1 |
On Wed, Jan 23, 2013 at 12:16 PM, Volker Armin Hemmann |
2 |
<volkerarmin@××××××××××.com> wrote: |
3 |
> Am 23.01.2013 16:35, schrieb Nikos Chantziaras: |
4 |
>> On 23/01/13 17:09, Nilesh Govindrajan wrote: |
5 |
>>> On Wednesday 23 January 2013 07:52:03 PM IST, Nikos Chantziaras wrote: |
6 |
>>>> [...] |
7 |
>>>> In my experience, most of the time you can overclock. The issue is |
8 |
>>>> with the user not knowing exactly how to do it. You need to |
9 |
>>>> understand a few things and how they affect each other. It's not just |
10 |
>>>> a knob you can turn. |
11 |
>>> |
12 |
>>> That pretty much applies to me. I don't know much about hardware stuff. |
13 |
>>> Regarding your 1 Ghz overclock, you probably have good components in |
14 |
>>> terms of RAM & SMPS. |
15 |
>>> When I bought this rig in 2008, I knew nothing about good components, |
16 |
>>> blindly trusted local vendor... also internet shopping wasn't advanced |
17 |
>>> here. |
18 |
>>> So pretty much substandard components. |
19 |
>> |
20 |
>> The part that's really important is the mainboard. RAM doesn't |
21 |
>> matter. In my case, I had pretty basic 800MHz DDR2 RAM. Raising the |
22 |
>> FSB would bring it above that, so I changed the DRAM ratio to 1:1, and |
23 |
>> the RAM then ran at only 600Mhz. |
24 |
>> |
25 |
>> That was the starting point to rule out RAM problems. After that, I |
26 |
>> raised FSB but kept the VCore constant until I hit the first |
27 |
>> instabilities. When that happened, I raised VCore a bit. Rinse and |
28 |
>> repeat, until the VCore was still below the maximum recommendation by |
29 |
>> Intel. That happened at 3.4GHz (378MHz FSB * 9 CPU multiplier = |
30 |
>> 3402MHz CPU clock.) The E6600 CPU I got was an average sample. |
31 |
>> Others were running it at 3.6GHz (or even higher with water cooling.) |
32 |
>> |
33 |
>> This was a process that took about 3 days to complete (needs a lot of |
34 |
>> stability testing.) The good thing about those older CPUs was that |
35 |
>> the performance boost I got by OCing wasn't just scaling linearly with |
36 |
>> the CPU frequency. It was scaling *better* than that, because raising |
37 |
>> the FSB also made the mainboard itself perform better and with lower |
38 |
>> latencies. |
39 |
>> |
40 |
> and here we are - the point where the suspension of disbelief ends. |
41 |
> |
42 |
> All you may have gained you threw away with the slower ram - and you are |
43 |
> trying to tell us that your rig was faster? |
44 |
> |
45 |
> You do know that with today's CPUs the CPU is not the bottleneck - the |
46 |
> slow as molasses, no speed bump for 10 years ram is. |
47 |
> |
48 |
> (just look at the internal clock rate of dram chips - and you realize |
49 |
> that ddr1-3 are pretty much the same crap). |
50 |
> |
51 |
|
52 |
Volker, in applications speficially tuned to keep their hot data small |
53 |
enough to stay in CPU cache (so, anything with a "frames per second" |
54 |
measurement), overclocking the CPU would still see performance |
55 |
improvements. Cache misses are always painful. |
56 |
|
57 |
|
58 |
-- |
59 |
:wq |