- Mar 30, 2010
- 87
- 0
- 18,640
- 1
Hi
I bought two separate G.Skill Trident Z Royal kits each at 4000mhz 17-17-17-37 and 1.35v (4x8gb) timings from NewEgg. Im wondering if you can tell me if my stress testing I’ve done is sufficient for gaming/steaming/moderate workloads or whether I should RMA them and wait for a 4 stick kit XMP to come available. I have about 10-12 days left to return to NewEgg so appreciate the quick response.
The four sticks have the same serial # except 87, 88, 89, 90 at the end – which leads me to think that they were all created in a row despite them being in separate kits. I would think that this would make them more compatible together since 4 stick kits are usually 4 sticks in succession (albeit XMP tested together as well, which isnt the case here since they are 2 seperate kits).
I stress tested both kits individual (2 sticks at a time) – both handled fine on XMP 1 and 2 under Prime Large FFT for 1 hour.
My motherboard is a ASUS Strix z490 Gaming G – WiFi (its on the QVL for the 4 stick pack and 2 stick packs of 4000mhz CL17). My PC has a i9-10900K at 5.0ghz (oc verified by Silicon Lottery, their OC is below for reference) where I bought it. My system is under custom water with 65-68c temps under full load.
Silicon Lottery binned the following OC on my CPU:
• Ai Overclock Tuner: XMP II (i changed it XMP 1 for this situation)
• AVX Instruction Core Ratio Negative Offset: 1
• CPU Core Ratio: By Core Usage
o Turbo Ratio Limit 0: 52
o Turbo Ratio Cores 0: 3
o Turbo Ratio Limit 1: 51
o Turbo Ratio Cores 1: 6
o Turbo Ratio Limit 2: 50
o Turbo Ratio Cores 2: 10
• Digi+ VRM
o CPU Load-line Calibration: Level 3 (they told me to do LLC 4 based on my mobo weaker vram cooling)
• Internal CPU Power Management
o Long Duration Package Power Limit: 230
o Short Duration Package Power Limit: 230
• CPU Core/Cache Voltage: Manual Mode
• CPU Core Voltage Override: 1.350
My mobo manual states that of the 4 dimm slots, you want to start with slots 2 and 4.
So from left to right I started with these sticks (ending 2 digits from serials) 89 – 87 – 90 – 88. As such, the kits were together in their respective dual channels
This config posted to windows, but failed to load Modern Warfare Warzone. XMP2 failed (understandably since that was rated for 2 sticks). XMP1 was the one that posted. I got a lot of BSODs though and seemed unstable
So then I switched the sticks around and went 87 – 89 – 88 – 90 and XMP1 (just timings rest auto in bios). So 89 and 90 were in slots 2 and 4 now.
This posted to windows, ran MW Warzone and GTA for 3 hours of gaming while streaming with Streamlabs OBS and no crashes. Further it handle a typical workload I do over night running between 50-70% cpu and 50% ram utilization for 12hrs with no crashes or BSODs.
However, it only ran stable in Prime Large FFT for 45 mins before threads started stopping. By the 1 hr mark I had 8 of 20 threads stop. But prime was still running on the remaining 12 threads
So my question is – are my real world uses of the system (gaming/stream, video rending, or other workload processing tasks) enough of a benchmark to let the RMA window run out and stick with these ram? Or should Prime really run stable at XMP indefinitely and should I RMA them, buy some cheap ram on Amazon that I can return later, and wait for a 4 pack of the 4000mhz cl17 come in stock?
Cost is not an issue, and i want 4 dimms filled of the royal since it goes with my builds looks. In case you want to recommend 2x16 ram im not interested.
Appreciate you thoughts and reply!
I bought two separate G.Skill Trident Z Royal kits each at 4000mhz 17-17-17-37 and 1.35v (4x8gb) timings from NewEgg. Im wondering if you can tell me if my stress testing I’ve done is sufficient for gaming/steaming/moderate workloads or whether I should RMA them and wait for a 4 stick kit XMP to come available. I have about 10-12 days left to return to NewEgg so appreciate the quick response.
The four sticks have the same serial # except 87, 88, 89, 90 at the end – which leads me to think that they were all created in a row despite them being in separate kits. I would think that this would make them more compatible together since 4 stick kits are usually 4 sticks in succession (albeit XMP tested together as well, which isnt the case here since they are 2 seperate kits).
I stress tested both kits individual (2 sticks at a time) – both handled fine on XMP 1 and 2 under Prime Large FFT for 1 hour.
My motherboard is a ASUS Strix z490 Gaming G – WiFi (its on the QVL for the 4 stick pack and 2 stick packs of 4000mhz CL17). My PC has a i9-10900K at 5.0ghz (oc verified by Silicon Lottery, their OC is below for reference) where I bought it. My system is under custom water with 65-68c temps under full load.
Silicon Lottery binned the following OC on my CPU:
• Ai Overclock Tuner: XMP II (i changed it XMP 1 for this situation)
• AVX Instruction Core Ratio Negative Offset: 1
• CPU Core Ratio: By Core Usage
o Turbo Ratio Limit 0: 52
o Turbo Ratio Cores 0: 3
o Turbo Ratio Limit 1: 51
o Turbo Ratio Cores 1: 6
o Turbo Ratio Limit 2: 50
o Turbo Ratio Cores 2: 10
• Digi+ VRM
o CPU Load-line Calibration: Level 3 (they told me to do LLC 4 based on my mobo weaker vram cooling)
• Internal CPU Power Management
o Long Duration Package Power Limit: 230
o Short Duration Package Power Limit: 230
• CPU Core/Cache Voltage: Manual Mode
• CPU Core Voltage Override: 1.350
My mobo manual states that of the 4 dimm slots, you want to start with slots 2 and 4.
So from left to right I started with these sticks (ending 2 digits from serials) 89 – 87 – 90 – 88. As such, the kits were together in their respective dual channels
This config posted to windows, but failed to load Modern Warfare Warzone. XMP2 failed (understandably since that was rated for 2 sticks). XMP1 was the one that posted. I got a lot of BSODs though and seemed unstable
So then I switched the sticks around and went 87 – 89 – 88 – 90 and XMP1 (just timings rest auto in bios). So 89 and 90 were in slots 2 and 4 now.
This posted to windows, ran MW Warzone and GTA for 3 hours of gaming while streaming with Streamlabs OBS and no crashes. Further it handle a typical workload I do over night running between 50-70% cpu and 50% ram utilization for 12hrs with no crashes or BSODs.
However, it only ran stable in Prime Large FFT for 45 mins before threads started stopping. By the 1 hr mark I had 8 of 20 threads stop. But prime was still running on the remaining 12 threads
So my question is – are my real world uses of the system (gaming/stream, video rending, or other workload processing tasks) enough of a benchmark to let the RMA window run out and stick with these ram? Or should Prime really run stable at XMP indefinitely and should I RMA them, buy some cheap ram on Amazon that I can return later, and wait for a 4 pack of the 4000mhz cl17 come in stock?
Cost is not an issue, and i want 4 dimms filled of the royal since it goes with my builds looks. In case you want to recommend 2x16 ram im not interested.
Appreciate you thoughts and reply!
Facebook
Twitter
Instagram