Pc for CAD

JonJ678 said:
i7 940 (22 +1) * 133 = 3059MHz
i7 870 (22 + 2) * 133 = 3192MHz

Turbo boost does run when all four cores are active. Indeed it provides precisely the boost I've ascribed to it when all four cores are active; it would have been irrelevant for me to say the 870 runs at 3.6ghz when only one core is active.

I've read the intel whitepaper on turbo boost and then spent a fair few hours looking into how it works. You've failed to grasp even the basics of its operation.
Click to expand...

You also fail to grasp the fact that the socket 1156 setup is faster out of the box,has more RAM and lower power consumption too.

JonJ678 said:
Further I'm capable of independent thought beyond this benchmark. For example, that an architect is going to be bright enough to achieve the trivial 4ghz (with all four cores active) overclock on either chip, making stock speed comparisons redundant.
Click to expand...

I know plenty of people who do high speed biological imaging. They have got better things to do then mucking around overclocking computers as it not only takes up their time stability testing but also means more potential issues in the long run. None of these systems or the ones used for processing the large data sets are overclocked.

http://blogs.technet.com/photos/gray_knowlton/images/2998979/original.aspx no hotlinking

Anyway the OP can make their mind up about what socket they want to go with.
 
Last edited:
The one you specified has more ram because you decided 4gb wasn't enough. If you want lots of ram, again it's X58 with 12gb for cheap rather than P55 with 8gb cheaply. 4gb sticks aren't common yet.

I know it's faster out of the box. I've even said this. I don't give a **** about power consumption, so freely admit that I ignore it.

Your name dropping is entirely irrelevant. The OP is a student. I'm yet to meet a student who doesn't have any free time, certainly not one in the engineering fields who needs his computer to be quick yet isn't willing to spend a day or so to make it much faster.

If I were currently working as an design engineer I'd be using a dual socket xeon system, at stock speeds. As I'm still a student, I'm using an overclocked 920 for cad work. It is ideal for this task.

You're quite right, you've irritated me. You're handing out bad advice.
 
JonJ678 said:
You're quite right, you've irritated me. You're handing out bad advice.
Click to expand...

You are also handing out bad advice too since you make the instant assumption that everyone knows to overclock their PC or wants to.

I have worked with software programmers,engineers and bio scientists and very few of them overclock.

Only the ones who are hardware enthusiasts like me can be bothered by such things.
 
Last edited:
Given the name of this site it's a reasonable assumption. Given stupid people fail to get onto a course in architecture and how simple overclocking an X58 920 is it's a reasonable assumption that the OP will be capable of it.

I feel your credibility was shredded when it turned out you had no idea what turbo is. You're not regaining this by telling me you know lots of intelligent people. The few professional engineers who I've spoken to have invariably been using the high end mac pro's for work. Students are unlikely to be able to afford this.

I wouldn't bother overclocking if the performance increase wasn't so high. It is, so I do. If I could afford a multiple xeon system instead, then yes I'd use that, and it wouldn't be overclocked for the good reason that server/workstation boards generally don't offer this.
 
JonJ678 said:
Given the name of this site it's a reasonable assumption. Given stupid people fail to get onto a course in architecture and how simple overclocking an X58 920 is it's a reasonable assumption that the OP will be capable of it.

I feel your credibility was shredded when it turned out you had no idea what turbo is. You're not regaining this by telling me you know lots of intelligent people. The few professional engineers who I've spoken to have invariably been using the high end mac pro's for work. Students are unlikely to be able to afford this.

I wouldn't bother overclocking if the performance increase wasn't so high. It is, so I do. If I could afford a multiple xeon system instead, then yes I'd use that, and it wouldn't be overclocked for the good reason that server/workstation boards generally don't offer this.
Click to expand...

Well said.
 
JonJ678 said:
Given stupid people fail to get onto a course in architecture and how simple overclocking an X58 920 is it's a reasonable assumption that the OP will be capable of it.
Click to expand...

You would be surprised by the type of people that do Architecture. I'm an Architecture student and there are loads of students that are incapable of sorting out any PC problems when it arises. 90% found it difficult to even understand how to get onto the Uni's wireless network via proxy even when explained to them.
 
Hey, I didn't want any arguments over things
not that I understand much of it.
but yea, I'm pretty sure I'll try overclocking, if only slightly more

Just wondering about the new amd chips I've been hearing about and if they're likely to work better than an i7 for CAD tasks.

also could someone explain the memory things to me and if this would affect what I want to do significantly
for example; Geil Black Dragon 6GB (3x2GB) DDR3 PC3-12800C8 (1600MHz) Tri-Channel (GB36GB1600C8TC) [GB36GB1600C8TC]

I've read up that C7 is faster than C8 etc.. and the higher MHz is better but am unsure!

cheers for everyones replys
 
It's easy to be drawn into ultimately petty arguments about hardware, don't worry about it. I was probably too hard on Cat.

Overclockers are a bit spoilt these days, even the lowliest i7 920 does 3.8ghz as easily as it does 3.0. I imagine you'll get carried away and end up running faster than you planned.

Ram timings / frequency doesn't matter very much for X58/i7. Faster is always nice, but it's nothing to loose sleep over. There's a detailed, and hopefully approachable article here. Timings/frequency might matter for amd or P55. I believe it's triple channel courtesy of Intel's high end which provides so much bandwidth that ram speed doesn't matter, the lesser chips are dual channel.

AMDs six core chips might be good, depends how multithreaded your programs are. No benchmarks out yet for amds hex core against the i7 so it's hard to know.
 
You must log in or register to reply here.
Back
Top Bottom