Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

Which ram is best for Ryzen 5900x?

Hello to everyone.

I use two of the Gskill TridentZ 2x16GB CL16 (F4-3200C16D-32GTZR) memory kit in my system.

So I have a total of 4x16gb Ram. First kit is Samsung B die, Second kit is Hynix A Die.

I've been reading comments on the internet that CL14 memory kits perform better.

They also talk about the DRAM ratio value. I think they say 1:1:1 is the best value.

 

Would changing my rams help me? Is it worth it? Or which memory kit might be better to switch to?

 

My other existing system components are as follows;
AMD Ryzen 9 5900x
Asus Rog Crosshair VIII
ZOTAC Rtx2060 mini
2 x XPG SX8200Pro 1TB Nvme SSDs
Be Quiet Straightpower 11 850w Gold PSU
Be Quiet SilentBase 802 White Case

 

I would be glad if you give me an idea. 

  
 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites

If you already have Samsung B-die just OC it to the speeds you want.

 

Getting 3600MT/s CL14 shouldn't be much of an issue with the B-die.

But if you're trying to use both kits at the same time this may be a very tricky thing to do.

 

Anyways, yes... CL14 is better than CL16 but there is more than that to the DRAM performance. For example, if you have Dual Rank sticks then they will still outperform your Singe Rank sticks at the same frequency even with a bit worse CL.

But running 4x16GB of Dual Rank sticks put a lot of strain on the memory controller so typically you want to avoid that if you are pushing for high frequency and low CL and just stick to two Dual Rank sticks in Dual Channel so (2x32GB if you need 64GB).

 

All that said, what you have now is fine, you don't need to buy new RAM. You can always manually OC and if you get newer faster RAM you may still need to manually tweak timings as XMP on higher speeds may not work without issues.

 

If you don't need 64GB then just use the 2x16GB of the Samsung B-die and OC it.

If you need 64GB then you probably won't be able to get anything much faster unless you are willing to OC (which you already can) or pay a huge price premium for a fast 64GB kit.

Link to post
Share on other sites
13 minutes ago, WereCat said:

If you already have Samsung B-die just OC it to the speeds you want.

 

Getting 3600MT/s CL14 shouldn't be much of an issue with the B-die.

But if you're trying to use both kits at the same time this may be a very tricky thing to do.

 

Anyways, yes... CL14 is better than CL16 but there is more than that to the DRAM performance. For example, if you have Dual Rank sticks then they will still outperform your Singe Rank sticks at the same frequency even with a bit worse CL.

But running 4x16GB of Dual Rank sticks put a lot of strain on the memory controller so typically you want to avoid that if you are pushing for high frequency and low CL and just stick to two Dual Rank sticks in Dual Channel so (2x32GB if you need 64GB).

 

All that said, what you have now is fine, you don't need to buy new RAM. You can always manually OC and if you get newer faster RAM you may still need to manually tweak timings as XMP on higher speeds may not work without issues.

 

If you don't need 64GB then just use the 2x16GB of the Samsung B-die and OC it.

If you need 64GB then you probably won't be able to get anything much faster unless you are willing to OC (which you already can) or pay a huge price premium for a fast 64GB kit.

Unfortunately, I cannot remove the other kit as I need 64GB of ram.

But I want to ask,
If I want to adjust these rams, I will have to adjust the settings according to the Hynix chip, right?

Because as far as I know, while Samsung can reach higher speed, that is, it is more flexible, Hynix A will offer more limited adjustment possibilities.
Am I thinking right?

Although the two ram kits are different molds, they belong to the same ram series.
They are referred to as F4-3200C16D-32GTZR. They have the same timings. They have CL16-18-18-38 speeds. And the DOCP profile is enabled and works fine.
We think Gskill can use different chips in the same model ram while producing.

Wouldn't using 4 dimm ram add a bit more to the performance?
Isn't 4 Dimm useful if we are not based on speed and CL value? 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
3 hours ago, Orodruin said:

Unfortunately, I cannot remove the other kit as I need 64GB of ram.

But I want to ask,
If I want to adjust these rams, I will have to adjust the settings according to the Hynix chip, right?

Because as far as I know, while Samsung can reach higher speed, that is, it is more flexible, Hynix A will offer more limited adjustment possibilities.
Am I thinking right?

Although the two ram kits are different molds, they belong to the same ram series.
They are referred to as F4-3200C16D-32GTZR. They have the same timings. They have CL16-18-18-38 speeds. And the DOCP profile is enabled and works fine.
We think Gskill can use different chips in the same model ram while producing.

Wouldn't using 4 dimm ram add a bit more to the performance?
Isn't 4 Dimm useful if we are not based on speed and CL value? 

The optimal number of sticks for dual channel is always two. The IMC has to work harder to run four sticks, which means less OC potential/looser timings. However, there's some benefit to running 4 ranks, which can be achieved with just two sticks, if they're dual rank, or four single ranked sticks.

 

Lower CL is always better, but there's very much diminishing returns in the cost to performance. 3600MHz CL14 doesn't provide enough of a performance benefit to warrant the extreme cost over 3600MHz CL16.

 

For what it's worth, I've got a kit of 2x16GB dual rank 3600MHz CL16 RAM that I OCed to just 3800MHz, mostly for the FCLK boost. With that, my 5900X is benching higher than any other I've seen, with like 23500 R23 MC and 9100 R20 MC, for example. Memory is part of the equation, but there's other factors as well, and you don't have to go insane on the memory to still get awesome performance.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites
1 hour ago, Chris Pratt said:

The optimal number of sticks for dual channel is always two. The IMC has to work harder to run four sticks, which means less OC potential/looser timings. However, there's some benefit to running 4 ranks, which can be achieved with just two sticks, if they're dual rank, or four single ranked sticks.

 

Lower CL is always better, but there's very much diminishing returns in the cost to performance. 3600MHz CL14 doesn't provide enough of a performance benefit to warrant the extreme cost over 3600MHz CL16.

 

For what it's worth, I've got a kit of 2x16GB dual rank 3600MHz CL16 RAM that I OCed to just 3800MHz, mostly for the FCLK boost. With that, my 5900X is benching higher than any other I've seen, with like 23500 R23 MC and 9100 R20 MC, for example. Memory is part of the equation, but there's other factors as well, and you don't have to go insane on the memory to still get awesome performance.

In other words, there is no positive cost in trying to switch to CL14 ram.

 

Then it means I will continue to use my current ram.
They work at their own standard values with CL 16-18-18-38 Timing.

 

So if I want to use 4000Mhz and above ram on Ryzen processors, will it be a problem?

Or what advantages or disadvantages does it have? 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
37 minutes ago, Orodruin said:

In other words, there is no positive cost in trying to switch to CL14 ram.

 

Then it means I will continue to use my current ram.
They work at their own standard values with CL 16-18-18-38 Timing.

 

So if I want to use 4000Mhz and above ram on Ryzen processors, will it be a problem?

Or what advantages or disadvantages does it have? 

At the same speeds c14 will always better than c16. There is a tangible difference, whether or not it’s worth it is up to you.. but the difference is there.

 

As for speed, you can buy whatever you want, it is entirely up to the cpu if it’s gonna run it that fast or not. Consider yourself to be very lucky if you can run 2000 1:1 whea free.

 

4 single rank sticks perform better than 2 single rank sticks, oc potential doesn’t really matter because you have to really lean on a single rank setup to get it to perform like a dual rank setup. For realsies.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
27 minutes ago, freeagent said:

At the same speeds c14 will always better than c16. There is a tangible difference, whether or not it’s worth it is up to you.. but the difference is there.

 

As for speed, you can buy whatever you want, it is entirely up to the cpu if it’s gonna run it that fast or not. Consider yourself to be very lucky if you can run 2000 1:1 whea free.

 

4 single rank sticks perform better than 2 single rank sticks, oc potential doesn’t really matter because you have to really lean on a single rank setup to get it to perform like a dual rank setup. For realsies.

So how do I know if the ram is even or odd rank without buying ram?
Gskill has kits with speeds of 3600-4000-4266-4400-4600 mhz. Would it be better to be promoted to one of these?
Does it make sense to switch to 4600 or 4800 mhz ram? 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
3 minutes ago, Orodruin said:

So how do I know if the ram is even or odd rank without buying ram?

Usually 16GB sticks with tight timings are dual rank B-Die. I don't know about others because I only know about my system.

 

5 minutes ago, Orodruin said:

Gskill has kits with speeds of 3600-4000-4266-4400-4600 mhz. Would it be better to be promoted to one of these?

That's up to you.. my 3200C14 kit is stable at 4200 C15, 4400 C16. Not saying yours will do that, but I'm not saying it wont either. To be fair I cant run more than 1900 with 4 sticks with this CPU and be stable. I also cant run 4 sticks faster than 1833 on my 5600X and be stable. But with 2 sticks my 5600X is stable at 2000 1:1 and can bench at 2100 1:1

 

10 minutes ago, Orodruin said:

Does it make sense to switch to 4600 or 4800 mhz ram?

For running with AMD? You could.. but I wouldn't. No point at all. Unless you can run them slower and clamp down on the timings.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
1 hour ago, Orodruin said:

In other words, there is no positive cost in trying to switch to CL14 ram.

 

Then it means I will continue to use my current ram.
They work at their own standard values with CL 16-18-18-38 Timing.

 

So if I want to use 4000Mhz and above ram on Ryzen processors, will it be a problem?

Or what advantages or disadvantages does it have? 

CL14 will be better, but in most workloads you won't be able to tell a difference. The goal should be to get the best ratio of cost to performance, not just pure raw performance, especially when you start delving into the area of micro-optimization like having RAM that is just 1ns less total latency.

 

4000MHz equally will be an improvement, but also likely won't be noticeable in most workloads. However, there's an additional complication there in that you need to be able to OC the FLCK to 2000MHz for it to even be worthwhile. If you can't get the FCLK that high, you'll either have to drop the clocks anyways, or run async, which will actually harm your performance, and likely cancel out any performance increase from the faster RAM anyways.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites
11 minutes ago, freeagent said:

Usually 16GB sticks with tight timings are dual rank B-Die. I don't know about others because I only know about my system.

 

That's up to you.. my 3200C14 kit is stable at 4200 C15, 4400 C16. Not saying yours will do that, but I'm not saying it wont either. To be fair I cant run more than 1900 with 4 sticks with this CPU and be stable. I also cant run 4 sticks faster than 1833 on my 5600X and be stable. But with 2 sticks my 5600X is stable at 2000 1:1 and can bench at 2100 1:1

 

For running with AMD? You could.. but I wouldn't. No point at all. Unless you can run them slower and clamp down on the timings.

 

I understand. Then I think it will be logical to buy 3600Mhz CL14 ram for me. AMD recommends 3733Mhz, but it's hard to find memory at that speed. Since I cannot use the kits I have together as CL14, I think I will buy 3600 Mhz CL14.

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
13 minutes ago, Chris Pratt said:

CL14 will be better, but in most workloads you won't be able to tell a difference. The goal should be to get the best ratio of cost to performance, not just pure raw performance, especially when you start delving into the area of micro-optimization like having RAM that is just 1ns less total latency.

 

4000MHz equally will be an improvement, but also likely won't be noticeable in most workloads. However, there's an additional complication there in that you need to be able to OC the FLCK to 2000MHz for it to even be worthwhile. If you can't get the FCLK that high, you'll either have to drop the clocks anyways, or run async, which will actually harm your performance, and likely cancel out any performance increase from the faster RAM anyways.

I understand. I think that upgrading the ram will hurt me in terms of all kinds of costs.

It seems logical to buy 3600MHZ CL14 ram for now. Would you recommend? As far as I remember, AMD recommended 3600Mhz rams. 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
4 minutes ago, Orodruin said:

I understand. I think that upgrading the ram will hurt me in terms of all kinds of costs.

It seems logical to buy 3600MHZ CL14 ram for now. Would you recommend? As far as I remember, AMD recommended 3600Mhz rams. 

As I said, I would recommend 3600MHz CL16. It has the best cost to performance ratio. If you can OC it to 3800MHz or higher and/or tighten the timings to CL14, great, but 3600MHz CL16 gives you the best starting point and the best overall performance out of the box without spending a fortune.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites

It might have the best cost to performance ratio, but its not the best performance. There is more than a 1ns difference, believe me.

 

If you buy C14 sticks, you wont have to worry if your C16 sticks can do it or not. Save some cash there by not having to buy another set when you do get into tuning.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
3 hours ago, Chris Pratt said:

With that, my 5900X is benching higher than any other I've seen, with like 23500 R23

I get that score within the stock clock range with my ram at 3200C14.

 

stock.thumb.PNG.4400c811be8e0203e6293e45d3238cb5.PNG

 

 

Edit:

 

And your 1900FCLK, still in stock clock range.

 

stock2.thumb.PNG.a001526903bf762224de28a7dab0aeaa.PNG

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
8 minutes ago, freeagent said:

I get that score within the stock clock range with my ram at 3200C14.

 

stock.thumb.PNG.4400c811be8e0203e6293e45d3238cb5.PNG

 

The point was that you don't need super fast RAM and/or very low latency to get top end scores. They help, of course, but there's no point in spending lots of extra money when you can get there just as easily without spending the lots of extra money.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites

I totally agree with you man. But.. if you do want to get into tuning, your going to want the best you can get right? To some people money doesn't matter too much, unfortunately, I am poor, so money does matter to me lol. But I don't make money with my computer, so its just a toy. And since I only do complete system upgrades bi-generationally, I end up using my computers longer than most enthusiasts.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
3 hours ago, Chris Pratt said:

As I said, I would recommend 3600MHz CL16. It has the best cost to performance ratio. If you can OC it to 3800MHz or higher and/or tighten the timings to CL14, great, but 3600MHz CL16 gives you the best starting point and the best overall performance out of the box without spending a fortune.

What do you say about 3800 mhz cl18 ram?  Does it make sense to get 3800 instead of 3600?

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
3 hours ago, freeagent said:

It might have the best cost to performance ratio, but its not the best performance. There is more than a 1ns difference, believe me.

 

If you buy C14 sticks, you wont have to worry if your C16 sticks can do it or not. Save some cash there by not having to buy another set when you do get into tuning.

Cl14 is better under all circumstances then do I understand correctly?

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
2 hours ago, Orodruin said:

Cl14 is better under all circumstances then do I understand correctly?

Yes, definitely.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
2 hours ago, Orodruin said:

What do you say about 3800 mhz cl18 ram?  Does it make sense to get 3800 instead of 3600?

Not really, because you can honestly probably get a 3600MHz kit at least to 3800MHz, anyways, and the looser timings are worse.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites
2 hours ago, Orodruin said:

Cl14 is better under all circumstances then do I understand correctly?

Lower CAS latency is always better. The question is whether that amount of "better" makes sense due to factors such as cost. That's true of pretty much anything. Unless you have just unlimited funds, and don't mind wasting a ton of money on getting an extra FPS here or an extra 0.1% perf boost there, then the eye should always be on best performance per dollar.

CPU: AMD Ryzen 9 5900X · Cooler: Artic Liquid Freezer II 280 · Motherboard: MSI MEG X570 Unify · RAM: G.skill Ripjaws V 2x16GB 3600MHz CL16 (2Rx8) · Graphics Card: ASUS GeForce RTX 3060 Ti TUF Gaming · Boot Drive: 500GB WD Black SN750 M.2 NVMe SSD · Game Drive: 2TB Crucial MX500 SATA SSD · PSU: Corsair White RM850x 850W 80+ Gold · Case: Corsair 4000D Airflow · Monitor: MSI Optix MAG342CQR 34” UWQHD 3440x1440 144Hz · Keyboard: Corsair K100 RGB Optical-Mechanical Gaming Keyboard (OPX Switch) · Mouse: Corsair Ironclaw RGB Wireless Gaming Mouse

Link to post
Share on other sites

I think your percentage gains may be under rated a little, but value wise I do agree. But if you do like to play around with benchmarks tighter and faster is always better. But if you are just an average joe type of user it will probably make little difference.

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites
11 hours ago, Chris Pratt said:

Lower CAS latency is always better. The question is whether that amount of "better" makes sense due to factors such as cost. That's true of pretty much anything. Unless you have just unlimited funds, and don't mind wasting a ton of money on getting an extra FPS here or an extra 0.1% perf boost there, then the eye should always be on best performance per dollar.

I bought TridentZ RGB 4000Mhz CL18 ram with Gskill F4-4000C18D-32GTZRB model code as two boxes.
It seems compatible with my motherboard. I think I won't have any problems. This ram is Single Rank. I think it has a Samsung chip.

If I have problems at 4000Mhz I will manually reduce the speed to 3733 or 3600.

I was going to get 3600Mhz, but I chose this because my motherboard didn't have any problems with 4000Mhz ram. 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites
9 hours ago, freeagent said:

I think your percentage gains may be under rated a little, but value wise I do agree. But if you do like to play around with benchmarks tighter and faster is always better. But if you are just an average joe type of user it will probably make little difference.

I bought Gskill F4-4000C18D-32GTZRB model ram. My goal was to buy the most optimal ram. Frankly, I don't like to play with timings or settings. This ram seems to be compatible with my motherboard and equivalent models of my motherboard. It has CL18 delay value. I think it will work fine. If there is a problem at 4000Mhz, I will manually set the speed to 3733 or 3600.

At first I thought of buying a 3600Mhz TridentZ, but I chose this ram because I knew that 4000Mhz was the most stable and highest speed I could get. As far as I know, delay values increase at higher speeds due to DRAM ratio and Infinity fabric ratio. 

AMD Ryzen 9 5900X | Asus ROG CROSSHAIR VIII HERO | Asus ROG Strix LC360 RGB | G.SKILL TridentZ RGB 4x16GB 3200Mhz | EVGA GeForce RTX 3070 Ti FTW3 ULTRA | Asus VG279QM 280Hz IPS | Asus ROG Strix SCOPE | Asus ROG GLADIUS II CORE | A-DATA XPG SX8200PRO 1TB | Seagate Barracuda 8TB | BeQuiet! STRAIGHTPOWER 11 850Watt 80+ Gold | BeQuiet! SILENTBASE 802 Window White Case

Link to post
Share on other sites

My 5900X throws a ton of whea’s at 2000 1:1. Good luck!

AMD R9 5900X | Thermalright Frost Commander 140

Asus Strix B550-F Gaming | 4x8GB G.Skill Trident Z 3200C14

EVGA RTX 3070 Ti FTW3 Ultra | WD SN750 500GB, 1TB

Prime GX-750 | Fractal Design Meshify C | Many CFM's

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×