooling so their rated clock speeds are irrelevant.

That's a combination of speed step and turbo at work. Processors have both a base clock rating and a turbo rating. Some don't have turbo at all so it's just the base clock. The speeds below the base clock are not relevant since it only runs at those speeds when under light or idle loads. If you don't know what to look for, I suppose it could seem useless to you, but it's really not that complicated.

 

[motrek]

That's a combination of speed step and turbo at work. Processors have both a base clock rating and a turbo rating. Some don't have turbo at all so it's just the base clock. The speeds below the base clock are not relevant since it only runs at those speeds when under light or idle loads. If you don't know what to look for, I suppose it could seem useless to you, but it's really not that complicated.

Okay great, you're basically just repeating what I've said, which is that these chips run at a wide variety of speeds depending on load and thermal state.

You've made no case that the "base" clock speed is relevant to anything, at least not for the MBAs, which seem to have more cooling than Intel specifies.

 

[2IS]

Okay great, you're basically just repeating what I've said, which is that these chips run at a wide variety of speeds depending on load and thermal state.

You've made no case that the "base" clock speed is relevant to anything, at least not for the MBAs, which seem to have more cooling than Intel specifies.

I don't think you understand how it works to be honest. I'm not saying the samething you are. The fact you think I am suggests even more so that you don't understand the significance of clock speeds.

A CPU under load will run anywhere between its base clock and turbo clock depending on the load type. A CPU with a higher base clock will perform faster than one with a lower base clock. Higher base clock CPU's typically also have a higher turbo ceiling. Just because there is a spectrum of clock speeds they can run at doesn't mean knowing what those speeds are is useless unless you don't know what you're looking at. By your understanding of how it works, a 2GHz CPU with a 2.5GHz turbo and a 3GHz CPU with a 4GHz turbo is not something worth knowing. Except that it is.

 

[Orr]

Kind of suspect the OP is trolling. Why would you spend such good coin on a big ticket item that you think may very well be slower than your 4 year old existing machine. A one minute google search could've answered such a seemingly important question. Very baffling/suspect logic.

 

[torana355]

Clock speed is a bit like RPM in an engine, you will get more power as the RPM rises but a v8 doing 6k rpm will have more power then a small 4 cylinder engine doing the same rpm. Basically your 2011 mac is a 4 cylinder and the new 2014 model is a v8 Some people are mistaken by thinking its only the clockspeed that determines how fast a CPU is, the 2011 cpu will do less work per cycle compared to the 2014 model which means it can run at a lower GHZ and still be as fast.

 

[motrek]

I don't think you understand how it works to be honest. I'm not saying the samething you are. The fact you think I am suggests even more so that you don't understand the significance of clock speeds.

A CPU under load will run anywhere between its base clock and turbo clock depending on the load type. A CPU with a higher base clock will perform faster than one with a lower base clock. Higher base clock CPU's typically also have a higher turbo ceiling. Just because there is a spectrum of clock speeds they can run at doesn't mean knowing what those speeds are is useless unless you don't know what you're looking at. By your understanding of how it works, a 2GHz CPU with a 2.5GHz turbo and a 3GHz CPU with a 4GHz turbo is not something worth knowing. Except that it is.

I said knowing the base clock speed is unimportant. I never said that all clock speeds are irrelevant.

For MBAs, knowing the turbo boost speed is the most important thing, because recent models have enough cooling to run at full turbo boost indefinitely.

So if the processor is going to run at full turbo boost speed when under load, tell me again why it's so important to know the base clock speed, which it's basically never going to run at?

----------

Clock speed is a bit like RPM in an engine, you will get more power as the RPM rises but a v8 doing 6k rpm will have more power then a small 4 cylinder engine doing the same rpm. Basically your 2011 mac is a 4 cylinder and the new 2014 model is a v8 Some people are mistaken by thinking its only the clockspeed that determines how fast a CPU is, the 2011 cpu will do less work per cycle compared to the 2014 model which means it can run at a lower GHZ and still be as fast.

Not really. There isn't a huge difference between the 2014 processors and 2011 in terms of architecture.

Most of the performance gain comes from the fact that the 2014 chips require less power, which means they require less cooling, which means the MBA has enough cooling that a 2014 chip can run at full turbo boost (~2.7GHz) indefinitely, whereas the 2011 model has to drop back to its base clock speed (~1.7GHz) after a minute or so.

 

[2IS]

I said knowing the base clock speed is unimportant. I never said that all clock speeds are irrelevant.

For MBAs, knowing the turbo boost speed is the most important thing, because recent models have enough cooling to run at full turbo boost indefinitely.

So if the processor is going to run at full turbo boost speed when under load, tell me again why it's so important to know the base clock speed, which it's basically never going to run at?

Read up on how turbo works.

http://www.intel.com/content/www/us...ology/turbo-boost/turbo-boost-technology.html

Essentially, processors cannot maintain turbo frequencies indefinitely. They will at some point and depending on the load, drop back down to their base frequency. If it's a high load that loads all your threads, you may never even go above base clocks at all. If you're encoding/transcoding a movie that loads up all the cores/threads and takes quite some time to do it, you will be running at your base clock essentially 100% of the encoding processes.

If you're running emulation software that runs entirely on a single core, you will be able to maintain turbo frequencies for a much longer period of time.

The base clock is the frequency the processor WILL be able to maintain indefinitely regardless of load, provided it's within thermal specifications.

 

[torana355]

Not really. There isn't a huge difference between the 2014 processors and 2011 in terms of architecture.

Most of the performance gain comes from the fact that the 2014 chips require less power, which means they require less cooling, which means the MBA has enough cooling that a 2014 chip can run at full turbo boost (~2.7GHz) indefinitely, whereas the 2011 model has to drop back to its base clock speed (~1.7GHz) after a minute or so.

Well what i said is actually correct, GHZ is only one part of the equation how much work that gets done per cycle is also very important. So you can have two 1.6ghz processors with one being much faster at the exact same GHZ. A great example is my old 2008 iMac is 2.8ghz but my MBA with only 1.7GHZ is actually faster.

 

[oo7ml]

Kind of suspect the OP is trolling. Why would you spend such good coin on a big ticket item that you think may very well be slower than your 4 year old existing machine. A one minute google search could've answered such a seemingly important question. Very baffling/suspect logic.

If you had read my comments carefully you would have seen that i spoke to Apple before my purchase and queried the same.

I was told that the new processor was a Duo, and my old one wasn't.

I also know that Apple, like any other manufacturer release higher spec machines with each new model release... however i was surprised when i clicked into 'About this Mac'... hence the reason of my post.

 

[motrek]

Read up on how turbo works.

http://www.intel.com/content/www/us...ology/turbo-boost/turbo-boost-technology.html

...

I'm getting tired of you telling me that I don't know what I'm talking about when it's actually the reverse.

Do yourself a favor and download Intel Power Gadget. It will show you a graph of your CPU clock speed. You can actually see how fast it's running vs. speculate about it based on your understanding of a web page written by Intel's marketing department.

FYI, Power Gadget has a bit of a design flaw, in that it only polls the clock speed once per second. If you're maxing out one core and it polls, the clock speed will drop down to the max dual-core turbo boost speed. So the graph will never show you the max single-core speed. But if you measure the performance of a task run single-core vs. dual-core, you will see that the CPU is really hitting the max single-core speed.

 

[2IS]

I'm getting tired of you telling me that I don't know what I'm talking about when it's actually the reverse.

Do yourself a favor and download Intel Power Gadget. It will show you a graph of your CPU clock speed. You can actually see how fast it's running vs. speculate about it based on your understanding of a web page written by Intel's marketing department.

FYI, Power Gadget has a bit of a design flaw, in that it only polls the clock speed once per second. If you're maxing out one core and it polls, the clock speed will drop down to the max dual-core turbo boost speed. So the graph will never show you the max single-core speed. But if you measure the performance of a task run single-core vs. dual-core, you will see that the CPU is really hitting the max single-core speed.

Doesn't matter if you're getting tired of it. It's obviously true. That link is from Intel themselves. You clearly don't know how turbo works. If you're getting tired of it, don't reply. If you do reply and maintain that base clock isn't important, I will maintain you don't have a clue what you're talking about.

I've played with turbo values, custom TDP settings, overclocking, etc etc for many years, I am WELL aware of how turbo works. You, on the other hand, are not. The fact that you think temperature is the only factor that dictates turbo is proof enough. Did you read the link I posted from Intel? It also says you're wrong.

Availability and frequency upside of Intel Turbo Boost Technology 2.0 state depends upon a number of factors including, but not limited to the following:

•Type of workload
•Number of active cores
•Estimated current consumption
•Estimated power consumption
•Processor temperature

 

[motrek]

If you had read my comments carefully you would have seen that i spoke to Apple before my purchase and queried the same.

I was told that the new processor was a Duo, and my old one wasn't.

I also know that Apple, like any other manufacturer release higher spec machines with each new model release... however i was surprised when i clicked into 'About this Mac'... hence the reason of my post.

Disappointing answer from Apple. The 2011 MBAs were Core i5s and i7s, not Core 2 Duos. :/

As I've explained in my earlier posts, the idea of measuring these new Core processors with one clock speed number is misguided. Rest assured that a 2015 MBA is significantly faster than your 2011 MBA.

----------

Doesn't matter if you're getting tired of it. It's obviously true. That link is from Intel themselves. You clearly don't know how turbo works. If you're getting tired of it, don't reply. If you do reply and maintain that base clock isn't important, I will maintain you don't have a clue what you're talking about.

Okay, I'll bite. I read the web page and it agrees with everything I've said.

It says "amount of time the processor spends in [turbo boost] depends on the workload and operating environment."

It goes on to specifically say that it will turbo boost as long as power consumption and temperatures permit.

I have maintained from the beginning that the MBA has enough thermal headroom to maintain max turbo boost speeds indefinitely. And it obviously has enough power headroom since the laptops haven't changed their power/cooling situation for ~5 years, and chips 5 years ago used way more power and produced way more heat.

If you monitor the CPU temperature of a current MBA under full CPU load, you will see that the fan only spins up to around 4000 RPM with ~70F ambient room temperature. The max fan speed is over 6000 RPM, so clearly there's thermal headroom to spare.

So what I've been saying agrees 100% with your web page.

And regardless of your opinions and "understanding" of turbo boost, it's trivial to download Power Gadget (it's free) and load down your CPU. You can probably confirm everything I've been saying in a matter of minutes. No speculation required.

 

[2IS]

Disappointing answer from Apple. The 2011 MBAs were Core i5s and i7s, not Core 2 Duos. :/

As I've explained in my earlier posts, the idea of measuring these new Core processors with one clock speed number is misguided. Rest assured that a 2015 MBA is significantly faster than your 2011 MBA.

----------



Okay, I'll bite. I read the web page and it agrees with everything I've said.

It says "amount of time the processor spends in [turbo boost] depends on the workload and operating environment."

It goes on to specifically say that it will turbo boost as long as power consumption and temperatures permit.

I have maintained from the beginning that the MBA has enough thermal headroom to maintain max turbo boost speeds indefinitely. And it obviously has enough power headroom since the laptops haven't changed their power/cooling situation for ~5 years, and chips 5 years ago used way more power and produced way more heat.

If you monitor the CPU temperature of a current MBA under full CPU load, you will see that the fan only spins up to around 4000 RPM with ~70F ambient room temperature. The max fan speed is over 6000 RPM, so clearly there's thermal headroom to spare.

So what I've been saying agrees 100% with your web page.

And regardless of your opinions and "understanding" of turbo boost, it's trivial to download Power Gadget (it's free) and load down your CPU. You can probably confirm everything I've been saying in a matter of minutes. No speculation required.

You don't know the power consumption of every type of workload so there's no way anyone can guarantee that turbo frequencies will be maintained indefinitely. In addition, you started out with one value (temp), and now added a second (power, with flawed reasoning on top of it).

So again, you don't know how turbo works and what you said isn't the same thing as what the links says. You're slowly making concessions and modifying your argument, but your position remains incorrect.

There is no speculation going on here. This IS how turbo works.

 

[motrek]

You don't know the power consumption of every type of workload so there's no way anyone can guarantee that turbo frequencies will be maintained indefinitely. In addition, you started out with one value (temp), and now added a second (power, with flawed reasoning on top of it).

So again, you don't know how turbo works and what you said isn't the same thing as what the links says. You're slowly making concessions and modifying your argument, but your position remains incorrect.

And you're making a lot of assumptions about what I do and don't know about CPUs, CPU load, power consumption, clock speed, etc.

I do scientific computing and the software I write runs as many threads as I want at 100% CPU usage. For this reason, I have been carefully monitoring my CPU temperatures, clock speeds, etc. for years. My software uses more power and heats up CPUs at least as much as any other software I've run/tested, including Prime95, Handbrake, and various raytracers.

So it's pretty funny that you're telling me that I don't know about turbo boost and CPU power/temps, and you're arguing from a position of... your interpretation of a web page.

But I don't want to turn this thread into an inane "my qualifications are better than yours" argument.

Seriously, all you have to do is download Power Gadget and check this stuff for yourself. It will take you a few minutes to find out that I'm right. So please do that. The longer you don't do it, the more ridiculous your position becomes.

And if anybody reading this thread has any doubts about anything I've said, I invite you to do the same.

 

[2IS]

And you're making a lot of assumptions about what I do and don't know about CPUs, CPU load, power consumption, clock speed, etc.

I do scientific computing and the software I write runs as many threads as I want at 100% CPU usage. For this reason, I have been carefully monitoring my CPU temperatures, clock speeds, etc. for years. My software uses more power and heats up CPUs at least as much as any other software I've run/tested, including Prime95, Handbrake, and various raytracers.

So it's pretty funny that you're telling me that I don't know about turbo boost and CPU power/temps, and you're arguing from a position of... your interpretation of a web page.

But I don't want to turn this thread into an inane "my qualifications are better than yours" argument.

Seriously, all you have to do is download Power Gadget and check this stuff for yourself. It will take you a few minutes to find out that I'm right. So please do that. The longer you don't do it, the more ridiculous your position becomes.

And if anybody reading this thread has any doubts about anything I've said, I invite you to do the same.

As I've told you, I've played with processors that use Turbo Boost for years. I've already SEEN it at work. I've SEEN it go from it's turbo frequencies down to it's base clocks (using CPUZ on a PC)

You can call my position ridiculous, I'll just point you to the Intel page that says what I'm saying and not what you're saying. You already have the link. Perhaps you should email Intel and tell them they're wrong?

Here's another excerpt

Note: Intel Turbo Boost Technology 2.0 allows the processor to operate at a power level that is higher than its TDP configuration and data sheet specified power for short durations to maximize performance.

But please, continue telling me how you're right and I'm wrong.

You want Intel to advertise their speeds based on nothing more than your own anecdotal evidence instead of how the CPU actually works. That's really what this boils down to.

 

[redheeler]

4 Years Later... Same Design and Display???

The answer to that is yes.