Ram... 64gigs? should I really?

Toggle sidebar Toggle sidebar
ZandaThebigbluePanda

ZandaThebigbluePanda

Reputable
May 21, 2014
487
0
4,860
Hey guys so my computer just died on me.
it was a gigabyte gaming gt + i7 4790k.
seems just my mobo is dead, I've pulled my 980s, i7, psu, and ram out.
but I basically said that it was time for an upgrade.

I'm so far looking at this:
i7 6800k
Gigabyte GA-X99M-Gaming 5
Deepcool Gamer Storm Captain 240EX AIO Liquid Cooling

but I don't know if I should get 64gb ram.
I'm used to 16 but alot of the time I was capping out, now I know I don't need it but I found a really nice kit for $400ish US.

I wanted to know what you guys thought about it, and what type of improvements you feel I may have with 64gb's.

my uses from highest to lowest priority:
gaming
website management
html website development
game server management (cs:go, gmod, rust, minecraft)
high use of the internet (30-50 tabs open at once)
modding (skyrim, kotor- making mods I mean, not just installing other peoples)
3d modeling
streaming
photo shop
 
Solution
JackNaylorPE
The Phoenix is a pretty good board, tho the aesthetics may take some getting used to. With 41% of Asus Strix board owners giving it the worse possible 1 egg rating, I'd take that of the list. Asus Strix = 41% / Asus X99-A = 15%

That's 41% is way way way too high a number to be excused by "some board owners being inexperienced". How could it be that all the inexperienced users are choosing the Strix and aren't choosing the X-99A ... that's almost a 3 to 1 ratio.... surely well beyond the statistical standard deviation.

I'd take a good hard look at:

Gaming Pro Carbon has 56% 5 eggs / 15% 1 egg
Asus X99-A has 53% 5 eggs / 12% 1 egg

For the inexperienced user argument to be in any way relevant, there would have to be a explanation...
Sort by date Sort by votes
For Gaming, the 6700k and 6800k run neck and neck in most games... but when there is a significant difference, it usually favors the 6700k. For example.... with GTX 980 at 4k, 6700k gets 31 fps to 6800k's 24 in SoM. 3D modeling would generally favor the 6800k ... PhotoScan favors the 6700k

http://www.anandtech.com/bench/product/1543?vs=1727

RAM requirements are typically dictated by the specific application. I'd go with 2 x 16GB modules in 6700k... more sticks is a bigger load on memory controller and can affect OCs. Also high speed / low CAS RAM stability goes down with more sticks.... ones that pass muster are very expensive.

If using an AIO, I'd highly recommend getting one that:

a) is expandable
b) Has copper radiator
c) Uses 140mm fans if possible

Swiftec H240-X2 satisfies all of the above.

Note that the Gigabyte board has some issues with SLI due to close PCI slot spacing. See newegg user reviews.
 
Upvote 0 Downvote
Have they upgraded the support, believe that mobo is still at 32GB max. If they have added the 16GB stick support then 64 is still a bit of overkill for your plans... 32 should be fine, if it's in the budget, and you did go 64GB you could create a RAMDrive which can prove useful and speedy for PhotoShop, the 3D modeling and other things like video. Otherwise would go 4x8GB in prob 2800, something like the GSkill RipJaws 5
 
Upvote 0 Downvote


Money isn't an issue for me, and the kit I was looking at (4x16gb 3200mhz CL15) is really cheap.
its $100 more than a kit of 4x8gb.

also all sources I looked at for that mobo says it supports 128gb's.
and in terms of a ram drive I do want to look into it, but for my needs I doubt I would need to.
I am a enthusiast in terms of photoshop, 3d modeling, and streaming, not a pro.
nor would I ever use them for professional work.


also while I'm gaming I normally have a youtube video on my other monitor, with like 20 tabs open.
and I'd much rather spend the little extra money on 4x16gb kit rather than have to upgrade later.
and I tested my rig (with out dual channel) on 24gb ram and I capped out on that while playing subnautica and doing the normal stuff I do.

I almost always have these open:
RDP
a game
chrome
firefox ( I use firefox and chrome at the same time, normally to test my website)
steam
geforce experience
OBS
netflix app

 
Upvote 0 Downvote


The thing is .... a $380 set of RAM is kinda "out of place" in a $165 MoBo.

http://www.newegg.com/Product/Product.aspx?Item=N82E16813128748

Newegg says the board supports DDR-3000

http://www.gigabyte.us/products/product-page.aspx?pid=5221#sp

Gigabyte says

Support for DDR4 2800(O.C.) / 2666(O.C.) / 2400(O.C.) / 2133 MHz memory modules
Click to expand...

Giga's Memory Support List )QVL) for that board goes to DDR4-2800
http://www.gigabyte.us/products/product-page.aspx?pid=5221#memory%20support%20list

Highly rated alternatives with higher speed memory support that I have previously considered for X99 ($450 max)... listed in order of best user rating by board owners on newegg.

MSI X99A XPOWER GAMING TITANIUM ($429) DDR4-3466
http://www.newegg.com/Product/Product.aspx?Item=N82E16813130935

ASUS X99-A ($225) DDR4-3200
http://www.newegg.com/Product/Product.aspx?Item=9SIA2F84D69020

MSI X99A GAMING PRO CARBON ($329) DDR4-3466
http://www.newegg.com/Product/Product.aspx?Item=N82E16813130934

SI X99A SLI KRAIT EDITION ($229) DDR4-3333
http://www.newegg.com/Product/Product.aspx?Item=N82E16813130840

ASUS X99-A ($225) DDR4-3200
http://www.newegg.com/Product/Product.aspx?Item=N82E16813132831

Highest rated Gigabyte board is the
GIGABYTE GA-X99-Ultra Gaming ($279) DDR3-3600
http://www.newegg.com/Product/Product.aspx?Item=N82E16813128928

User ratings by board owners are by no means the "be all and end all" of selection criteria especially when there are too few reviews to be statistically significant. But if a large % of users are having problems and wish they hadn't bought that board, especially when they are the same problems, it's worth finding out why and avoiding the ones with large amount of negative reviews. Suggest narrowing down your list, finding good tear down reviews and learning as much as you can before deciding
 
Upvote 0 Downvote


yeah, didn't really look to much at that mobo, now looking at either the Gigabyte GA-X99-Phoenix SLI Intel X99 LGA2011-3 DDR4 ATX or the ASUS STRIX X99 Gaming
 
Upvote 0 Downvote
And that's one of the biggest problems with NewEgg reviews....you have many, many newer builders building rigs and not knowing what they are doing. You see tons of problems with things like DRAM, drives, etc, and often the only problem is the builder never updated the BIOS. Many of the X99 mobos out on the shelfs, have an older BIOS that was based more on the older and earlier DDR4 (mosyt all of which was made to base JEDEC specs, where most all of the newer lines and upgraded models are made to higher performance criteria, (tighter CL and timings, higher data rates). The mobo makers are constantly updating the BIOS for the newer DRAM, drives like the multitude of M.2, etc drives coming out, newer USB devices, et al. Then you have the new builders simply loading the drivers and utilities off the disk that came with the mobo, and to be honest I haven't built with a board (even a brand new model) in quite awhile, that had a disk with drivers that were up to date, so the new builders are also using outdated drivers 😉 It's no wonder there are so many negative reviews. This is also one of the biggest reasons knowledgeable systems folks tend to ignore reviews at NewEgg, Amazon, etc. Oh we take a look, but when you get into the cons and negative remarks, you can more often than not tell what the problem is by simply seeing what they say. Could give examples galore.

With X99 in particular, it's also a good idea to check with the manufacturer, not all have good/full support built in for 16GB sticks for the simple reason there basically weren't any really available when the boards were initially released. Advertising is slow to catch up with the mobo's actual capabilities.

As far as the amount of DRAM, if the mobo and budget supports it, no reason not to go for it. 😉
 
Upvote 1 Downvote
The Phoenix is a pretty good board, tho the aesthetics may take some getting used to. With 41% of Asus Strix board owners giving it the worse possible 1 egg rating, I'd take that of the list. Asus Strix = 41% / Asus X99-A = 15%

That's 41% is way way way too high a number to be excused by "some board owners being inexperienced". How could it be that all the inexperienced users are choosing the Strix and aren't choosing the X-99A ... that's almost a 3 to 1 ratio.... surely well beyond the statistical standard deviation.

I'd take a good hard look at:

Gaming Pro Carbon has 56% 5 eggs / 15% 1 egg
Asus X99-A has 53% 5 eggs / 12% 1 egg

For the inexperienced user argument to be in any way relevant, there would have to be a explanation as to why all these inexperienced users are drawn to some motherboards and not others.... I have yet to hear that explanation. Why is it that some MoBos can have > 40% 1 eggs and some can be in single digits. Why aren't these inexperienced users buying the single digit ones ? Why are all the > 40% 1 eggers buying those boards ?

If we say that 10% of all reviews have issues due to user inexperience.

Asus Strix = 41% - 10% = 31% legit problems
Asus X-99A = 15% - 10% = 5% legit problems

If anything, accounting for inexperienced users only makes the high scorers look worse.

Knowledgeable system folks (Knowledge being defined as information, understanding, or skill that you get from experience or education) have one thing in common. They use information to make judgements. Experience is valuable but hard to obtain since you would need to have built a statistically significant number of builds for every MoBo model.

Knowledgeable users evaluate the validity of the information available but, as indicated above, by no means take it at face value. As was stated, the % are not the be all and end all of the decision making process. Reading the reviews, a knowledgeable user can immediately see when the problem is the user and not the board. So lets try that.

Looking at the Strix reviews .... 8 on the first page were for **working** boards that failed in the 1st 2 months. Seems the user managed to build and use it just fine . Not working 1 week or 2 months later, RMA issued,some DRAM slots not working ... that ya gotta lay on the board.

The GTX 970 was the most popular graphics card ever. The Asus Strix, MSI Gaming and EVGA SC 970s all had 7-8% 1 egg user ratings .... the Gigabyte G1's was 3 times higher ? So what accounts for the difference ?

A. We can assume that these inexperienced users were overwhelmingly buying the Gigabyte G1 cards and all the smart users were buying the other 3.

or

B. The G1 card had issues that the users of other cards did not experience.

I'd bet the farm on B.

Unless you buy a board immediately after it comes out, you can expect never to get all drivers up to date.... even then, sometimes updates are already available by the time it hits the doorstep. I recommend never buying early... let those anxious folks live on the bleeding edge and experience the problems of 1st stepping boards.

To check support, the BIOS notes generally indicate any changes that result in support for faster speed RAM, no such are listed. Only RAM notation was for F4B Beta that added support for 16 GB in 08/2015. The QVL has not been updated since 12/2014

Note that version 1.1 of the board Giga's site also says support only up to 2800. With newegg saying 3000, if you still leaning to that board, I'd give them a call on Tuesday to confirm (Closed today - Monday in US)

EDIT: regarding the Pro Carbon and X99-A.... I forgot about ya GFX cards, with two air cooled 980s, I'd want a board with a larger slot spacing. The Phoenix has adequate spacing
 
Upvote 0 Downvote
Solution
Actually, no. The BIOS notes seldom ever list more than a single item or two that's included in the update, there are often tons of little updates, and often the majority are for DRAM and various new XMP profiles from new or updated models of DRAM. Updates to accommodate drives, GPUs, etc are also included. Typically you might see "Improved System stability" or maybe "Improved memory compatibility" they don't go into details, it more of a generalist note(s). Detailed BIOS update logs are a long lost thing of the past. Need to be realistic here, you talk about DRAM and the BIOS, for XMP to work, the BIOS has to be able to take the SPD info (XMP profile) and implement those settings into the BIOS, the initial BIOSs on the Z170 mobos were primarily geared to know DDR, both JEDEC and what little was available in higher performance sticks. Since Skylakes release we've seen DDR4 data rates rise, seen continual improvements in timings vs data rate, i.e. originally we saw 3200 sticks with CL16, since then it's become availa in CL15 and not too far back CL14, and may well see CL13 before long, all these updates/upgrades have to be programmed into the BIOS.

As far as NewEgg reviews, I'd just suggest watching them for awhile, they are very fickle and things jump around alot, just what a couple weeks ago, NewEgg showed the MSI Gaming XPower Titanium as the 'Best Rated' Z170, now it's dropped to about 13th (according to NewEgg), other mobos have started with slow/low reviews from users, then soared, and at the same time many got early on rave reviews and then sunk to the lower ends. A good example was the Z87 Hero, that was the first of the line for the Hero, it started out with middling Egg reviews and then soared to be up at the top of the best rated. (Same with the Z97 Hero), and am guessing same will happen with the Z170 Hero. I've got all three and have helped many, many with all three right here in the forums (often out of date drivers and BIOSs were the problem 😉 ). The X99 Strix is mentioned above and as a new line of mobo, appears to have had some problems, but I'm looking at X99 for a new build for myself right now, and both it and the Sabertooth on are my initial list of mobos for consideration, along with a couple of Rock mobos and a GB or 2.
 
Upvote 1 Downvote


The Deluxe has often been better rec'd than most of the RoG boards ... not up there with the Sabertooth but oft better than most. The 2nd slot in the Deluxe 2 is a bit too close to the 1st for air cooled 980s. From the pic here, it appears that you can use the 1st and 3rd slot for 2 cards

http://www.bit-tech.net/hardware/2016/08/01/asus-x99-deluxe-ii-review/2

 
Upvote 0 Downvote


All I can say is that hasn't been my experience ... fixing bugs is not something manufacturers tend to bring attention to whereas expanded features is. Nevertheless, it's certainly worth checking. It's answered my question many a time. As is the case **with the board in question** where they noted the addition of 16 GB module support which is something I had also wanted to check.

NewEgg showed the MSI Gaming XPower Titanium as the 'Best Rated' Z170, now it's dropped to about 13th (according to NewEgg
Click to expand...

Again, if ya look at the statement in context, it's all about how the boards actually compare ... not the web sorting function which newegg previously acknowledged didn't appear to be working properly.

First things first, as I have always stated, the more reviews a component receives, the more accurate the numbers are. Back then it was at 4% 1 egg, now it's at 9% ... not exactly a wide swing. Let's ignore the web sorting function which doesn't work properly and lets see if your argument holds water when we look at the actual data which the site is attempting, and failing to properly, sort.

In it's price range ($200-300) using the web site sorting function by best reviewed, the Titanium is 10th ... Of the 10, 2 of them have 4 and 3 reviews. One of them is the new Whetstone version of the Asus Hero w/ 3 reviews ... so what do we place our reliance on... the statistically insignificant 3 people who bought the new Whetstone version or the 215 folks who bought the Hero where 40% gave it 5 eggs and 37% who gave it 2 eggs or less ?

We can't ignore the science of statistics; it's not about blindly looking at a ranking ... it's about using them in a meaningful way and ignoring things data that is obviously statistically insignificant. It's also about actually looking at the real numbers not the web sorting function which obviously doesn't produce accurate results:

Ranking then by "Best Rating" between $200 and $300, using the sorting function puts 3 MSI boards in the "top ten". Does the actual data support this web sorting function ? ... not even close.

I don't see the M5 being ranked higher than the 7. I see the 8th ranked Asus Ranger as equal to the M5 from a user satisfaction % and don't think it should be ranked 3 places higher than the Ranger.

5th place (111 Reviews) MSI Gaming M5
5 eggs - 45% (50)
4 eggs - 16% (18)
3 eggs - 11% (12)
2 eggs - 8% (9)
1 egg - 20% (22)

7th Place (104 reviews) MSI Gaming 7
48% (50)
16% (17)
9% (9)
9% (9)
18% (19)

10th place (54 reviews) MSI Titanium XPower
69% (37)
4% (2)
6% (3)
13% (7)
9% (5)

Huh ? What math are they using ? If 3 eggs are neutral....

The 10th place Titanium has 73% positive and 21% negative
The 7th place Gaming 7 has 64% positive and 27% negative
The 5th place Gaming M5 had 51% positive and 28% negative

Isn't 73 bigger than 51 ? ... Isn't 21 less than 28 ?

How is the 5th place 20% 1 eggers better than 7th place 18% or the 10th place 9% ?
How is the 5th place 45% 5 eggers better than 7th place 48% or the 10th place 69% ?

If 69% is 53% better than 45%, why is 45% sorted higher ?
If 9% is 122% better than 20% why is 20% sorted higher ?

Let's try this ... apply the following values to the egg rating ...I'm thinking better of worse than average and 3 eggs getting a neutral 0 ... more + points for being better than average. more - points for being worse than average

5 eggs = +2
4 eggs = +1
3 eggs = 0
2 eggs = -1
1 egg = -2

So 69 (%) x +2 = 138 and so on

The 10th place Titanium gets +111 ... 7th place Gaming 7 gets +67 and 5th place Gaming M5 gets a +58

Use number of eggs ... i.e 69 (%) x 5, 4(%) x 4 etc and the 10th place Titanium gets +414 ... 7th place Gaming 7 gets 367 and 5th place Gaming M5 gets a 358. Use any kinda math you want, you can't make the 5th ranked card *as sorted* better than the 7th or 10th place card *as sorted*.

The Sabertooth has always been a great board which attracts serious users (included myself) and I oft recommend it. But other than that .... no board between $200 and $300 has better review history than the Titanium. The actual data conclusively and irrefutably contradicts the sort ranking.

And yes, I have called and asked newegg .. was 2-3 years ago when they actually answered the phone. Asked how it works and what I was seeing ... the support person I spoke to agreed that the sorted ranking "wasn't making sense" ... they said they'd get back to me but I never did get a follow up call.
 
Upvote 0 Downvote
Exactly why I stated it's rather fickle in how it gives ratings. Very deceptive. This doesn't really change the fact that again it comes down to the users/builders, all too often negative reviews come from builders that simply don't know what they are doing, if you actually read the reviews, negative reviews are often a simple fix (BIOS, drivers), boards are reported DOA which are (from what we see in the forums) often missing a standoff or the like, many times an underpowered PSU, mobo won't run memory at it's rated spec (no XMP enabled, trying to run faster DRAM than the CPU can handle), USB not working or SATA not working (wrong/old drivers), GPU . As I mentioned before many are common newb mistakes. They get frustrated and write a negative review. For outsiders that see the reviews and don't really know systems they take the complaints at face value, rather than give thought to troubleshooting steps, which is hard as few negative reviews give full system specs. Those of us that help people troubleshoot problems more often than not ask for for system specs pertinent to what the problem or problems appear to be. The big thing to look for is consistent problems that can identify a fallacy in the actual mobo and overall you seldom find that, when you do then that particular mobo calls for a closer look.
 
Upvote 0 Downvote
I understood what you were saying the in the previous 2 posts. What I was looking / asking for was reasoning to support this claim. So far, this has not been forthcoming. What is it that you think accounts for these uneducated users overwhelmingly buying only certain boards and ignoring others ?

Seems we agree that the web site sorting is faulty ... I just ranked all Asus Z170 boards using the web sort feature and the ASUS Z170-PREMIUM came in last. It only has 9 reviews, but yet it scores 56% 5 eggs and 0% 1 egg ? The sort function should therefore place that board way up near the top with 0% negative reviews and > 50% 5 eggs, but it was listed last .... so the sorting function is clearly deficient.

However, the numbers are the numbers ... math isn't fickle. If people say they are unhappy with their purchase, you can count the results. Math doesn't have gray areas, the numbers are the numbers, you can verify them. If math was fickle, I'd be scared to practice my profession... as would other engineers, accountants, scientists etc. I have read all the reviews for many boards in several price ranges and I see nothing to suggest the knowledge deficient users populate the reviews more for some boards than others.

The science of statistics does have a gray area; it's called the standard deviation or reliability which is depends on sample size. Such is generally presented as the data "being accurate to within a +/- %". But, with adequate sample sizes, this gray area represents a rather narrow margin which shrinks more and more as the sample size grows larger over time. This is why we have to ignore small sample sizes.... at 100, I'm very comfy ... at 50 it's a pretty good bet, below 20 I wouldn't even look.

A student's grade of 85% might not accurately indicate his knowledge level ... but when ya have 10 grades between 70 and 85% for one guy and 10 grades between 85 and 100% for another, I think it's clear that one's knowledge level is superior to another.

The Titanium remains one of the best boards available with 1 egg ratings still in single digits. How many Z170 boards are there between $200 and $300 that have:

a) A statistically significant number of reviews
b) 1 egg reviews in single digits
c) more than 2/3 of reviews being 5 eggs

Answer: Sabertooth and Titanium XPower....

But more importantly, is there an explanation as to why this mass of uneducated users are apparently not buying either of these boards ? It would appear that they are as the 1st newegg review I read on the Titanium was about a dude trying to plug a 6+2 PCI-E cable into the MoBo. Clearly he / she wasn't at the top of their game.

So the relevant question one has to ask is, for example, why are 30% of owners extremely unhappy with the Z170 Hero and only 6% extremely unhappy with the Z170 Sabertooth ? That's a 5 to 1 ratio. If I'm to entertain this position of knowledge deficient users being the cause, I need an explanation as to why these knowledge deficient users are purchasing the Hero in droves and apparently completely ignoring the Sabertooth ?

You find X99 boards with similar variances .... X99-A gets 12%, Rampage Extreme gets 3 times as many with 36%. Even among the same board line / different chipset ..... the Z170 Sabertooth gets 13% ; the X99 version gets 31%. Why are X99 Sabertooth users 2.5 times as unhappy than Z170 owners ? These variances are well outside the standard deviation

Is there any data or even reasonable hypothesis as to why the distribution of uneducated users isn't even among all brands and boards ? Is there any data or even reasonable hypothesis as to why the distribution of uneducated users isn't even among the 4 AIB 970s presented above ? ... or better said why 3 of them are neck and neck at 7-8% and 1 stood out with 3 times the 1 egg ratings .... sampling size all greater than 400 (MSI 621 / Asus Strix 470 / EVGA SC 427 / Gigabyte G1 * ) as the competition ? How can we possibly conclude that all the smart builders were buying Asus, MSI, EVGA 970s and all the knowledge deficient ones were buying the Gigabyte ?

* the original G1 Windforce is no longer there just the newer OC Edition, but it originally was running slightly behind the MSI in # of reviews B4 the 10xx series came out.

If ya can offer a reasonable hypothesis which would explain why one board / GFX card gets such a staggering amount of knowledge deficient users such as to result in 5:1 and 3:1 ratios of 1 egg reviews, I would be happy to to take another look and perhaps get a better understanding of how this might be relevant .... but I have now asked this question several times and no explanation has been put forth which would explain this presumption.

 
Upvote 0 Downvote
Popularity plays in, so do website/publication reviews, past performance of the model line. There are lots of tangibles that play in to why one buys a mobo. There are also many problems with using NewEgg reviews as a basis, as you and I have both agreed on, their ranking system is flawed and without knowing what type of formula they use to determine rankings, it can come down to luck of the draw. Same with how they rank best selling products - there is no way to determine what is actually happening as far as what 'Best selling' means - it used to be based on overall sales but appears they have changed their 'formula' less than a year or so ago as you now see many relatively new products popping up as the 'Best Selling', i.e. the Hero/Whetstone a newer Z170 (not even on the Asus website yet) is there on Newegg as the 5th best selling. Is it based on sales for the day, week, month or what? The mobo hasn't even been out but a month or two.

As far as statistics, what exactly determines a 'significant number of reviews' and how would that apply to overall sales, number of reviews vs overall sales. You seem to think and position your hypothesis on sitting number of reviews. And in part you appear to base number of reviews vs overall sales.

Just as an example which I refered to earlier, some mobos start off with great reviews and head downhill, others start with lousy reviews and climb....and as mentioned also the number of reviews have nothing to do with overall sales, yet you seem to want to compare a mobo with 50+ reviews vs with 5 times as many reviews - where will the XPower stand when it hits 250 reviews? Or the Sabertooth for that matter it has what maybe 40.....

You just never know what will happen based on the reviews, as I mentioned before with the Z87 Hero it started with so-so reviews (a brand new line of mobo) and many ridiculed it as you do with the Z170 Hero - on the other other hand MSI had it's Z87-G45 which fired out great guns comparatively - when the dust settled in the end Hero walked away as better (according to your formula above by 10% positive over the G45 and total reviews the G45 ended up with 103 total reviews vs the Hero's 374.

What would be nice if they were to actually list total sales of a given mobo. If you were to take number of reviews and try and to compare that to overall sales then the Hero apparently would have a ton more happy customers (5:1).
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom