1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Hardware Some questions about our new CPU cooler test systems

Discussion in 'Article Discussion' started by Combatus, 30 Apr 2015.

?

Select your preference below

  1. De-lid the LGA1150 CPU

    5 vote(s)
    9.8%
  2. Don't de-lid the CPU

    42 vote(s)
    82.4%
  3. Show all the results in one graph

    21 vote(s)
    41.2%
  4. Show graphs for either air or liquid coolers

    2 vote(s)
    3.9%
  5. Split the graphs but include results of the cooler on test in both

    16 vote(s)
    31.4%
  6. Something else (please comment)

    9 vote(s)
    17.6%
Multiple votes are allowed.
  1. Combatus

    Combatus Bit-tech Modding + hardware reviews Lover of bit-tech Super Moderator

    Joined:
    16 Feb 2009
    Posts:
    2,761
    Likes Received:
    89
  2. DbD

    DbD Minimodder

    Joined:
    13 Dec 2007
    Posts:
    519
    Likes Received:
    14
    Don't de-lid - we want to see how the cooler would perform with the sort of cpu we will be trying to cool which won't be de-lided. That said you need something that produces a lot of heat to test maximum cooling capacity.
    I would mix up air and water coolers but just make it clear what is what. tbh when I look at a cooling chart I'd like to know air/water, cost and noise.
    I speak as someone who has always customised their case for cooling/noise, o/c and used a custom cooler, although not the to the extreme - these days I am lazy and just use the m/b one click o/c controls.
     
  3. loftie

    loftie Multimodder

    Joined:
    14 Feb 2009
    Posts:
    3,173
    Likes Received:
    262
    Don't delid the processors, it's better to know how the coolers will perform with processors that are standard. Otherwise you could introduce undervolting too. I'd add a footnote mentioning Intel's recent design choices.

    As for the coolers, i'd have them ranked in order of temperature but i'd introduce another colour to the graph, so you have one for water, one for air, and one for the current one being tested.
     
  4. blacko

    blacko What's a Dremel?

    Joined:
    30 Jul 2012
    Posts:
    138
    Likes Received:
    2
    Don't delid.
     
  5. Corky42

    Corky42 Where's walle?

    Joined:
    30 Oct 2012
    Posts:
    9,648
    Likes Received:
    388
    Personally I wouldn't de-lid the CPU on a test system, consistent result over a period of time are far more important than lower temperatures, wouldn't damage to the exposed core mean having to throw away past results and start again?
     
  6. Dr. Strangelove

    Dr. Strangelove What's a Dremel?

    Joined:
    13 Mar 2005
    Posts:
    345
    Likes Received:
    1
    Don't de-lid.
    Also keep the coolers in the same graph but color them differently.
    Might be nice to add the cooler price to the graph, so you can not only compare performance but also price. (you could put it to the left inside the individual bars)

    Perhaps you should only show the 10-15 coolers which performed similar to and/or same price range of whatever you are reviewing and then have a link to the full blown ridiculously long graph if someone wants to compare all the coolers in the universe.
    -Jacob
     
  7. Hustler

    Hustler Minimodder

    Joined:
    8 Aug 2005
    Posts:
    1,039
    Likes Received:
    41
    De-lidding is a minority sport amongst the minority pursuit of overclocking, so I really don't see how such results would be relevant to 99% of your readers.

    Even if it showed them what could be if they did de-lid..realistically how many of them would risk a near £200 CPU for the sake of running cooler?
     
  8. ChaosDefinesOrder

    ChaosDefinesOrder Vapourmodder

    Joined:
    6 Feb 2008
    Posts:
    712
    Likes Received:
    8
    How about a trial with an additional graph of "Delta-T per Pound-as-reviewed"?

    That would definitely make the relative value for performance a lot clearer...
     
  9. Deders

    Deders Modder

    Joined:
    14 Nov 2010
    Posts:
    4,053
    Likes Received:
    106
    If you are going to de-lid then that should be in a separate chart to avoid confusion. You should keep a normal CPU for testing so that people can see what results they are likely to get without de-lidding.

    It's nice to see how air coolers match up to AIO water systems. I think the best way for this would be keep them on the same chart but have them a different colour.
     
  10. Fingers66

    Fingers66 Kiwi in London

    Joined:
    30 Apr 2010
    Posts:
    8,873
    Likes Received:
    1,054
    Don't de-lid so that the results are comparable to the vast majority of people that don't do it. It also allows us to (roughly) compare results from other generations of processors (newer or older).

    Regarding the charts, I see your problem regarding them getting too big so splitting the charts into AIO and air cooled results is sensible.

    If you are going to only show results for the same cooler type, please provide a link on the chart page to the most recent alternative cooler chart so a reader can quickly compare for themselves. This way you only need to show, for example, AIO results in an AIO review but readers can look at the latest air cooler chart if they want to compare.

    To be clear, I have no problem splitting the AIO and air cooler charts but we still need to compare both so a link to the latest chart for both (outside of reviews) would be good if you are not going to include both in the same review.
     
  11. John_T

    John_T Minimodder

    Joined:
    3 Aug 2009
    Posts:
    533
    Likes Received:
    23
    Q1 - De-Lidding:
    I'm with the consensus here - don't de-lid.

    Like Hustler said, even in a place like this, de-lidding is going to be a ridiculously niche pursuit. I know some posters in here have the budget for multiple 4k monitors with the multiple 980's to power them - then change half of it every few months to gain another couple of fps. I'm sure those people will love it, as risking a brand new component won't be a big deal to them. I don't think that's most of us. If I spend two, three (or more) hundred quid on a new CPU then I'm going to be pleased as hell with it - the very last thing I'm going to do is break out the tools and hack it open! I've happily overclocked over the years, (which admittedly is not without risk) but that's not quite the same thing.

    I can understand the technical argument for what you're suggesting, but I do feel it will inevitably give unrealistic expectations for the vast, overwhelming majority of us. What's the point of adding the caveat: "Remember to add, roughly, about 10 or so degrees to what you see here, approximately, give or take" - you may just as well just give the accurate, real-world figure.

    Q2 - Graphs:
    Not fussed either way to be honest - are large graphs a problem? I wouldn't particularly mind seeing separate all air and all water charts, (perhaps with a link to the other chart) but as you said, I think different colours on one chart would work better. Could be pale blue for air and dark blue for water, (or whatever).
     
  12. Spigsy

    Spigsy What's a Dremel?

    Joined:
    28 Sep 2009
    Posts:
    59
    Likes Received:
    1
    I'm on the bandwagon! I think the results using a de-lidded proc would be irrelevant to the majority of your readers. However, I would love to see an article once per (Intel) generation that explains the process and also gives an idea of the gains to be had.

    I think you should stick with the one chart irrespective of size as I'm too lazy to cross-reference between charts. Using different colours for water/air colours would be inarguably sensible.

    Kudos for asking the opinions of your readership- I'm more of a lurker here but have visited the site pretty much every day for the last seven years and it's always pleasant to be asked your opinion!
     
  13. Xir

    Xir Modder

    Joined:
    26 Apr 2006
    Posts:
    5,412
    Likes Received:
    133
    Pretty much the same as said above...

    Q1 don't delid, use "normal" CPU's. That said, isn't the i5-4690K supposed to be replaced this summer?

    Q2 I'd like to see a differentiation between air and water, be it in colour or in split graphs.
    With powerfull yet quiet aircoolers reaching the efficiency of similarly priced AIW's, I don't see the benefit of using a more complex and probably louder AIW system, so I'd like to be able to rule them out at a glance.
    At least I'd like to know which are which (colourcoding).

    Oh, the same goes for SLI/crossfire/multiGPUsetups in your Graphics-charts, please coulour code those too.

    Last whish: bring back the Hardware Buyers Guide :D
    ...quarterly or tri-annual updates would suffice.
     
  14. spolsh

    spolsh Multimodder

    Joined:
    4 Feb 2012
    Posts:
    1,559
    Likes Received:
    821
    Don't delid. Most people interested in delidding would be more interested in custom loops parts reviews I think - which could be a kinda-cool idea to do as well ...

    I like spigsys idea of a bang per buck comparison, think that would maybe throw in some surprises.

    I think colour to differentiate - maybe green for air coolers, blue for water, red for current cooler on test.
     
  15. TheStockBroker

    TheStockBroker Modder

    Joined:
    19 Nov 2009
    Posts:
    1,533
    Likes Received:
    110
    Wow, not what I expected from BT.

    Seems I'm the only one who would want the de-lidded results, as in my opinion these better reflect the true capacity/quality of the cooler... and this is what is under review!

    I understand that you will have to pander to the majority, like any business, but I definitely appreciate you asking the question :)

    Edit: FYI Delidding is not risky, nor should it be reserved for custom water loops! A good AIO will allow a delidded CPU (with sensible voltages, mind) reach it's max overclock while being well within it's thermal specification.
     
  16. SchizoFrog

    SchizoFrog What's a Dremel?

    Joined:
    5 May 2009
    Posts:
    1,574
    Likes Received:
    8
    Q1 - I am glad to be like minded with the rest here and will say 'don't de-lid'. I think a 'stock' level needs to be maintained. What others do with their own CPUs is up to them but once you start to customise it will become very hard to truly evaluate performance, especially against expected performance at home.

    Q2 Part 1 - I would certainly like some way to distinguish AIO coolers from air coolers so I would like them to be different colours.

    Q2 Part 2 - This is where it gets more difficult to know what is best to do as trying to fit everything in to the space of a compact chart often means that specific products have to be omitted. This often leads to many people asking 'How does this compare to ********' in the comments rather than discussing the cooler on test. My main gripe with omitting specific products is that often, very like performing products are featured while maybe lesser performers are the ones to be first left out, leaving it very difficult to judge things across the board. So, whether or not you keep things in one chart ot split them I would like you to remember this if possible...

    Please keep a baseline, value product in ALL charts. In respect to air coolers something like the Silverstone Argon AR01. Then include 1 or 2 of the best performing 'ultra tower', twin fan, tower coolers and to cover all the bases please also include a recommended 'low profile' cooler as well... especially as AIOs will often be used in ITX builds where tower coolers may not fit or will have their cooling performance affected.

    Thanks. :)
     
  17. SchizoFrog

    SchizoFrog What's a Dremel?

    Joined:
    5 May 2009
    Posts:
    1,574
    Likes Received:
    8
    I disagree. a review is supposed to offer a relative comparison and de-lidding will only show the results of a personal experience which may vary to a large extent for all sorts of reasons, type of TIM used for example. AS Bit-Tech will not doubt be unable to review each and every cooling product, it also removes the possibility of comparing figures from Bit-Tech to other sites for comparison.

    I also disagree that it is not risky... It voids the warranty of your product and that is a risk. End of.
     
  18. wolfticket

    wolfticket Downwind from the bloodhounds

    Joined:
    19 Apr 2008
    Posts:
    3,556
    Likes Received:
    646
    I'm not against having de-lidded results as well.
    However, just having de-lidded results when the majority of people reading won't have a de-lidded CPU and wont want to do it (irrespective of how easy and safe it is) just doesn't make sense.
     
  19. rollo

    rollo Modder

    Joined:
    16 May 2008
    Posts:
    7,887
    Likes Received:
    131
    I'd not see a lot of point in reading a cpu cooler review with just de lidded results on it. I'm not personally going to do it anytime soon and the amount who are is a tiny % of a tiny market.

    Sure it would be nice to see on a cpu review when a new one comes out what the gains are but guess that's close to impossible also unless your buying the cpu. As Intel would surely not send you another if you de lidded a review chip.

    It's not that risky but it does void warranty and most just have no interest in messing around that much.
     
  20. RedFlames

    RedFlames ...is not a Belgian football team

    Joined:
    23 Apr 2009
    Posts:
    15,401
    Likes Received:
    2,996
    1 - Don't De-lid... you make a point of testing the cases as they come, the coolers as they come, so test the CPU as it came.

    2 - 1 Chart, If your chart is getting a bit long maybe have the full chart elsewhere and may be just show the cooler being reviewed relative to a few 'Award' winners.
     
Tags: Add Tags

Share This Page