Processors: more cores vs. higher clock speeds? (desktop, RAM, HDD)
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
I was wondering if I could please ask a question, on the issues of computer processor hardware: specifically, on the issue of having more cores vs. higher stock-clock speeds?
For example, I currently have 2 PC desktop workstation systems set up at home (please see additional specs below), one that is older and has a higher clock speed and less cores, and one that has a lower clock speed, but more cores. Just wondering if it is possible for the novice hardware user how to determine how fast each of these is in relation to each other, versus just looking at the stock-clock speed alone?
[System 1]
x2 Xeon (Harpertown series, Quad Core, LGA 771) X5492 processors, running at 3.4 GHz each
32 GB FB-DIMM RAM installed (max RAM capacity of 128 GB)
[System 2]
x2 Xeon (Sandy Bridge-EP series, Six Core, LGA 2011) E5-2640 processors, running at 2.5 GHz each
16 GB DDR3(?) RAM installed (max RAM capacity of 512 GB)
The irony is, even though System 1 is almost 1 GHz faster in terms of stock-clock speed, and has twice as much RAM installed, System 2 still outperforms it, in terms of overall speed -- not really sure I understand how exactly though, tbh, other than the fact that the processors in System 2 are newer in terms of when they were released?
I was wondering if I could please ask a question, on the issues of computer processor hardware: specifically, on the issue of having more cores vs. higher stock-clock speeds?
For example, I currently have 2 PC desktop workstation systems set up at home (please see additional specs below), one that is older and has a higher clock speed and less cores, and one that has a lower clock speed, but more cores. Just wondering if it is possible for the novice hardware user how to determine how fast each of these is in relation to each other, versus just looking at the stock-clock speed alone?
[System 1]
x2 Xeon (Harpertown series, Quad Core, LGA 771) X5492 processors, running at 3.4 GHz each
32 GB FB-DIMM RAM installed (max RAM capacity of 128 GB)
[System 2]
x2 Xeon (Sandy Bridge-EP series, Six Core, LGA 2011) E5-2640 processors, running at 2.5 GHz each
16 GB DDR3(?) RAM installed (max RAM capacity of 512 GB)
The irony is, even though System 1 is almost 1 GHz faster in terms of stock-clock speed, and has twice as much RAM installed, System 2 still outperforms it, in terms of overall speed -- not really sure I understand how exactly though, tbh, other than the fact that the processors in System 2 are newer in terms of when they were released?
You can really only compare those specs within a family of processor.
The newer processor has a different architecture that allows it to do more with less cycles.
It has more cache on the chip (15 Mb vs. 12 Mb), it has two more cores (so it can theoretically process two more instructions per cycle), and can also turbo to 3.0 GHz in some circumstances.
You also don't say what board / FSB the two systems are on, HDD architecture, or the spec of the RAM. These things change with each new generation as well, allowing each part to do it's job faster, and allowing the separate parts to communicate with each other faster.
You also don't say what board / FSB the two systems are on, HDD architecture, or the spec of the RAM. These things change with each new generation as well, allowing each part to do it's job faster, and allowing the separate parts to communicate with each other faster.
Please feel free to correct me, if I have accidentally misinerpreted what it is you are asking for, in the below items:
[System 1]
Board: LGA 771 server-compatible board
2x 3.4 GHz Harpertown X5492 Xeon processors: 1600 MHz FSB, 150 W TDP
HDD architecture: 250 GB (not sure about additional specs)
RAM specs: 32 GB DDR2-generation, Fully-Buffered (FB-DIMM), RAM runs slightly slower than conventional desktop PCs b/c of server board
Both RAM and the HDD have an interface spec. The RAM will have a speed (667, 800, 1333, 1600, etc), and the HDD will spec a particular interface (IDE, SCSI, Sata II, Sata III, etc). The board will support some variant of these speeds as well.
I actually missed the fact that the SB CPU is a HT CPU, so it's really a 12 core chip, not a 6 core chip. The CPU can potentially process 3 times as many items per CPU cycle, so even though it's not running as many cycles per second, it can do so many more things per cycle that it doesn't matter. On top of that, it can do it with about 1/3 less power than the older CPU (95 watts vs. 150).
Also, a computer is more than just a CPU. It's really the sum of it's parts, and which ones matter will depend upon the primary job of the machine.
Last edited by Skunk Workz; 11-05-2013 at 09:52 AM..
Also, a computer is more than just a CPU. It's really the sum of it's parts, and which ones matter will depend upon the primary job of the machine.
That sums up my opinion. It all boils down to what OS and application will be installed and used on the machine.
Who cares if the machine has a 50-core CPU if the applications will only be utilizing 1?
You know what I mean?
It all boils down to what OS and application will be installed and used on the machine.
The current PC is intended to be as a test/evaluation environment server system, to be able to play around with the following software, all installed together on the same box:
Windows Server
SharePoint Server
(possibly) SQL Server
(possibly) Project Server
The current PC is intended to be as a test/evaluation environment server system, to be able to play around with the following software, all installed together on the same box:
Windows Server
SharePoint Server
(possibly) SQL Server
(possibly) Project Server
Sounds like a job for a VM, that's a very expensive test rig, and putting it back back into a pristine state after testing any of that is beyond painful.
Again, I agree with Skunk. Just two of the ones on the list alone would be an annoying challenge at the very least so better off creating a VM environment for this. Honestly, most everything we do, from app updates or script testing to creating new ws images, we use VM.
Sounds like a job for a VM, that's a very expensive test rig, and putting it back back into a pristine state after testing any of that is beyond painful.
One other reason -- aside from the server programs that I had planned to install on the second system -- that I chose this particular configuration was because of expandability, scalability, and relative future-proofing. For example, I bought the first system I listed almost 6 years ago, and it still today either meets or exceeds modern and contemporary hardware specs. You can't really do that (i.e., future-proofing against inevitable hardware inflation), if you buy a consumer-driven, mainstream computer system, because they are locked-down, in terms of their maximum hardware expandability. For example: if I had bought my first system with mainstream consumer hardware, at that time 6 years ago, it would have very likely to have been locked down to max RAM scalability of 4-8GB. I like to buy hardware that lasts for a 7+ year lifecycle, not the typical 2-3 year lifecycle that mainstream hardware faces. With my latest recent hardware upgrades to the first system (upgraded to 32GB RAM and 2x3.4GHz proceessors), I can probably still get at least a few more years out of it, maybe perhaps even up to a lifecycle of 10 years total, since the RAM maxxes out at 128GB, if I decide to do further memory upgrades later on.
The second system is intended as the successor of the first, also intended for a long life-cycle. I figure that if/when I need to later on in the future, I can simply upgrade the processor specs. The RAM maxxes out at 512GB. Subsequently, I am planning on a 7+ year lifecycle for this one, as well...we'll see if it hopefully works out that way
So it is not just the server software intended for testing itself -- the hardware specs also matter a great deal to me as well, which is probably why in my case, a VM in itself would still not be enough...again since I prefer to have hardware specs that are much more scalable than conventional consumer desktop systems.
ETA: I take backup images of the 2 systems I had mentioned almost religiously, so that I can always rollback to a stable configuration, if I need to. Right now Windows Server is installed, configured, and working fine (I'm glad I did several backup images, b/c I have already had to rollback to one once already); I just have to install and configure the other software items that I had mentioned, atm...
ETA: Agreed, the test rig is very expensive, but it is the only way I am currently able to train myself in a hands-on way for the referenced software packages such as Windows Server, SharePoint Server, etc., since my workplace does not really have any kinds of hands-on training, for these programs...and although I work with SharePoint Server every day with elevated rights, my workplace locks down the rights so that I do not have access to components such as Central Admin, etc., that would allow me to train myself, on-the-job. So training at home on my own time, on my own home server test rig is realy my only viable option...
Last edited by Phoenix2017; 11-07-2013 at 08:39 AM..
Again, I agree with Skunk. Just two of the ones on the list alone would be an annoying challenge at the very least so better off creating a VM environment for this. Honestly, most everything we do, from app updates or script testing to creating new ws images, we use VM.
Please understand, I am definitely not trying to ask a question that sounds dumb or uninformed in any way here, but is there a particular reason why you would want to use a VM envionment, if you could already have the actual physical hardware itself instead?
Also j/c, do you mind if I ask what VM software you are currently using? I have not really had a chance to experiment with VM COTS software yet so far, so that may be worth looking into more, for me...
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
