1 |
On Wed, Jul 25, 2012 at 4:32 PM, Florian Philipp <lists@×××××××××××.net> wrote: |
2 |
> Am 25.07.2012 22:14, schrieb Volker Armin Hemmann: |
3 |
>> Am Mittwoch, 25. Juli 2012, 16:05:29 schrieb Philip Webb: |
4 |
>>> I've listed what's available at the local store, |
5 |
>>> which I trust to stock reliable items, tho' I wouldn't ask their advice. |
6 |
>>> |
7 |
>>> All the AMD's are 32 nm , while the Intel recommended by one commenter |
8 |
>>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm -- |
9 |
>>> is 22 nm : it costs CAD 230 & they have 3 in stock, |
10 |
>>> which suggests demand, but not the most popular ( 9 in stock). |
11 |
>>> |
12 |
>>> Isn't 22 nm going to be faster than 32 nm ? |
13 |
>> |
14 |
>> no |
15 |
>> |
16 |
> |
17 |
> Lower transistor size gives you two advantages: Lower current (-> |
18 |
> potentially lower power consumption and heat) and more transistors to do |
19 |
> something. The practical effects depend on what the chip maker does with |
20 |
> this. |
21 |
|
22 |
I second this; the feature size limit of the process isn't really |
23 |
something a consumer should care about at _all_. Its only real impact |
24 |
is on what architectural options are open to the manufacturer, which |
25 |
in turn drives how much they can get out of a performance and feature |
26 |
balance. |
27 |
|
28 |
What you really care about is what the manufacturer builds, not the |
29 |
tools and materials they had available to them. |
30 |
|
31 |
> |
32 |
>>> |
33 |
>>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W) |
34 |
>>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in stock). |
35 |
>>> |
36 |
>>> How do you compare cores vs nm ? |
37 |
>> |
38 |
>> who cares? |
39 |
>> |
40 |
> |
41 |
> You cannot really compare this. If you can use more cores, e.g. because |
42 |
> you have an embarrassingly parallel application, by all means, get it. |
43 |
> Otherwise you should probably care more about single core performance. |
44 |
|
45 |
I'll note that emerge -e @world with parallel emerge and parallel make |
46 |
qualifies. |
47 |
|
48 |
So does running a browser like Chromium which gives each tab its own process. |
49 |
|
50 |
> |
51 |
>>> How far is cache size important ( 6 vs 8 MB )? |
52 |
>> |
53 |
>> depends on the architecture. |
54 |
>> |
55 |
> |
56 |
> In short, for all three questions: Look at benchmarks and look at the |
57 |
> TDP ratings if that is important to you. |
58 |
|
59 |
Good points. |
60 |
|
61 |
> |
62 |
> nm numbers don't tell you anything that can be directly translated into |
63 |
> performance or other qualities. They only allow educated guesses. If you |
64 |
> really want to delve so deep into chip design, you could as well look at |
65 |
> pipeline depths, cache associativity and such alike (not that you should). |
66 |
|
67 |
Not that that isn't fun. ^^ |
68 |
|
69 |
|
70 |
|
71 |
-- |
72 |
:wq |