Review: Patriot Viper Steel 4000 MHz DDR4

#5632914 Posted on: 01/29/2019 12:46 PM

---

No RGB? Where's my wallet...

#5632941 Posted on: 01/29/2019 01:50 PM

---

Oh plz no to RGB. Instead of wasting money on beautification spend it on a better heat-sink or higher binned die

#5632951 Posted on: 01/29/2019 02:04 PM

---

Can they do something high on amd ryzen?

#5632971 Posted on: 01/29/2019 02:58 PM

---

BLACK IS BACK !!!

#5632980 Posted on: 01/29/2019 03:15 PM

---

I have never understood the allure of high speed memory considering a 1 to 2 fps gain and most benchmarks within the margin of error from regular memory. What does appeal to me is no nonsense (no RGB BS) good cooling design, small footprint and solid lifetime reliability.

#5632983 Posted on: 01/29/2019 03:33 PM

---

You are wrong
In cpubound games with 2080/2080ti, higher memoryspeed (higher bandwidth) can increase minimum fps by 10-20%. Read: New Tombraider, BF V etc...

If you trying in non- cpubound and memorybound scenarios, the result is equal to zero.

#5633051 Posted on: 01/29/2019 06:35 PM

---

Consumer and most professional workloads aren't really bound by memory speeds. While lower latency is always preferable (for single-threaded bottlenecks), architectural and physical limitations of the current DRAM tech present a very steep curve to any meaningful solution. As far as the bandwidth, applications that benefit it already rely excursively on the GPU, while the rest (large DB, data mining, etc.) are way out of scope for the PC users. From my experience, paying premium for anything over DDR4-3200 is not worth it.

#5633064 Posted on: 01/29/2019 07:08 PM

---

Thanks boss! Another awesome review!

Also it's articles/reviews like this that remind why and that I glad spend a bit more on my 2x TridentZ 16GB 4266C19 kits (1 used kit and 1 new) running at 4000 17-17-17-38@1.425v memetest stable.

Obviously 3200C14 and 3600C15-16 are usually the sweetspot performance/price wise but 4000-4400c19 aren't bad either. :p

#5633156 Posted on: 01/29/2019 09:24 PM

---

Guys, do you think that I should upgrade RAM in combination with Hashwell-E? currently I have Kingston 16GB KIT DDR4 2133MHz CL14 HyperX Fury Black Series and I have 2 questions. Should I upgrade to 32 GB kit CL15 3000-3200 MHz? Should I use dual or quad channel in combination with ASUS X-99A? It supports DDR4 up to 3200 MHz.

Thanks!

#5633241 Posted on: 01/30/2019 04:49 AM

---

You are wrong
In cpubound games with 2080/2080ti, higher memoryspeed (higher bandwidth) can increase minimum fps by 10-20%. Read: New Tombraider, BF V etc...

If you trying in non- cpubound and memorybound scenarios, the result is equal to zero.

Running asrock taichi 370 with 8700k oc'd to 5ghz, vega 64, I originally purchased a cheap g.skill 2066 set 16GB from the qvl list for the build, a few months later I decided to upgrade the memory to a g.skill 4266 16gb set B-die also on the qvl list in hopes of getting more performance.

I literally noticed no difference with the upgrade in real world gaming, benchmarks or anything else, maybe a few fps here and there, a slightly quicker boot time, but I did notice a much more responsive bios, never understood that one.

#5633255 Posted on: 01/30/2019 06:17 AM

---

Realizing this was tested on an i7 - Has anyone here ran DDR4-4000MHZ with a Ryzen 2xxx by chance post AGESA 1006 with stable results? Starting to lick my chops watching those kits on Newegg... hmmm

#5633263 Posted on: 01/30/2019 07:13 AM

---

Running asrock taichi 370 with 8700k oc'd to 5ghz, vega 64, I originally purchased a cheap g.skill 2066 set 16GB from the qvl list for the build, a few months later I decided to upgrade the memory to a g.skill 4266 16gb set B-die also on the qvl list in hopes of getting more performance.

I literally noticed no difference with the upgrade in real world gaming, benchmarks or anything else, maybe a few fps here and there, a slightly quicker boot time, but I did notice a much more responsive bios, never understood that one.

You are most likely gpu bound...

#5633265 Posted on: 01/30/2019 07:21 AM

---

I think it is nice set. But people should check beforehand if they will benefit from given clock.

#5633424 Posted on: 01/30/2019 04:53 PM

---

Running asrock taichi 370 with 8700k oc'd to 5ghz, vega 64, I originally purchased a cheap g.skill 2066 set 16GB from the qvl list for the build, a few months later I decided to upgrade the memory to a g.skill 4266 16gb set B-die also on the qvl list in hopes of getting more performance.

I literally noticed no difference with the upgrade in real world gaming, benchmarks or anything else, maybe a few fps here and there, a slightly quicker boot time, but I did notice a much more responsive bios, never understood that one.
GPU bound.
On same setup, one with Vega 64 second time with 1080 gaming x there is 20-40% more FPS on later, with same game settings in various games I tried.
Like 1 spot in Americas army proving grounds Vega shows 130 140fps, 1080 shows around 200. Other spot Vega 200, 1080 280+.
Vermintide 2, Keep, ~140 Vega, 1080 shows ~180.
But. My CPU runs 5ghz. Ram runs 4200 17 18 18 32. Very tight secondary, tertiary timings, optimized for minimum latency (38ns Aida64) and throughout ~64 read write, 6k less copy.

And games ( amost all of them)runs different for me with Vega and 1080. The later in same scenarios is faster. Have both cards, as Vega sits in my sons PC, 1080 in my.

#5633574 Posted on: 01/30/2019 11:21 PM

---

Possibly, but I wasn't really getting any notable performance boosts anywhere else either other than a more responsive bios, maybe an extra second shaved off my boot time, but that was about it.

I have a second vega 64 laying around for crossfire if I want but I haven't tried it since none of the games I usually play are supported, maybe the faster ram will help in crossfire? I just wouldn't see the gains.