Neoseeker.com Forum Thread: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations - page 1

reprinted from http://www.neoseeker.com/forums/
original thread: http://www.neoseeker.com/forums/17/t1796786-i7-3930k-vs-xeon-e5-2630-for-long-duration-financial-calculations/


Author:   dave777
Date:   Nov 24, 12 at 5:53am (PST)
Subject:   i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
Hello,

I wonder if you can help me to build the perfect PC setup for the following task:

I am going to perform daily financial market analysis on multiple instruments using Excel (large workbooks around 300 and 500 mbs), Matlab and Automation software. The price will be loaded into Excel, the results would be fed into Matlab to create charts (this sequence will be repeated many times). I plan to sell the resultant analysis at my website for a set monthly subscription – so this can be considered a production workstation. The total process would take 5-7 hours daily. I need this process to run as error-free as possible – absolutely predictably on autopilot. So I am not planning to over-clock the CPU.

I am deciding between a workstation built on Intel Core i7-3930K or the one built on single Intel Xeon E5-2630. These CPUs are roughly the same in price (same number of cores, different speed though) with the I7 being much faster one. But I am more concerned with reliability and stability of this setup. Do you think the ECC memory can help eliminate system crashes when the analysis job is running? I need to be able to connect this PC remotely to initiate the analysis jobs as well. I am also thinking of Intel Core i7-3930K which is very fast but not sure about its stability for long-duration number crunching sessions (it can overheat if run at full speed for many hours?).

Please let me know what you think,

Dave




Author:   dave777
Date:   Nov 24, 12 at 1:05pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I am just concerned if i7 will be able to run the continuous number crunching session lasting MONTHS ? I might not power-off the computer for months! you think XEON can handle such workload? and I7?



Author:   Crusad3r
Date:   Nov 24, 12 at 4:39pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I don't think either of them would be able to do it for months on end without a reboot.

Your work sounds pretty CPU intensive though and I'm not sure if the Xeon will be able to handle it in a timely fashion.

My personal opinion, go with the i7.



Author:   harbin
Date:   Nov 24, 12 at 6:06pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
Xeons are designed to go in servers, which are expected to run endlessly, which is why EEC RAM is valued in those cases, since it corrects memory errors instead of just chucking an error in your face because it messed something up. Since you are needing it to run number crunching programs, then EEC would be very advantageous. Xeons support EEC RAM, i7s do not... in stuff like this, it is a must since it restricts data loss. EEC RAM costs alot more than normal RAM though.

Overheating isn't usually an issue on most PCs since they've adequate enough cooling but if you are worried then after market cooling would be better, either a better air cooler or an liquid cooler. Hard to find a proper benchmark comparison since typing i7-3930k vs e5-2630 comes up with loads of duplicate threads around the internet (fair enough, you'll want to gather as much info), but the one benchmark website I often check (although I doubt it's reliability sometimes) puts the i7-3930k as a better performing CPU. Better server CPUs exist but for much higher prices, the best I can find for price (based on that website) is the AMD Opteron 6272 which has 16 cores but I'm not entirely sure on how many cores those applications are best optimised for. AMD performance has suffered in the past due to core-on-core performance.

Also, if you are interesting in a build, better to recommend a budget



Author:   Benedict
Date:   Nov 25, 12 at 1:38am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
You'd probably be better off renting a server, actually. It depends on what you need, because costs get high on both renting and buying.



Author:   dave777
Date:   Nov 25, 12 at 3:01am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
thank you for your interesting reply.....well I have 50 currency pairs to analyse daily - 4 time frames each - each time frame shoudl be analyzed using a few different large excel files (5 or so) - around 300 mbs on average....meaning on average the 300 mbs excel file would need to be recalculated 5*4*50=1000 times including getting the data into it and outputting to Matlab that would need to be recalculated 1000 times as well per day....

do you think this process can be speeded up if I break each of the 4 time frames into different virtual machines and place the analysis pipeline into them?.....you think a 4 core E3 1275 V2 might be enough for it???



Author:   Crusad3r
Date:   Nov 25, 12 at 4:09am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I don't think virtual machines are the answer to your problem as virtual machines themselves require a certain amount of CPU power to run.

Your needs require a lot of CPU power and reliability. You can get the power from the i7, but it will require frequent rebooting. The Xeon can give you the reliability, but you won't be able to get the power unless you're willing to spend thousands of dollars for a higher end model.

You might have to look into getting a dual CPU motherboard.

I wouldn't recommend getting the AMD at all since they're notorious for performance issues.



Author:   dave777
Date:   Nov 25, 12 at 4:39am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I am also thinking - for the price of XEON 2630 -634$ (http://www.newegg.com/Product/Product.aspx?Item=N82E16819117268) I could get two lower end processors from the 2600 series like Intel Xeon E5-2609 ($299.99) (http://www.newegg.com/Product/Product.aspx?Item=N82E16819117270) on a two CPU motherboard? to maximize the number of cores?

Or I can go with single Xeon E3 1275 V2 altogether and lots of ECC ram - can you recommend a setup on this CPU?

thanks!



Author:   dave777
Date:   Nov 27, 12 at 1:00am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I am stuck with excel for now as taking it to an executable code would add a year to the project....do you think EXCEL 2010 on 4 cores will run faster than EXCEL 2003 on one core?



Author:   ashantiqua
Date:   Nov 27, 12 at 8:57am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
harbin91 has it right.

server hardware is purpose built for what youre looking for, that is, sustained, uninterrupted, reliable operation.

ECC memory is also a bit slower than non-ECC memory, but for sustained operation, its your guy.

as for multicore performance, heres some info concerning excel 2007 to present:

"Most workbooks show a significant improvement in calculation speed on a system with multiple cores. The degree of improvement depends on how many independent calculation trees the workbook contains. If you make a workbook that contains one continuous chain of formulas, it will not show any multithreaded calculation (MTC) performance gain, whereas a workbook that contains several independent chains of formulas will show gains close to the number of processors available."

"Results showed calculation times ranging from no improvement to better than theoretical (2x/4x) improvement on both the dual-core and quad-core systems. Typical (median) improvement for a system with an Intel dual-core Pentium 4 at 3.0 GHz with 1 GB of RAM compared to the same file calculating in Excel 2003 was 48 percent, or a 1.92x speedup. Typical (median) speedup for a system with an Intel quad-core Xeon at 3.0 GHz with 4 GB of RAM was 76 percent, or a 4.17x speedup. Similar speed improvements were observed on other processors and platforms. Improvements beyond theoretical speedup (because of multithreading) are attributed to other performance improvements starting in Excel 2007, such as improvements to the speed of function execution."

looks like if the type of calculation youre doing is supported with multicore processing, youll get great performance gains. its scalable, so the more cores you have, the better performance you should see.

some forms of calculating are not supported by multithreading / multicores however.

"Some Excel features do not use multithreaded calculation. For example:
Data table calculation (however, structured references to tables do use MTC).
User-defined functions (however, XLL functions can be multithread-enabled).
XLM functions.
INDIRECT, CELL functions that use either the format2 or address options.
GETPIVOTDATA and other functions referring to PivotTables or cubes.
Range.Calculate and Range.CalculateRowMajorOrder.
Cells in circular reference loops."



Author:   dave777
Date:   Nov 27, 12 at 9:42am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
quote ashantiqua
harbin91 has it right.

server hardware is purpose built for what youre looking for, that is, sustained, uninterrupted, reliable operation.

ECC memory is also a bit slower than non-ECC memory, but for sustained operation, its your guy.

as for multicore performance, heres some info concerning excel 2007 to present:

"Most workbooks show a significant improvement in calculation speed on a system with multiple cores. The degree of improvement depends on how many independent calculation trees the workbook contains. If you make a workbook that contains one continuous chain of formulas, it will not show any multithreaded calculation (MTC) performance gain, whereas a workbook that contains several independent chains of formulas will show gains close to the number of processors available."

"Results showed calculation times ranging from no improvement to better than theoretical (2x/4x) improvement on both the dual-core and quad-core systems. Typical (median) improvement for a system with an Intel dual-core Pentium 4 at 3.0 GHz with 1 GB of RAM compared to the same file calculating in Excel 2003 was 48 percent, or a 1.92x speedup. Typical (median) speedup for a system with an Intel quad-core Xeon at 3.0 GHz with 4 GB of RAM was 76 percent, or a 4.17x speedup. Similar speed improvements were observed on other processors and platforms. Improvements beyond theoretical speedup (because of multithreading) are attributed to other performance improvements starting in Excel 2007, such as improvements to the speed of function execution."

looks like if the type of calculation youre doing is supported with multicore processing, youll get great performance gains. its scalable, so the more cores you have, the better performance you should see.

some forms of calculating are not supported by multithreading / multicores however.

"Some Excel features do not use multithreaded calculation. For example:
Data table calculation (however, structured references to tables do use MTC).
User-defined functions (however, XLL functions can be multithread-enabled).
XLM functions.
INDIRECT, CELL functions that use either the format2 or address options.
GETPIVOTDATA and other functions referring to PivotTables or cubes.
Range.Calculate and Range.CalculateRowMajorOrder.
Cells in circular reference loops."

yes thanks for this helpful answer...I do have formula streams that are independent of one another but which converge at the final value - do you think this can be considered a multi-=threaded ready setup? thanks a lot!)




Author:   dave777
Date:   Nov 28, 12 at 2:33pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
I wonder if you have tried running your i73930k for long periods of time..like 10 hours at the maximum load....I wonder how it behaved? lets say for 8-10 hours in a row....at 95% (no over-clocking) - did you see any hints of instability in its performance?
Thanks!



Author:   Crusad3r
Date:   Nov 28, 12 at 5:43pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
quote dave777
I wonder if you have tried running your i73930k for long periods of time..like 10 hours at the maximum load....I wonder how it behaved? lets say for 8-10 hours in a row....at 95% (no over-clocking) - did you see any hints of instability in its performance?
Thanks!
I have an i7 2600k, the earlier generation, and I've ran it at close to max load for around 5 hours with no problems (it was a high end game). I've also had this computer run for up to a week before I started getting errors. This is with the CPU underclocked though, since it is really hot where I am so I'd rather it produce less heat.



Author:   Luis_GT
Date:   Nov 28, 12 at 8:02pm (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
quote dave777
I wonder if you have tried running your i73930k for long periods of time..like 10 hours at the maximum load....I wonder how it behaved? lets say for 8-10 hours in a row....at 95% (no over-clocking) - did you see any hints of instability in its performance?
Thanks!
CPU's are actually a lot more capable than you think, I had a i7 950 setup (Gen 1) that I ran 24/7 under full load, while overclocked to 4ghz, using folding @ home, only time it wasn't at full load was when the power went out, this was used for 2 years, until I upgraded this week. Same story with GPU's as I have had my 5830 running at constant full load 24/7 while mining bitcoins (ie. lots and lots of calculations) for over a year.


as long as the CPU has proper cooling... it can almost run forever at full load.



Author:   dave777
Date:   Nov 29, 12 at 10:19am (PST)
Subject:   re: i7-3930K vs Xeon E5-2630 For Long-duration Financial Calculations
-------------------------------------------
thank you for your generous attention!)

I need to introduce one more element into the equation...sorry to have left it out before....I need to spend quite some time outside of the Europe and USA - in eth Western Europe - where the server grade parts take a week to arrive - a week is intolerable time to wait for any part in case the setup fails....I will be providing the analysis each day and it needs to be updated on a daily basis...new signals can orccur any time...

for a server-based example, I have been recommended the followign setup

CPU: Xeon E3 1275 V2
MOBO ASUS P8C WS
Kingstong RAM 16GB DDR3 1600 ECC Unbuff. (4x4gb)

the cpu and motehrboard are more or less easily available here but the RAM 16GB DDR3 1600 ECC Unbuff which is only accepted by this setup needs to be ordered and will take from 7 to 10 days to arrive.....I have not foudn otehr unuffered ECC on teh market now...

I mean if any part of this seup fails - Especially memory - I will need to replace it...but only within one weeks time....


if i go with the i7 3930k based setup - I shoudl be able to get any kind of the memory in no time at all from teh local PC shops...The XEON+ECC+SERVER mobo looks very reliable but if anythign beraks down I stand the chance of replacing the broken part much faster with the 3930k based setup....considering the power of 3930k I mitgh be able to crunch all data in a matter of 5 o6 hours so no need for 24/7 operation for now....

I will be doign maintenance myself at first - do you think a server-based setup mitghbe harder to maintain than a desctop based setup?

thanks!)


Copyright Neo Era Media, Inc. 1999-2014.
All Rights Reserved.