Need advice in deciding if RAM-memory is good enough for my system.

[dardy77]

Hey guys! I recently bought this system, though I wonder if my memory is good enough for the other parts and if you think I should swap them for other, better memories. My purpose with this system is gaming in 1080p.

CPU: Intel Core i7-4770K 3.5GHz Quad-Core Processor
CPU Cooler: Corsair H100i CFM Liquid CPU Cooler
Motherboard: MSI Z87 MPOWER ATX LGA1150 Motherboard
Memory: Kingston HyperX Predator Series 8GB (2 x 4GB) DDR3-2400 Memory CL 11
Storage: Kingston SSDNow V300 Series 120GB 2.5" Solid State Disk
Video Card: MSI GeForce GTX 770 2GB TF OC
Case: Corsair 330R ATX Mid Tower Case
Power Supply: Corsair CX 750W 80+ Bronze Certified Semi-Modular ATX Power Supply

 

[tiny voices]

Your memory is perfectly fine. Changing it will not affect the performance on the computer or game performance.

 

[Jameson Clark]

No that is better than good ram

 

[Mciahel]

Hey guys! I recently bought this system, though I wonder if my memory is good enough for the other parts and if you think I should swap them for other, better memories. My purpose with this system is gaming in 1080p.

CPU: Intel Core i7-4770K 3.5GHz Quad-Core Processor
CPU Cooler: Corsair H100i CFM Liquid CPU Cooler
Motherboard: MSI Z87 MPOWER ATX LGA1150 Motherboard
Memory: Kingston HyperX Predator Series 8GB (2 x 4GB) DDR3-2400 Memory CL 11
Storage: Kingston SSDNow V300 Series 120GB 2.5" Solid State Disk
Video Card: MSI GeForce GTX 770 2GB TF OC
Case: Corsair 330R ATX Mid Tower Case
Power Supply: Corsair CX 750W 80+ Bronze Certified Semi-Modular ATX Power Supply

The RAM for the most part isn't going to change the performance that much. Pretty much as long as its DDR3-1600 its fine. Though I have heard that the CAS timing affects FPS, but only up to 5 FPS so not too much.

TL: DR The ram you have is fine.

 

[Alec Mowat]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

 

[dardy77]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

Well, I wasn't so updated about what CAS latency when I bought my memory and I am willing to buy new if it is worth it for what I will gain from it

 

[Alec Mowat]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

Well, I wasn't so updated about what CAS latency when I bought my memory and I am willing to buy new if it is worth it for what I will gain from it

The higher the frequency, the higher the CAS timings. Basically, you don't need new RAM if you want to lower the frequency and the CL. You can just manually change it in the BIOS.

A higher frequency RAM (like yours, at 2400) will give better performance when compared to a lower frequency RAM with a lower CL. The CL11 timing only applies to the 2400 speed. I would go into your BIOS and confirm what timings and frequency your RAM is actually running at. You can lower the frequency and decrease the RAM timings if you want to compare, but I found ultimately that higher frequency performed better than a 1 or 2 decrease in CL (CAS Latency).

But ultimately, you are looking at less than 1% performance difference. The quality of the RAM is really only applicable if you are going to over clock.

If anything, I would recommend getting at least 16 GB of RAM. I don't feel 8 is sufficient now that games are released to run in x64.

 

[tiny voices]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

Well, I wasn't so updated about what CAS latency when I bought my memory and I am willing to buy new if it is worth it for what I will gain from it

The higher the frequency, the higher the CAS timings. Basically, you don't need new RAM if you want to lower the frequency and the CL. You can just manually change it in the BIOS.

A higher frequency RAM (like yours, at 2400) will give better performance when compared to a lower frequency RAM with a lower CL. The CL11 timing only applies to the 2400 speed. I would go into your BIOS and confirm what timings and frequency your RAM is actually running at. You can lower the frequency and decrease the RAM timings if you want to compare, but I found ultimately that higher frequency performed better than a 1 or 2 decrease in CL (CAS Latency).

But ultimately, you are looking at less than 1% performance difference. The quality of the RAM is really only applicable if you are going to over clock.

If anything, I would recommend getting at least 16 GB of RAM. I don't feel 8 is sufficient now that games are released to run in x64.

8gb is plenty for modern gaming and for gaming to come. Even 64-bit games barely use more than 4gb RAM and most games are still 32-bit. Get 8gb now and then in a year or two add 8gb more if games advance to that point by then.

 

[Alec Mowat]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

Well, I wasn't so updated about what CAS latency when I bought my memory and I am willing to buy new if it is worth it for what I will gain from it

The higher the frequency, the higher the CAS timings. Basically, you don't need new RAM if you want to lower the frequency and the CL. You can just manually change it in the BIOS.

A higher frequency RAM (like yours, at 2400) will give better performance when compared to a lower frequency RAM with a lower CL. The CL11 timing only applies to the 2400 speed. I would go into your BIOS and confirm what timings and frequency your RAM is actually running at. You can lower the frequency and decrease the RAM timings if you want to compare, but I found ultimately that higher frequency performed better than a 1 or 2 decrease in CL (CAS Latency).

But ultimately, you are looking at less than 1% performance difference. The quality of the RAM is really only applicable if you are going to over clock.

If anything, I would recommend getting at least 16 GB of RAM. I don't feel 8 is sufficient now that games are released to run in x64.

8gb is plenty for modern gaming and for gaming to come. Even 64-bit games barely use more than 4gb RAM and most games are still 32-bit. Get 8gb now and then in a year or two add 8gb more if games advance to that point by then.

I disagree on this on. May as well get the 16 now, you won't regret it in the short term.

 

[tiny voices]

CAS timings can affect FPS when comparing RAM, but higher bus speed (2400 vs 1600) will give the overall better performance. When comparing two similar types, that's when CAS matters,

Well, I wasn't so updated about what CAS latency when I bought my memory and I am willing to buy new if it is worth it for what I will gain from it

The higher the frequency, the higher the CAS timings. Basically, you don't need new RAM if you want to lower the frequency and the CL. You can just manually change it in the BIOS.

A higher frequency RAM (like yours, at 2400) will give better performance when compared to a lower frequency RAM with a lower CL. The CL11 timing only applies to the 2400 speed. I would go into your BIOS and confirm what timings and frequency your RAM is actually running at. You can lower the frequency and decrease the RAM timings if you want to compare, but I found ultimately that higher frequency performed better than a 1 or 2 decrease in CL (CAS Latency).

But ultimately, you are looking at less than 1% performance difference. The quality of the RAM is really only applicable if you are going to over clock.

If anything, I would recommend getting at least 16 GB of RAM. I don't feel 8 is sufficient now that games are released to run in x64.

8gb is plenty for modern gaming and for gaming to come. Even 64-bit games barely use more than 4gb RAM and most games are still 32-bit. Get 8gb now and then in a year or two add 8gb more if games advance to that point by then.

I disagree on this on. May as well get the 16 now, you won't regret it in the short term.

Buying 16gb RAM is a waste when nothing needs it. In a year, you can add more if you REALLY think you need it, but you won't just for gaming. IF you have the money to blow, sure get 16gb, but it is BY NO MEANS necessary at all, and it won't even do anything for the time being.

 

[Alec Mowat]

I'd get 16GB over faster RAM. A year from now, RAM may increase it price. You also need pair it with the exact same make and model, which could be difficult. If you do decide you want 16GB, you may end up spending the same money or more.

I'd just buy the 16GB now and be done with it. If you are running anything other than games, or have Origin, Steam, Firefox, a half finished video and are playing BF4 at the same time, 16 GB wont hurt.
If you plan on running some Virtual Machines for fun, 16 won't hurt.

There's no situation where 8GB is worse than 16GB. You could even create an 8GB RAM disk with the extra, if you're up to it.

"64K is more than enough for anyone"
- Bill Gates

 

[tiny voices]

RAM prices only go up when there is a natural disaster or political conflict that affects the country(s) it is made in. This is not possible to predict. 8gb RAM used to be $40 late in 2012 but then the floods happened and boom $70 for 8gb. This is nothing that can be predicted.

Also the RAM only working with the sme exact RAM is not really true. It is just something people say to cover their butt JUST IN CASE something doesn't work perfectly. I have been using 3 different sticks of RAM for the last 8 months together. They are different sizes, manufacturers, speeds, and timings. They however all work perfectly with each other, and easily were clocked to 1600mhz CL9 timings, and even run in dual channel.

Even if you don't set the speeds and timing yourself, they will automatically default to the slowest of the multiple sticks and all run at that speed. RAM compatibility issues are really a thing of the past. The ONLY thing to really care about is the voltage. All my sticks are 1.5v

 

[Alec Mowat]

RAM prices only go up when there is a natural disaster or political conflict that affects the country(s) it is made in. This is not possible to predict. 8gb RAM used to be $40 late in 2012 but then the floods happened and boom $70 for 8gb. This is nothing that can be predicted.

Also the RAM only working with the sme exact RAM is not really true. It is just something people say to cover their butt JUST IN CASE something doesn't work perfectly. I have been using 3 different sticks of RAM for the last 8 months together. They are different sizes, manufacturers, speeds, and timings. They however all work perfectly with each other, and easily were clocked to 1600mhz CL9 timings, and even run in dual channel.

Even if you don't set the speeds and timing yourself, they will automatically default to the slowest of the multiple sticks and all run at that speed. RAM compatibility issues are really a thing of the past. The ONLY thing to really care about is the voltage. All my sticks are 1.5v

For DDR3, it is true. That's why they release the RAM in kits now, instead of individual sticks.

So, no. Don't build a franken computer. It will impact stability, especially if you want to overclock.

This advice is terrible. Don't take shortcuts when you are building a high performance system. That's like buying two different disks for your cars breaks, or mismatching spark plugs.

 

[tiny voices]

RAM prices only go up when there is a natural disaster or political conflict that affects the country(s) it is made in. This is not possible to predict. 8gb RAM used to be $40 late in 2012 but then the floods happened and boom $70 for 8gb. This is nothing that can be predicted.

Also the RAM only working with the sme exact RAM is not really true. It is just something people say to cover their butt JUST IN CASE something doesn't work perfectly. I have been using 3 different sticks of RAM for the last 8 months together. They are different sizes, manufacturers, speeds, and timings. They however all work perfectly with each other, and easily were clocked to 1600mhz CL9 timings, and even run in dual channel.

Even if you don't set the speeds and timing yourself, they will automatically default to the slowest of the multiple sticks and all run at that speed. RAM compatibility issues are really a thing of the past. The ONLY thing to really care about is the voltage. All my sticks are 1.5v

For DDR3, it is true. That's why they release the RAM in kits now, instead of individual sticks.

So, no. Don't build a franken computer.

Pretty much any DDR3 RAM of the same voltage is compatible. The BIOS will even sort out the speed and timings for you. RAM 'incompatibilities' are few and far between.

 

[Alec Mowat]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

 

[tiny voices]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

Actually using Passmark's memory test, I average BETTER OR THE SAME results than the 2x4gb sets they have benchmarked at the same speed and CAS rating.

Of course it is not the best option to buy 3 different sticks and go for it, I never said that. BUT, what I am saying is that IF you were in the situation where you had different sticks, chances are they will be perfectly fine though.

I never told anyone to just buy non-matching RAM for the hell of it. Simply explaining that generally mismatched RAM is not a problem anymore as it used to be with DDR2 for example.

 

[Alec Mowat]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

Actually using Passmark's memory test, I average BETTER OR THE SAME results than the 2x4gb sets they have benchmarked at the same speed and CAS rating.

Of course it is not the best option to buy 3 different sticks and go for it, I never said that. BUT, what I am saying is that IF you were in the situation where you had different sticks, chances are they will be perfectly fine though.

I never told anyone to just buy non-matching RAM for the hell of it. Simply explaining that generally mismatched RAM is not a problem anymore as it used to be with DDR2 for example.

No, higher speed ram is more of a problem mis matching now than it was with ddr2.
You can mismatch DDR and sd ram. Ddr2 wasn't recommended, ddr3 will cause corruption and it's difficult to find similar pairs. They are usually rated by their test results after manufacturing, they aren't always built to specifics.

Buy one ram kit and stick with it. If you wan to change ram, sell the current kit and buy a new one. Don't mis match. It's one of those little quirks that will cause bsod down the road.

Regardless; are you saying now that adding more ram improved your pass mark score?

 

[tiny voices]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

Actually using Passmark's memory test, I average BETTER OR THE SAME results than the 2x4gb sets they have benchmarked at the same speed and CAS rating.

Of course it is not the best option to buy 3 different sticks and go for it, I never said that. BUT, what I am saying is that IF you were in the situation where you had different sticks, chances are they will be perfectly fine though.

I never told anyone to just buy non-matching RAM for the hell of it. Simply explaining that generally mismatched RAM is not a problem anymore as it used to be with DDR2 for example.

No, higher speed ram is more of a problem mis matching now than it was with ddr2.
You can mismatch DDR and sd ram. Ddr2 wasn't recommended, ddr3 will cause corruption and it's difficult to find similar pairs. They are usually rated by their test results after manufacturing, they aren't always built to specifics.

Buy one ram kit and stick with it. If you wan to change ram, sell the current kit and buy a new one. Don't mis match. It's one of those little quirks that will cause bsod down the road.

Regardless; are you saying now that adding more ram improved your pass mark score?

No, this is not what I am saying at all. I am saying that my mismatched 8gb setup scores about 1-5% higher than the 2x4gb kits they have benchmarked. My kit consists of 3 sticks running in XMP dual channel mode. The point isn't that my setup is slightly better. It is that I have lost NOTHING in performance from this setup.

One Crucial 2gb stick
One PNY 2gb stick
One Dell OEM 4gb stick
All timings set to CL9 and speed set to 1600mhz.

I'm not saying to go and mismatch RAM, all I am saying is that it is way less of a big deal than you make it seem. I have had this setup for 6 months, and before that I had 2x2gb with the PNY and Crucial mismatched sticks for a year. NEVER had a single BSOD. I have run them through memtest for 10 passes multiple times. ZERO errors.

 

[Alec Mowat]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

Actually using Passmark's memory test, I average BETTER OR THE SAME results than the 2x4gb sets they have benchmarked at the same speed and CAS rating.

Of course it is not the best option to buy 3 different sticks and go for it, I never said that. BUT, what I am saying is that IF you were in the situation where you had different sticks, chances are they will be perfectly fine though.

I never told anyone to just buy non-matching RAM for the hell of it. Simply explaining that generally mismatched RAM is not a problem anymore as it used to be with DDR2 for example.

No, higher speed ram is more of a problem mis matching now than it was with ddr2.
You can mismatch DDR and sd ram. Ddr2 wasn't recommended, ddr3 will cause corruption and it's difficult to find similar pairs. They are usually rated by their test results after manufacturing, they aren't always built to specifics.

Buy one ram kit and stick with it. If you wan to change ram, sell the current kit and buy a new one. Don't mis match. It's one of those little quirks that will cause bsod down the road.

Regardless; are you saying now that adding more ram improved your pass mark score?

No, this is not what I am saying at all. I am saying that my mismatched 8gb setup scores about 1-5% higher than the 2x4gb kits they have benchmarked. My kit consists of 3 sticks running in XMP dual channel mode. The point isn't that my setup is slightly better. It is that I have lost NOTHING in performance from this setup.

One Crucial 2gb stick
One PNY 2gb stick
One Dell OEM 4gb stick
All timings set to CL9 and speed set to 1600mhz.

I'm not saying to go and mismatch RAM, all I am saying is that it is way less of a big deal than you make it seem. I have had this setup for 6 months, and before that I had 2x2gb with the PNY and Crucial mismatched sticks for a year. NEVER had a single BSOD. I have run them through memtest for 10 passes multiple times. ZERO errors.

You are using 3 sticks of RAM, of different makes and sizes on a dual-channel motherboard?

LOL

Get out of here.

 

[tiny voices]

Yes, but DDR3 is tested in a lab and paired up for a reason. It's very finicky when it's running at such a high speed. It might run fine, but you could very well be losing performance without knowing.

Don't mismatch DDR3. This advice is awful.

Actually using Passmark's memory test, I average BETTER OR THE SAME results than the 2x4gb sets they have benchmarked at the same speed and CAS rating.

Of course it is not the best option to buy 3 different sticks and go for it, I never said that. BUT, what I am saying is that IF you were in the situation where you had different sticks, chances are they will be perfectly fine though.

I never told anyone to just buy non-matching RAM for the hell of it. Simply explaining that generally mismatched RAM is not a problem anymore as it used to be with DDR2 for example.

No, higher speed ram is more of a problem mis matching now than it was with ddr2.
You can mismatch DDR and sd ram. Ddr2 wasn't recommended, ddr3 will cause corruption and it's difficult to find similar pairs. They are usually rated by their test results after manufacturing, they aren't always built to specifics.

Buy one ram kit and stick with it. If you wan to change ram, sell the current kit and buy a new one. Don't mis match. It's one of those little quirks that will cause bsod down the road.

Regardless; are you saying now that adding more ram improved your pass mark score?

No, this is not what I am saying at all. I am saying that my mismatched 8gb setup scores about 1-5% higher than the 2x4gb kits they have benchmarked. My kit consists of 3 sticks running in XMP dual channel mode. The point isn't that my setup is slightly better. It is that I have lost NOTHING in performance from this setup.

One Crucial 2gb stick
One PNY 2gb stick
One Dell OEM 4gb stick
All timings set to CL9 and speed set to 1600mhz.

I'm not saying to go and mismatch RAM, all I am saying is that it is way less of a big deal than you make it seem. I have had this setup for 6 months, and before that I had 2x2gb with the PNY and Crucial mismatched sticks for a year. NEVER had a single BSOD. I have run them through memtest for 10 passes multiple times. ZERO errors.

You are using 3 sticks of RAM, of different makes and sizes on a dual-channel motherboard?

LOL

Get out of here.

Excuse me? They run in DUAL CHANNEL thanks to intel XMP mode. They benchmark HIGHER than a normal 2x4gb kit. I'm sorry, you don't know what you are talking about.

 

[Alec Mowat]

You have 3 sticks of RAM, 2 x 2 and 1 x4

And it's running in dual channel?

Of course your benchmark is better, you doubled your RAM.

All of your advice is awful. Anyone reading this thread should disregard your comments. Even if it works, it's bad practice to begin with.

 

[tiny voices]

You have 3 sticks of RAM, 2 x 2 and 1 x4

And it's running in dual channel?

Of course your benchmark is better, you doubled your RAM.

All of your advice is awful. Anyone reading this thread should disregard your comments. Even if it works, it's bad practice to begin with.

Please clam down and listen for a minute. I'll assume you do not know what FLEX modes are for. They allow you to run non-traditional sets of RAM in dual channel. YES, my THREE sticks run in dual channel. CPUZ, my BIOS, PASSMARK all confirm this.

And I will also assume you have no knowledge of passmark. The memory test DOES NOT take into account total amount of RAM. ONLY speed and timings and calculations per UNIT. NOTHING to do with 4gb vs. 8gb. So NO, adding more RAM does NOT help your passmark score at all.

I NEVER told anyone to go out and run memory like I am. I simply stated that it IS possible to mix and match and have equally as good results as buying a 2x4gb kit. Is it recommended? NO, of course not, but it is possible. That is all I am saying.

You don't need to come in and beat me up over saying that something is possible and backing it up with proof. You can disagree with me, but I have shown proof that the way I run my RAM works perfectly fine with no ill effects or performance loss.

http://www.intel.com/support/motherboards/desktop/sb/cs-011965.htm#flex please read this. It explains how 3 sticks can be run in dual channel.

I never came in here to start a fight or anything. Simply explain some things. That's all. I'm sorry you took what I said as a personal attack. It was never intended to be that way at all. All I was doing was providing simple and accurate information.

 

[Alec Mowat]

Look, you can do whatever you want with your own system. Passmark will improve performance when increasing RAM from 4 to 8, simply because Windows uses 2 GB of RAM to operate correctly. Of course you got better Benchmarks.
That's why I suggest buying 16GB now.

If you are upgrading a work station or bringing an end of life system up to spec, sure. Just stick some RAM in, go in the BIOS and set the timings and voltages.

If you are spending $2200+ on a gaming rig, running a brand new i7 and brand new 2400 mhz RAM (which will be even more susceptible to read/write errors when not paired correctly), why on earth would you do anything that ~might~ cause integrity failures?

For a brand new build, recommending mixing RAM is just a bad idea. Just simply don't recommend it, or don't do it. Sure, it will work if you HAVe to

Just buy 16 GB of 2400 mhz RAM today, replace the videocard in 3 years, and replace the whole system 5 years. End of story.

Passmark is Passmark, I'm very impressed you can stick RAM in a slot and it works.

 

[tiny voices]

Look, you can do whatever you want with your own system.

If you are upgrading a work station or bringing an end of life system up to spec, sure. Just stick some RAM in, go in the BIOS and set the timings and voltages.

If you are spending $2200+ on a gaming rig, running a brand new i7 and brand new 2400 mhz RAM (which will be even more susceptible to read/write errors when not paired correctly), why on earth would you do anything that ~might~ cause integrity failures?

For a brand new build, recommending mixing RAM is just a bad idea. Just simply don't recommend it, or don't do it. Sure, it will work if you HAVe to

Just buy 16 GB of 2400 mhz RAM today, replace the videocard in 3 years, and replace the whole system 5 years. End of story.

Passmark is Passmark, I'm very impressed you can stick RAM in a slot and it works.

Can you please show me where I told someone to buy mixed RAM for a NEW computer?

 

[Alec Mowat]

Look, you can do whatever you want with your own system.

If you are upgrading a work station or bringing an end of life system up to spec, sure. Just stick some RAM in, go in the BIOS and set the timings and voltages.

If you are spending $2200+ on a gaming rig, running a brand new i7 and brand new 2400 mhz RAM (which will be even more susceptible to read/write errors when not paired correctly), why on earth would you do anything that ~might~ cause integrity failures?

For a brand new build, recommending mixing RAM is just a bad idea. Just simply don't recommend it, or don't do it. Sure, it will work if you HAVe to

Just buy 16 GB of 2400 mhz RAM today, replace the videocard in 3 years, and replace the whole system 5 years. End of story.

Passmark is Passmark, I'm very impressed you can stick RAM in a slot and it works.

Can you please show me where I told someone to buy mixed RAM for a NEW computer?

That's the question!!!
"Hey guys! I recently bought this system, though I wonder if my memory is good enough for the other parts and if you think I should swap them for other, better memories. My purpose with this system is gaming in 1080p."

I'm done here.