I always thought an i7 was better than an i5, which was better than an i3. Based on the above, is the Acer CPU better (as in more powerful, efficient, and faster) than either of the i7s? And why would one i7 (the SYX) be about 75% faster and more powerful than the Sony i7?
To call it confusing only scratches the surface of the processor nomenclature and configuration. In my opinion, it’s more complex than mere mortals can comprehend.
Sadly, I am but a mere mortal.
However, I will share my priorities, which the average consumer may share when selecting a computer. In so doing, I’m sure I’ll annoy some of the geekier members of my audience.
I’ll also look at a few more things about the processor configuration mix, and compare the three processors you list.
My take on what matters
For the average consumer, I focus processor selection on three things: cores, cycles, and cost.
You’ve probably heard the term “core” being thrown about quite liberally.
In a nutshell, a core is the working part of a processor: the CPU or Central Processing Unit. Think of it as your computer’s engine. The CPU is the part that actually does stuff by following the instructions of the computer program that it’s running.
These days, most processors have more than one core or CPU on a single chip or package. This means the processor can quite literally do two to four things at the same time. In a dual core processor, there are two CPUs. Thus, it can process two sets of instructions at the same time. A quad core has four CPUs.
In Should I get a dual core or a quad core processor?, I recommend that you should get at least a dual core processor. There’s no reason to get a single-core processor any more, and in fact, I believe single-core processors are difficult to find these days.
By cycles, I mean the megahertz or gigahertz (MHz or GHz) measurement associated with a processor. This measures the raw speed of the processor.
While often limited by other reasons, a 3.0Ghz processor is roughly twice as fast as a 1.5Ghz processor. One gigahertz represents a processor’s ability to perform a billion – 1,000,000,000 – operations per second1, and a megahertz is 1000 times less, or one million operations per second. Thus, an older 500Mhz processor would be considered one third the speed of a 1.5Ghz processor.
It’s easy to fall into the “I want as much speed as I can get” trap. However, raw CPU speed plays only a small role in common tasks, like web surfing, video playing, emailing, and the like. You’d be hard pressed to notice the difference between a 1.5Ghz processor compared to 3.0Ghz, if web surfing is all you do.
On the other hand, if you regularly perform CPU-intensive tasks, like video creation, desktop gaming, or other computationally heavy operations, you might well notice.
Naturally, it’s safest to get more than you need, but only within the limits of the next criterion.
The absolute fastest, most capable processors are typically significantly more expensive than models that are only slightly less capable.
For example, on one random machine I examined, the difference between a 2.93Ghz and 3.06Ghz version of the same processor was roughly 10% the price of the entire machine: a 10% cost increase for less than a 5% increase in speed. I’d be shocked if you’d ever notice the difference.
The same goes for the number of cores. Technically, four cores are twice as fast as two, although other limitations – the ability of your software to even use more than one core, for instance – make that a rarity. If the quad upgrade is comparatively inexpensive, it’s an easy one to take. As the incremental upgrade cost rises, the choice becomes less obvious.
Only you know your budget and your needs. When presented with various options in speed, compare the percentage performance gain that you might get to the cost.
I have to stress that my comments are directed at the average consumer. Clearly, if your needs dictate that you need more cores, you need the most cycles. If you’re not sensitive to expense, you might well have different considerations.
And, if you are that average consumer, it’s pretty safe to stop reading here. In my opinion, you have the basics of what you need to decide the configuration of your next computer.
Other processor differences
There are many other factors that come into play besides cores, cycles, and cost.
Traditionally, a processor reads instruction from RAM one at a time. Processors have been getting faster and faster, but the speed of RAM hasn’t always kept pace. To compensate, processors “cache”, loading blocks of RAM into faster on-chip memory.
How CPU caching works is the stuff of both doctoral theses and nightmares; I don’t have the first and I don’t want the second. Suffice it to say that bigger is generally better, and one of the differences between processors, even within the same “i” family, is the size of the cache that it uses.
A faster processor uses more power. That part is pretty simple, but it’s at odds with extending battery life in portable computers.
Many variants of processors are built to use less power at the cost of some of the processor’s other features. For example, it might operate more slowly, or lack other processor features.
Other variations in processors include the type of socket used on the motherboard, the bus (electrical interface used to connect the CPU to the other components on the motherboard), on-chip circuitry to perform graphics operations, the maximum amount of RAM the processor is configured to accept, and probably more.
As you can see, there are a wide variety of combinations, which result in an incredibly wide variety of processors.
OK, but what about the i-mess?
Unfortunately, it’s not as simple as i7s being better than i5s, which are better than i3s. You’ll note that I ignored the whole i-mess in the discussion above. Unless you’re a serious gamer or someone who really needs to tweak every ounce of whatever out of your system, the whole processor nomenclature is fairly irrelevant. (This is where the hardware folks’ heads explode. ) I just can’t advocate the average consumer taking the time to understand it all.
If we do dive in just a little, the first thing to realize is that i3, i5, or i7 isn’t enough to identify all the characteristics of the processor. You really do need the whole processor model number, as you’ve provided. To be honest, I still can’t tell you what makes an i3 an i3 versus an i7. There’s a complex combination of features and technologies that go into the mix.
So, let’s look at some basic characteristics of your three examples.
(Specs are from Intel’s specific product data sheets: i7-950, i7-740QM, i5-650. Cost is “Release Price”, per assorted Wikipedia articles on Intel processors. This is not meant to be accurate. It’s provided for comparison.)
So, yes, that i5-650 is “faster” in GHz, but it has half as many cores as the i7s. (Nope, not all i7s are quad core. It wouldn’t be that simple, but the two that you mention happen to be).
The i7-740QM is clearly designed for the mobile market, given its significantly lower power consumption, which is probably due to its slower speed. (The trailing “M” in the CPU identifier is, indeed, an indicator that the processor is intended for the mobile market.)
Which is better?
That depends on how you define “better”.
Among these three processors, the i7-740QM is probably a better choice for a laptop if you plan to run on batteries frequently. The i7-950 might make for a good workhorse desktop machine, and you might find the i5-650 in a lower-end desktop.
But I use “probably” and “might” on purpose. The devil, as they say, is in the details.
And I’ll bet that even he has a headache by now.