4 Cores? 6 Cores? Do You Care?

An anonymous reader writes "Intel has updated its processor price list earlier today. Common sense suggests that Intel may not care that much anymore whether its customers know what they are actually buying. One new six-core processor slides in between six-core and quad-core processors – and its sequence number offers no clues about cores, clock speed, and manufacturing process. If we remember the gigahertz race just a decade ago, it is truly stunning to see how the CPU landscape has changed. Today, processors carry sequence numbers that are largely meaningless."

Comments

More Cores, More Power

[Anonymous Coward]

Would you want to have a 4 inch penis? Don't you think a healthy 6 or 8 inches might be better?

I have a quad core, which I'm confident will soon become the equivalent of a 4 inch penis. I'll have to upgrade my e-peen when it become affordable.

Seriously though, if you like to game on your computer there is no such thing as too much power.

Re:More Cores, More Power

[thomasinx]

Not necessarily. I could very easily envisage a 6 core system that plays games/handles most tasks worse than a quad core system (emphasis on most). More cores doesn't necessarily mean more power. There are many other statistics to take into account before a judgement can be made, especially when it comes to gaming. Your e-peen is safe for now. Put it to good use.

Re:More Cores, More Power

[kanto]

Pfft.. this reminds me a bit of the jump to DirectX 9 graphics cards; in general the old cards performed better in brute force triangles per second whereas the new ones would perform better at the more technically advanced stuff (read: the things you disable when you're serious about fps). How much use is it having 6 or 8 cores if the program being run only efficiently uses 2 or 4 of them most of the time? It's not like everything can just be multithreaded like that and even if it can, there's bound to be some overhead for doing it.

Re:More Cores, More Power

[Locke2005]

You say that like you have some personal experience with women looking elsewhere...

Re:More Cores, More Power

[cynyr]

because some of us run more than one thing at a time....

Re:More Cores, More Power

[Locke2005]

My wife says it doesn't matter how many cores I have!

Of course, my wife also said she wanted 10 inches... I just told her, "Screw you! I am NOT having it shortened!"

Re:More Cores, More Power

[Grishnakh]

Maybe I'm missing something, but unless the 6-core system is clocked slower than the 4-core one, the 6-core system should outperform it easily in all tasks.

Where it becomes questionable is when you're comparing higher-clocked fewer-core systems to slower-clocked, greater-core systems, because then it comes down to the software you're running and how well it's architected for multiple processes and parallelization. Obviously, a single-threaded application will generally run better on the faster-clocked system, unless that system is being loaded down with a lot of other processes.

Re:More Cores, More Power

[ultranova]

How much use is it having 6 or 8 cores if the program being run only efficiently uses 2 or 4 of them most of the time?

The program? I dunno about you, but I run plenty of programs at once. And having 4 cores means that I have a few on standby whenever I feel like doing input, even when the machine is busy processing stuff.

The real issue I see is memory access. Even with a single core did we run into memory bandwidth/latency bottleneck; with 4-6 cores those are 4-6 times as much. In the long run we have to give up Neumann architechture; it simply can't scale to our needs. A NUMA might be an acceptable compromise, but in the long run we need to change to a dataflow architechture, and that also means a step beyond C/C++ and other Algol-descended languages which have dominated our thinking these past decides.

We need to switch to a system with lots of cores, all with their own local memory, and able to send each other messages. As an added bonus, such a system is also a natural fit for artificial intelligence.

It's not like everything can just be multithreaded like that and even if it can, there's bound to be some overhead for doing it.

True, but most hard problems can be redefined as search problems, and those can be efficiently multithreaded. Our current programming languages just make multithreading a pain, since you have to worry about everything manually.

Re:More Cores, More Power

[Surt]

The 6 core system is slower in non-parallel tasks because the OS has per-core overhead. So all single-threaded tasks get slower as the number of cores rises.

Imagine a task running on an otherwise idle core. It is running as fast as possible, with only OS overhead getting in the way of using 100% of that cpu. Now add more OS overhead to that cpu for core management. There's also cpu (hardware)-level overhead to consider, and the possibility that caches aren't ramped to the same level, so now more cores may be sharing a same-sized cache ... etc.

Lots of reasons for the performance of a single core to drop as the number of cores goes up.

Re:More Cores, More Power

[flappinbooger]

Here's my take. Gaming is not the question here. Gaming is, has, and will be about clock speed and graphics card.

No computer should be built today with less than 2 cores, that much is a given. Anyone who is at all a "power user" should consider a 3 core. AMD's triple cores are really stinkin snappy. Quad core systems? Of course they will become the norm, after a while. Intel and AMD have basically said that since they can't go up in speed they're going to go sideways with cores.

As screens get bigger they will fill up with this feed, that feed, weather, streaming video, multiple website tabs, flash games, a few trojans, printer drivers, chat clients, etc. Lots and lots of things going on at the same time, more cores will make future computing a much more enjoyable experience.

Hey, Intel - shrink the atom core, clock it at 2.5 to 3 ghz, give us 50 atoms on one chip and save yourself some hassles with this iWhatever confusion. Make the model number the number of atom cores.

Regarding TFA? The marketing guy must have been laid off, this numbering system is stupid (intel and amd!). I3 = dual core, I5 is both 2core and 4core, I7 is 4 core, but is now 6 cores. Yeah, that makes sense. Uh huh.

Re:More Cores, More Power

[Cylix]

Even dismissing core overhead and other details you did not mention there are still some wildly important facts to remember.

Only modern games have been designed to take advantage of multi-processor systems. There is also a scaling factor which needs to be considered on an engine by engine basis.

I believe valve only recently made updates to the hl2 engine which optimize for greater then four core systems. While you could vary well purchase a dual proc host and fit it with 4 or 6 core processors the engine may not be able to scale on greater then 2 to 4 threads.

At one point the multi-core support wasn't so hot and I had to disable it on my dual core system. Again, this is an application by application basis and not all experiences will be equal. Certainly, in the future we can assume most newer systems will utilize the hardware better, but it is no guarantee for today.

Re:More Cores, More Power

[The Snowman]

The 6 core system is slower in non-parallel tasks because the OS has per-core overhead. So all single-threaded tasks get slower as the number of cores rises.

Imagine a task running on an otherwise idle core. It is running as fast as possible, with only OS overhead getting in the way of using 100% of that cpu. Now add more OS overhead to that cpu for core management. There's also cpu (hardware)-level overhead to consider, and the possibility that caches aren't ramped to the same level, so now more cores may be sharing a same-sized cache ... etc.

Imagine a single-threaded game with a high CPU demand that consumes all time on a single CPU, while background processes such as the OS, drivers, pr0n downloads, etc. run on the second core.

Yes, each core has overhead, but in general, more cores does increase the system's potential performance even if it maintains or decreases an application's performance.

Honestly, this is not always a big deal. I have a quad core, but often have to wait for the hard drive to do anything.

Re:More Cores, More Power

[recoiledsnake]

Just add 4 more cores to handle the core overhead.

Re:More Cores, More Power

[L4t3r4lu5]

The Jeremy Clarkson school of Computer Science.

POWERRRRRRRR!

Re:More Cores, More Power

[vlm]

while background processes such as the OS, drivers, pr0n downloads, etc. run on the second core.

More likely, given todays throw away mentality of, "if its infected, just throw it out and buy a new PC", using a 6-core processor means you get 5 "free" worm/virus infestations before you "have to" buy a new computer. Unless you run mac or linux of course.

It's in their best interests

[mikael_j]

The average consumer just thinks "bigger is better" and by creating a mess of hard to understand sequence numbers they can make it harder for the semi-knowledgable customer to pick the right CPU. The same can be seen with graphics cards and many other products (if there is some kind of system behind your sequence numbers you do have to remember to change the system every now and then to further confuse everyone).

Re:It's in their best interests

[pelrun]

Yep, much like the mobile phone industry - make the whole mess so utterly confusing that instead of picking an appropriate product that suits your budget, you're tricked into buying at an inflated price.

Re:It's in their best interests

[pclminion]

What benefit is there in confusing your customers as to which product they should purchase? When I, as a consumer, feel overwhelmed or confused about a product choice, I usually respond by simply purchasing nothing at all. And I'm sure I'm not alone in that.

Re:It's in their best interests

[spazdor]

There are others, who respond to the same stress by spending indiscriminately. And their reaction might, on the economic whole, outweigh yours.

Re:It's in their best interests

[Pharmboy]

The problem now is that you have to do a tremendous amount of research before you buy now. It used to be much simpler: Pentium 60, 66, 75 or 100, pick one. Later it was still simple with Celeron or P2/P3/P4, as you are picking bigger cache and faster bus speed. Now to get the highest return on partially defective silicon, they offer too many models, many that overlap each other, and can be very confusing, with some dual core models that outperform quad core, etc. A year ago I finally settled on a Q9550 but it took reading 50 articles to figure out that it was, at the time, the best bang for the upper middle buck. So yes, the average consumer will get boned.

Re:It's in their best interests

[ultranova]

Picking the right CPU is quite easy, it's the motherboard that's the problem, especially with the current fad of putting on the board as few PCI slots as possible.

To be fair, most boards nowadays have both networking and sound integrated, so it's not like the average consumer needs that many (or any, to put it bluntly) PCI slots. Add a graphic card, and that's it.

No.

[bertoelcon]

The headline asked a question, I answered it.

Ideally the best metric would be

[snooo53]

Some combination that measures both how many operations per second, and how much power it's going to take to do said operations (i.e. Watts/computing unit). I don't know if even FLOPS is sufficient anymore to describe current computing tasks. Heck, I'd be happy with any sort of standardization.

Re:Not at all

[Anonymous Coward]

One core is sufficient for 99% of office workstations that only run a browser and MS Office applications.

Absolutely not. There are so many crappy applications that will max out a single processor doing stupid things (like rendering javascript on a webpage), that a 2nd core is very very useful.

Since most software still isn't multithreaded, a crappy application will only max out one core, allowing you to still get work done.

Re:Not at all

[mikael_j]

Multicore plus enough RAM is generally a lot better performance-wise than singlecore plus low amount of RAM and an SSD.

If you're having performance issues in everyday office use that go away when switching from a regular hard drive to an SSD you could just try accepting that these days you need 1+ gigs of RAM instead of trying to implement a bunch of workarounds that don't address the actual problem (that your computer keeps swapping out stuff to the hard drive because you're running out of RAM).

Unfortunately it's pretty common to see regular office desktops with fast multicore CPUs and ridiculously low amounts of RAM (I've seen C2D 2+ GHz CPUs coupled with 512 megs of RAM, it ran slower than a low-end P4 with 2 gigs of RAM).

My gaming system is...

[mindwhip]

... overdue for its 2 year processor and motherboard upgrade. It is overdue because when I started to look at what processor met my ideal performance/cost ratio it was impossible to figure out.

I don't have time spare to sit with a spreadsheet and a matrix of 30 different processors to work it out so I won't be upgrading now until something breaks. You lost a sale Intel, and I will have to pay more attention to system requirements of games for a while.

I would guess I am not the only one choosing not to buy because its so unclear...

Re:They make perfect sense to a ph.d. professor

[Anonymous Coward]

What about telling him the truth ?

The sequence number is assigned by the marketing department in order to confuse you. By making it harder for you to know what you're buying, they decrease your bargaining power which allow them to charge you more.

I have no idea what's good anymore

[Chewbacon]

I remember the clockspeed race and it was much simpler to decide what CPU you needed when looking at system requirements. Just a week ago I was looking at a game's requirements and had no idea if my CPU met them. If I were to upgrade, I wouldn't know which CPU would satisfy the requirements. I'm pretty handy with computers and I find picking a processor with today's marketing daunting, I can imagine being totally in the dark if I knew little about computers. Intel could do a better job indicating which CPU is better than the other and letting you know what you're buying.

They are just late to the party

[DeadboltX]

Nvidia and ATI have been giving their graphics cards arbitrary numbers for years.

Is a 330m better than a 220m? maybe.
What about a 9600 vs an 8800? who knows.

Intel didn't invent the random product model numbering scheme, they are just joining the ranks.

Lies, damn lies, and CPU speed numbers

[Locke2005]

If the chip can't run all the cores at full speed due to heat/power considerations and therefore either throttles back each core's speed or disables some cores under heavy load, than core counts are really just a deceptive pissing contest, aren't they?

Bring back the good old days of MIPS!

[jayveekay]

Misleading Indicator of Performance Statistic was the worthless number we had back then, and we liked it!

One guess why

[Locke2005]

Have you considered that the reason the processor numbers tell you nothing is that ALL the chips are fabbed with 6 cores and the ones that have one or two bad cores in testing have 2 cores disabled and are sold as quads?

I couldn't care less

[adenied]

When I decide a new computer (usually because the current one is out of warranty) I just buy whatever the newest Mac laptop is that seems to fit my use case. I might look at the specs a bit but frankly I couldn't tell you what processor is in the one I currently have.

I used to care a lot about this. When I was in high school. I have a lot more interesting things to care about and I think 99% of the public does too. I'm not trying to diss anyone here. If being a processor geek is your thing, more power to you. But I think people decide for whatever reason that at some point they need a new computer and just buy whatever fits their price bracket and feature needs.

If I was say, building a huge server farm, or spec'ing out computers for a big group of people I'd obviously do a lot more homework. But those are edge cases in the grand scheme of things.

Re:Here's a short summary of TFA.

[mindwhip]

Yeah but Intel need to label the chips with those 3 things for it to work rather than make up a random number and leave you scrambling around looking for useful info...

Re:They make perfect sense to a ph.d. professor

[aquila.solo]

THEY CAME FIRST for the marketing department, and I said "They're on the third floor, next to legal. Tell them I sent you."

Can you show examples?

[Sycraft-fu]

Everything I see shows that modern OSes not only don't have an overhead with more cores, it helps things. Reason is what OSes really have is a heavy context switching overhead. If a processor is doing something, and the OS needs it to do something else, it has to generate an interrupt, push everything on to the stack, switch to the kernel, switch to the net process, etc. It is a hefty overhead. However that all goes away if instead multiple things run at the same time on hardware. They don't switch contexts, they just keep running.

This is the reason why web/DB heavy servers like to have lots of cores, even if less powerful. Sun's new chips are designed with that in mind. Each core can handle 8 threads in hardware, meaning it acts like a 64-core CPU though only having 8 actual cores. Why? Context switching. The tasks it normally deals with are not high load, but they switch around a lot. The more than can run side-by-side from the OSes perspective, the less overhead and the more efficient use of processor resources.

In a desktop the tasks are more intense so it is less useful to have lots of threads/CPU (currently 2 is the highest in the Core i3/5/7 series) but more cores are still quite useful. It allows for more things to happen at the same time, from an OS perspective, and lowers overhead.

You notice too, using a multi-core, multi-threaded system. Things are damn responsive.