![]() |
|
|
#243 |
|
Dec 2014
3×5×17 Posts |
Thanks for the input.
You all raised valid points, but most I can not measure the impact of. How much does RAM speed or MB chipset affect the iteration time? I am sure they do affect it, but I don't have a way to measure the impact. The C6100 basically has 4 motherboards in a 2U case. Each MB has 24GB RAM which is more than the others but I am not sure how to reflect that in a score. I have 2 Xeon 1U servers (by Dell and HP). They seem to scale up fine when using all cores. My electric Utility charges $0.06 per kWh. The C6100 draws 500W under full load which is 12 kW / day or $0.72 / day or $22 / month. Hmm. I am a little worried about the heat... |
|
|
|
|
|
#244 | |
|
"Kieren"
Jul 2011
In My Own Galaxy!
2×3×1,693 Posts |
Crosspost:
Consider the difference grouping makes: [LIST=1][*]score=(benchmark divided by cores) times cost[*]score=benchmark divided by (cores times cost)[/LIST] One puts cost in the numerator. The other puts cost in the denominator. Let's say 'benchmark' = 4 ms 'cores' = 4 'cost' = 400 [LIST=1][*]x=(4/4)*400=400[*][STRIKE]x=4/(4*400)=0.75 In #2 'cost' vastly decreases 'score', which only gets worse as cost increases. So #1 seems more sensible, especially if the 'benchmark' is iteration time, which is the only way the whole thing really makes sense. However, Quote:
So #1 is the intended form, but what is 'benchmark'? [/STRIKE]Duh. 'Cost' is electricity? What is benchmark? I should stop
Last fiddled with by kladner on 2015-11-20 at 06:00 |
|
|
|
|
|
|
#245 |
|
Jun 2003
116738 Posts |
E5507 is nothing like E5462. I guess an i7 920 (benchmark = 77.84ms) is a better comparison, but with probably worse performance. We should look at something like 85-90ms.
Also, i5-6400 score = 17.91/4*538= 2409 (not 2847). And I don't see any benchmark numbers for E3-1240. Where did you get the 23ms number? |
|
|
|
|
|
#246 |
|
Jun 2003
5,051 Posts |
|
|
|
|
|
|
#247 | |
|
"Kieren"
Jul 2011
In My Own Galaxy!
1015810 Posts |
Quote:
Even if one needs to replace the memory, it is awfully cheap, even if it is four generations back, or so. Second Edit: Consider how much you will spend for power for a 2xxx CPU. Long-term costs can easily outweigh purchase price. FINALLY (or finale): I completely overlooked detailed explanations from the OP.
Last fiddled with by kladner on 2015-11-20 at 08:11 |
|
|
|
|
|
|
#248 |
|
Jan 2014
2×73 Posts |
|
|
|
|
|
|
#249 | |
|
Nov 2008
3·167 Posts |
Quote:
|
|
|
|
|
|
|
#250 |
|
Jun 2003
5,051 Posts |
|
|
|
|
|
|
#251 |
|
(loop (#_fork))
Feb 2006
Cambridge, England
144238 Posts |
Is this just companies being unused to international shipping and picking a single ludicrously expensive option? Or does it really cost eight hundred dollars to ship a 250lb three-foot-by-three-foot-by-one-foot cardboard box from Michigan to the UK?
|
|
|
|
|
|
#252 |
|
Undefined
"The unspeakable one"
Jun 2006
My evil lair
22×1,549 Posts |
It depends upon how long you want to wait for the package. Plus that is heavy so going by air will be pricey. And general handling will require more than just one guy in a van.
|
|
|
|
|
|
#253 |
|
Bamboozled!
"πΊππ·π·π"
May 2003
Down not across
29×3×7 Posts |
Wouldn't be surprised. The best price I could get for shipping a Victorian family bible to Florida was > Β£80, or ~$125. That weighed much less than 250lb. In the end I hand-delivered it to the recipient when she visited the UK. Even if she had to pay excess baggage charges it would still have been much cheaper.
|
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Hardware Benchmark Jest Thread for 100M exponents | joblack | Hardware | 284 | 2020-12-29 03:54 |
| Garbage hardware thread | PageFault | Hardware | 21 | 2004-07-31 20:55 |
| Old Hardware Thread | E_tron | Hardware | 0 | 2004-06-18 03:32 |
| Deutscher Thread (german thread) | TauCeti | NFSNET Discussion | 0 | 2003-12-11 22:12 |
| Gratuitous hardware-related banana thread | GP2 | Hardware | 7 | 2003-11-24 06:13 |