EVGA

Considering 4000CL19 B-die RAM vs. 3200CL14 B-die

Author
Post
Essentials Only Full Version
kevinc313
CLASSIFIED ULTRA Member
  • Total Posts : 5004
  • Reward points : 0
  • Joined: 2019/02/28 09:27:55
  • Status: offline
  • Ribbons : 22
2019/12/04 09:08:23 (permalink)
Currently have 2x 32gb kits of G.Skill 3000CL16 ram on a MSI Z390 Gaming Edge AC MB with a 9900K.  Have been able to clock them to 2666CL15 1T with 1.15/1.2v IO/SA.  Not really stable at 3000mhz with my basic memory OC skills, the kit is not on the QVL for the MB and it is two kits together.  Dual rank 16gb sticks.  Thought it would be cool to have 64GB, but don't need it at all.
 
Have had my eye on a 3200CL14 32GB (2x16) b-die kit for $220-240, but there's a variety of faster kits for around $300 or less.  4000CL19 G.Skill b-die is QVL'd for my MB, along with 3600CL16, 3733CL17 and 3866CL18 b-die.  Looking to get better performance in 3dmark benches and in single/low thread gaming.   SOME testing shows pretty dramatic FPS increases with the high clock ram, but other people say lower true latency is key.  Your thoughts?
post edited by kevinc313 - 2019/12/04 13:59:08
#1

5 Replies Related Threads
    Cool GTX
    EVGA Forum Moderator
    • Total Posts : 30978
    • Reward points : 0
    • Joined: 2010/12/12 14:22:25
    • Location: Folding for the Greater Good
    • Status: offline
    • Ribbons : 122
    Re: Considering 4000CL19 B-die RAM vs. 3200CL14 B-die 2019/12/04 20:52:04 (permalink)
    See this thread:  RAM Explained   Check those links in Post 53 & 54

    Learn your way around the EVGA Forums, Rules & limits on new accounts Ultimate Self-Starter Thread For New Members

    I am a Volunteer Moderator - not an EVGA employee

     https://foldingathome.org -->become a citizen scientist and contribute your compute power to help fight global health threats

    RTX Project EVGA X99 FTWK Nibbler EVGA X99 Classified EVGA 3080Ti FTW3 Ultra


    #2
    kevinc313
    CLASSIFIED ULTRA Member
    • Total Posts : 5004
    • Reward points : 0
    • Joined: 2019/02/28 09:27:55
    • Status: offline
    • Ribbons : 22
    Re: Considering 4000CL19 B-die RAM vs. 3200CL14 B-die 2019/12/05 07:19:48 (permalink)
    Cool GTX
    See this thread:  RAM Explained   Check those links in Post 53 & 54




    Hey thanks, I've seen that.  The Techspot test is not particularly useful because they hold the ram timings constant while varying the frequency, which will obviously produce gains, but then test at 1440p where the system is somewhat GPU bound.
     
    Here is the Tom's test where they test a variety of realistic speeds and timings on a 9900K, with 720p gaming so the system is CPU bound.  The results are pretty dramatic:
     
    https://www.tomshardware.com/reviews/best-ram-speed,5951-6.html
     
    On the other hand, here's the Techpowerup test with a 8700k that shows very subtle improvements:
     
    https://www.techpowerup.com/review/intel-core-i7-8700k-coffee-lake-memory-performance-benchmark-analysis/9.html
     
    Guess I'm just looking for consensus that 4000CL19 is faster in the real world and worth the money.
    post edited by kevinc313 - 2019/12/05 17:23:22
    #3
    kevinc313
    CLASSIFIED ULTRA Member
    • Total Posts : 5004
    • Reward points : 0
    • Joined: 2019/02/28 09:27:55
    • Status: offline
    • Ribbons : 22
    Re: Considering 4000CL19 B-die RAM vs. 3200CL14 B-die 2019/12/06 18:09:20 (permalink)
    Did a bunch more research, it turns out my MB uses a daisy chain memory trace config, this means the primary two dimm slots will clock better than the other two, while a 4x dim config will be pretty limited in it's clocking.  4x8gb up to about 3600mhz, maybe 3733 or 3866 with lots of skill, or 2x8gb 4000mhz fairly easy, up to about 4400.  On dual rank 2x16gb, it should be able to handle 4000mhz, which is the highest bin 2x16gb commonly available and the best performing I'm going to get for 32gb, plus the dual rank speed bonus.  Questionable overclocking/timing headroom though.
     
    I've also been able to clock my current 4x16gb ram to 15-16-16-36 1T 3000mhz with 1.15/1.2/1.4 IO/SA/DRAM voltages and it seems to be working so far. 
     
    We'll see if I end up picking up the better stuff, it keeps coming up on newegg around $300 but quickly selling out.  The other option is picking up a $125 set of the TEAM T-Force 2x8gb 4133CL18 sticks and using that for benchmarking and some gaming, but that's kind of silly.
    post edited by kevinc313 - 2019/12/06 19:26:00
    #4
    kevinc313
    CLASSIFIED ULTRA Member
    • Total Posts : 5004
    • Reward points : 0
    • Joined: 2019/02/28 09:27:55
    • Status: offline
    • Ribbons : 22
    Re: Considering 4000CL19 B-die RAM vs. 3200CL14 B-die 2019/12/17 13:22:13 (permalink)
    Just ended up ordering the 2x16GB Trident Z 4000CL19 for $270+tax on ebay from newegg with coupon, I guess that's a pretty darn good deal.  Still spendy though.  Did vaguely consider switching to a ASUS MB (Strix, Hero or WS) with T-memory topology and getting two sets of the Team T-Force 2x8GB 4133CL18, but it seemed like too much work.  Hopefully I get some good results with this kit.  
     
    The 3000CL15 I had set previously was not stable, went back to 2666CL16.  So if this kit runs stable at the approved qvl XMP 4000CL19 it should be a pretty substantial upgrade.
    post edited by kevinc313 - 2019/12/17 13:26:50
    #5
    kevinc313
    CLASSIFIED ULTRA Member
    • Total Posts : 5004
    • Reward points : 0
    • Joined: 2019/02/28 09:27:55
    • Status: offline
    • Ribbons : 22
    Re: Considering 4000CL19 B-die RAM vs. 3200CL14 B-die 2019/12/24 17:17:53 (permalink)
    Got the b-die above, seems to be running really good so far, just straight XMP, booted right up.  In the couple cpu benches I've run so far, Passmark and Timespy CPU, it does about 8-9% better than my 2666CL16.  Totally worth it considering I could have paid $230 for straight 3200CL14 Trident Z in 2x16 at the normal price. 
    #6
    Jump to:
  • Back to Mobile