New i7 Rig, need advice

[nightkhaos]

Chaps
Could use some advice fairly soon as im gonna need to order the components by tomorrow afternoon (UK Time)

See my post above, I've been pretty much saying "go for it". But someone might disagree, so give it a few hours. Sorry, this is just the forums process. Someone posts something, someone else retorts, exventually they all agree, and then you get the best advice possible.

To be helpful, sometimes it might be worth posting supplimental information. Which you have been, I see you want it for Video Editing and modest high variable (SupCom) gaming? So CPU over GPU that translates into tech terms at the moment cause GPU accerlated encoding/transcoding leaves much to be desired.

[ta2]

6GiBs is not a bit of waste. It just means more of your data can be wired instead of sitting in the swap file, preventing you from switching applications quickly. And when those applications are games, well, you notice. Haven't you noticed that with a modest amount of RAM it can take 5 minutes to alt tab out of a game?

If you look at task manager/resource monitor you'll notice that a lot of swap is being used even if you have a couple of gigs of memory still available. This is because when developers are using the windows API they set an affinity for certain IO operations to use memory or swap. This is because if the memory was kept full at all times, you'd have to move the data to storage/swap before you could use it for a new process.

[nightkhaos]

If you look at task manager/resource monitor you'll notice that a lot of swap is being used even if you have a couple of gigs of memory still available. This is because when developers are using the windows API they set an affinity for certain IO operations to use memory or swap. This is because if the memory was kept full at all times, you'd have to move the data to storage/swap before you could use it for a new process.

Hehe. Thanks for another thing I already know. I have a tendency to over simply instead of baffle with science. Yes, things are swaped out, but the core processes, that means in general everything required to alt tab to another open window, is kept in active RAM just in case. What is swapped out generally is supplemental information, like the other five thousand tabs you have open in Chrome, not the active one.

Programmers optimise it so that they reduce the active memory footprint without impedding performace, Windows further optimises it if it's required, but generally speaking, with more RAM, Windows doesn't need to optimises what's active and what's not.

And finally, the increased performance by having more RAM has an added benefit, and that's reduced swap transactions, which are by far the lastest bottleneck of memory operations. With 6GiBs of RAM you can, althrough I recommend you don't, remove the swap file competely.

[ta2]

I know... you just basically regurgitated what I said.

Anyway, back to helping the OP.

OCZ Obsidian 6GB (3x2GB) DDR3 PC3-12800C9 1600Mhz - £108.33 delivered

That looks like a steal.

[nightkhaos]

I know... you just basically regurgitated what I said.

Reguritated and expanded to 3 paragraphs with supplemental information and examples! Ya gotta admit, that's skill.

OCZ Obsidian 6GB (3x2GB) DDR3 PC3-12800C9 1600Mhz - £108.33 delivered

It's CL9 through... still, I doubt anyone can really notice much a difference between CL8 and CL9 RAM.

[kalniel]

It's CL9 through... still, I doubt anyone can really notice much a difference between CL8 and CL9 RAM.

What speed do you actually run RAM at during an OC? If it's under 1600 then it'd probably run at least CL8 anyway.

[FREEJAFFA]

I know... you just basically regurgitated what I said.

Anyway, back to helping the OP.

That looks like a steal.

and this works well with the Gigabyte UD5?

[nightkhaos]

What speed do you actually run RAM at during an OC? If it's under 1600 then it'd probably run at least CL8 anyway.

Depends on the OC. Mine runs at 1440 CL7, see below. It's rated for 1600 CL8.

[Fraz]

Clearly you have no understanding of thermal dynamics, a push pull, i.e. 2 fans, system on a heatsink will result in increased airflow.

Hmm, I read some quite good tests that showed having a twin fan setup on your heatsink was entirely pointless... it was a while back - I shall have to see if I can find it.

It makes some reasonable sense to me - if you've got a fan pushing however many CFM into the ~enclosed volume of the heatsink, is a second fan really going to achieve anything extra by sucking the same number of CFM out the other side? The air has to come out of the other side anyway. Fans are good at "refocusing" airflow, but in the ~1 inch of heatsink, is there really any point in second fan? It seems doubtful to me.

[nightkhaos]

Hmm, I read some quite good tests that showed having a twin fan setup on your heatsink was entirely pointless... it was a while back - I shall have to see if I can find it.

It makes some reasonable sense to me - if you've got a fan pushing however many CFM into the ~enclosed volume of the heatsink, is a second fan really going to achieve anything extra by sucking the same number of CFM out the other side? The air has to come out of the other side anyway.

No, what the second fan does is pushes air out of the way from the exit point, inrease the fludity of the flow. Not by much I'll admit. There is also the added advantage of if the fans are going at different speeds you can create a postive pressure enviroment within the heatsink, which can cause the air to stagnate slightly while it's flowing through the heatsink, causing a slight increase in thermal transfer, or in the case of a negative pressure enviroment, it causes the air to be sucked out, which will actually cause a slight vacuum, and this increase the airflow, also increasing the thermal transfer rate.

The advantages are small, but they are not zero.

[Fraz]

No, what the second fan does is pushes air out of the way from the exit point, inrease the fludity of the flow. Not by much I'll admit. There is also the added advantage of if the fans are going at different speeds you can create a postive pressure enviroment within the heatsink, which can cause the air to stagnate slightly while it's flowing through the heatsink, causing a slight increase in thermal transfer, or in the case of a negative pressure enviroment, it causes the air to be sucked out, which will actually cause a slight vacuum, and this increase the airflow, also increasing the thermal transfer rate.

The advantages are small, but they are not zero.

Doesn't seem worth the bother to me. I expect that for an arbitrary choice of heatsink you are just as likely to cause harm as good with push-pull fans. I really don't believe you can apply some vague hand-wavy physics to something as complicated as airflow. I guess the point I'm trying to make is that it's not a good idea to always assume "more is better". If tests show it does work, then well and good, but don't assume 2 fans are better in every situation.

[nightkhaos]

Doesn't seem worth the bother to me. I expect that for an arbitrary choice of heatsink you are just as likely to cause harm as good with push-pull fans. I really don't believe you can apply some vague hand-wavy physics to something as complicated as airflow. I guess the point I'm trying to make is that it's not a good idea to always assume "more is better". If tests show it does work, then well and good, but don't assume 2 fans are better in every situation.

Well that's a good assumption. Fortunately I do always do the research when recommeding push pull. The two heatsinks I recommend push pull on, the TRUE and the Megahalems, actually do reduce the temperature by about 3 and 2 degrees respectively by being in push pull.

I also think that you should be more careful who you're talking to my friends. I don't do "hand-wavy" physics. I'm an engineer. I base it on facts.

Based upon my understanding of fluid dynamics the problem occurs if there is not enough room between the fans, that is when the "harm" starts to happen. You're not allowing the distribed air to settle. This would probably become a particular problem in thiner heatsinks, like for example the Cogage TRUE. It is likely that in that heatsink there will be no benefit, or push-pull will run slightly hotter.

I understand what you're saying, throughing more fans at the problem won't make it go away. That's why I usually take considerable care checking the airflow of my case when setting up a rig. I often go as far as to replace every single fan in the case, like I did with Cy.

Airflow is complicated, hard to understand, and to be perfectly honest, impossible to predict.

[Fraz]

I also think that you should be more careful who you're talking to my friends. I don't do "hand-wavy" physics. I'm an engineer. I base it on facts.

Well, I'm a physicist. As you said at the bottom of your post, airflow is impossible to predict, and yet you are - broadly speaking - predicting it:

Clearly you have no understanding of thermal dynamics, a push pull, i.e. 2 fans, system on a heatsink will result in increased airflow.

If you have done actual tests that show an improvement, then fair enough.

[nightkhaos]

Well, I'm a physicist. As you said at the bottom of your post, airflow is impossible to predict, and yet you are - broadly speaking - predicting it:

I'm predicting it based upon experimental results. How is that different to what you do as a physicist in your day to day job?

But yousa point is well noted.