Board 8 > Can someone tell me what the difference between these two things are

(+) Yeah! (+)

https://www.newegg.com/g-skill-32gb-288-pin-ddr4-sdram/p/N82E16820232218?Item=N82E16820232218

https://www.newegg.com/g-skill-32gb-288-pin-ddr4-sdram/p/N82E16820232091?Item=N82E16820232091

Because they look the same to me but one is $80 cheaper

---

You can pay $80 more for a latency of 14 instead of 16

---

Speed, CAS latency. The returns are diminishing, but that $200 model is pretty quick.

---

PrinceReva posted...

Speed, CAS latency. The returns are diminishing, but that $200 model is pretty quick.

Is it really worth $80 more?

---

_stingers_ posted...

Is it really worth $80 more?

Eh, to someone who insists on every slight advantage, Those clocks are milliseconds, so what's .02 worth to you? Probably not worth it to me.

---

Wall of text warning.

Showing up a bit late to this topic. The difference is a bit more substantial than CAS Latency of 16 vs 14 for certain people (i.e. overclockers). If you weren't aware, you can overclock RAM to force it to have higher frequency or better CAS latency (amongst other things). Certain RAM kits have a lot more headroom for overclocking than others.

Background: Even though there are many sellers/brands of RAM (G.Skill in this case), only three companies actually produce it. Those companies are Samsung, SK Hynix, and Micron and they each have their own processes for the production of RAM at various qualities/densities.

The 3200C16 kit was originally manufactured by SK Hynix using their 8Gbit MFR process. This manufacturing process is not known for overclocking well. So basically, you wouldn't expect to be able to overclock the performance much above the originally promised 3200C16 spec.

The 3200C14 kit on the other hand was made by Samsung using their process often just called "B-Die". This process is very well-known for producing RAM that can overclock well-beyond the original spec. The $80 price differential could very well be worth it to someone looking for top-shelf overclocking potential.

To someone who is uninterested in overclocking, the $200 price-tag doesn't make much sense. There are cheaper 3200C14 kits out there that aren't B-Die, and would perform just fine by enabling the XMP/DOCP profile and calling it a day.

EDIT: Actually, 3200C14 doesn't really exist outside B-die at this density. But the point still stands, for someone just using XMP/DOCP, a 3200C16 kit gives extremely similar performance.

---

_stingers_ posted...

Is it really worth $80 more?

No chance in hell unless you are rolling in money.

tyder21 posted...

Wall of text warning.

Showing up a bit late to this topic. The difference is a bit more substantial than CAS Latency of 16 vs 14 for certain people (i.e. overclockers). If you weren't aware, you can overclock RAM to force it to have higher frequency or better CAS latency (amongst other things). Certain RAM kits have a lot more headroom for overclocking than others.

Background: Even though there are many sellers/brands of RAM (G.Skill in this case), only three companies actually produce it. Those companies are Samsung, SK Hynix, and Micron and they each have their own processes for the production of RAM at various qualities/densities.

The 3200C16 kit was originally manufactured by SK Hynix using their 8Gbit MFR process. This manufacturing process is not known for overclocking well. So basically, you wouldn't expect to be able to overclock the performance much above the originally promised 3200C16 spec.

The 3200C14 kit on the other hand was made by Samsung using their process often just called "B-Die". This process is very well-known for producing RAM that can overclock well-beyond the original spec. The $80 price differential could very well be worth it to someone looking for top-shelf overclocking potential.

To someone who is uninterested in overclocking, the $200 price-tag doesn't make much sense. There are cheaper 3200C14 kits out there that aren't B-Die, and would perform just fine by enabling the XMP/DOCP profile and calling it a day.

EDIT: Actually, 3200C14 doesn't really exist outside B-die at this density. But the point still stands, for someone just using XMP/DOCP, a 3200C16 kit gives extremely similar performance.

Thank you for this. I learned a lot today. And you helped me make a very educated decision ,

---

tyder21 posted...

Wall of text warning.

Showing up a bit late to this topic. The difference is a bit more substantial than CAS Latency of 16 vs 14 for certain people (i.e. overclockers). If you weren't aware, you can overclock RAM to force it to have higher frequency or better CAS latency (amongst other things). Certain RAM kits have a lot more headroom for overclocking than others.

Background: Even though there are many sellers/brands of RAM (G.Skill in this case), only three companies actually produce it. Those companies are Samsung, SK Hynix, and Micron and they each have their own processes for the production of RAM at various qualities/densities.

The 3200C16 kit was originally manufactured by SK Hynix using their 8Gbit MFR process. This manufacturing process is not known for overclocking well. So basically, you wouldn't expect to be able to overclock the performance much above the originally promised 3200C16 spec.

The 3200C14 kit on the other hand was made by Samsung using their process often just called "B-Die". This process is very well-known for producing RAM that can overclock well-beyond the original spec. The $80 price differential could very well be worth it to someone looking for top-shelf overclocking potential.

To someone who is uninterested in overclocking, the $200 price-tag doesn't make much sense. There are cheaper 3200C14 kits out there that aren't B-Die, and would perform just fine by enabling the XMP/DOCP profile and calling it a day.

EDIT: Actually, 3200C14 doesn't really exist outside B-die at this density. But the point still stands, for someone just using XMP/DOCP, a 3200C16 kit gives extremely similar performance.

This post certainly deserves a VAPE.

---