Early this week at the Intel Developer Forum in San Francisco, California G.Skill showed off new low latency DDR4 memory modules for desktop and notebooks. The company launched two Trident series DDR4 3333 MHz kits and one Ripjaws branded DDR4 3333 MHz SO-DIMM. While these speeds are not close to the fastest we have seen from them, these modules offer much tighter timings. All of the new memory modules use Samsung 8Gb chips and will be available soon.
On the desktop side of things, G.Skill demonstrated a 128GB (8x16GB) DDR4-3333 kit with CAS latencies of 14-14-14-34 running on a Asus ROG Rampage V Edition 10 motherboard with an Intel Core i7 6800K processor. They also showed a 64GB (8x8GB) kit clocked at 3333 MHz with timings of 13-13-13-33 running on a system with the same i7 6800K and Asus X99 Deluxe II motherboard.
G.Skill demonstrating 128GB DDR4-3333 memory kit at IDF 2016.
In addition to the desktop DIMMs, G.Skill showed a 32GB Ripjaws kit (2x16GB) clocked at 3333 MHz running on an Intel Skull Canyon NUC. The SO-DIMM had timings of 16-18-18-43 and ran at 1.35V.
Nowadays lower latency is not quite as important as it once was, but there is still a slight performance advantage to be had tighter timings and pure clockspeed is not the only important RAM metric. Overclocking can get you lower CAS latencies (sometimes at the cost of more voltage), but if you are not into that tedious process and are buying RAM anyway you might as well go for the modules with the lowest latencies out of the box at the clockspeeds you are looking for. I am not sure how popular RAM overclocking is these days outside of benchmark runs and extreme overclockers though to be honest.
Overclocking Innovation session at IDF 2016.
With regards to extreme overclocking, there was reportedly an "Overclocking Innovation" event at IDF where G.Skill and Asus overclocker Elmor achieved a new CPU overclocking record of 5,731.78 MHz on the i7 6950X running on a system with G.Skill memory and Asus motherboard. The company's DDR4 record of 5,189.2 MHz was not beaten at the event, G.Skill notes in its press release (heh).
Are RAM timings important to you when looking for memory? What are your thoughts on the ever increasing clocks of new DDR4 kits with how overclocking works on the newer processors/motherboards?
I probably wouldn’t bother
I probably wouldn’t bother buying high speed memory. I would just look up the most common speed grade of DDR4 on newegg and buy the cheapest name brand stuff. It just doesn’t seem to be worth it to buy higher speed memory any more. Most consumer applications are very cacheable; with 8 MB of L3 and aggressive prefetchers, applications are not going to be very memory dependent. If we get consumer level systems with HBM-based APUs, then we will have even less dependence on external memory. Even a single 1 GB stack acting as a cache would make the speed of external memory mostly irrelevant. If they make a system with HBM2, then it could be 8 or 16 GB. For those they could just go directly to non-volatile storage. A pair of m.2 SSDs could provide a huge amount of bandwidth for swap.
Hehe, yeah I find myself
Hehe, yeah I find myself doing that lately as well, just getting the cheapest name brand stuff (usually g.skill heh) at a reasonable speed for builds. When I eventually upgrade my own desktop I will probably splurge a bit though :).
Since when are the
Since when are the ultra-shitty timings of 14-14-14 became “very low latency”? Sometimes I really can’t tell if PcPer’s article writers just trying to troll or if they’re actually that utterly unprofessional at some things…sheesh. :
Do you know many faster ddr4
Do you know many faster ddr4 banks? Or are you comparing to older gen stuf which had lower latency ratings? If the later is true I’d like to remind you that the real latency is the Clock latency divided by the clock rate and is measured in (nano) seconds.
See for instance:
“Latency moves from DDR3-1600 at CL 11 to DDR4-2133 at CL 15, which was an expected jump as JEDEC tends to increase CL by 2 for a jump in frequency. While having a latency of 15 clocks might come across as worse, the fact that the clocks are at 2133 MTs ensures that the overall performance is still comparable. At DDR3-1600 and CL11, time to initiate a read is 13.75 nanoseconds, compared to 14.06 nanoseconds for DDR4-2133 at CL15, which is a 2% jump.”
Of course if you overclock
Of course if you overclock and use more voltage you can get tighter timings. Out of the box, these new DIMMs do appear to be the lowest latency ones at that capacity and speed. Corsair has 16GB DIMMs running at DDR4-3333 with CL16 and that was the lowest I was seeing on Newegg. Of course the lower capacity kits have lower cas latencies out of the box depending on what speed you buy.
Are there good benchmarks fro
Are there good benchmarks fro DDR4/Skylake comparing ram speeds on real world performace?
Was thinking of building a gaming pc ITX (currently have i7/3770k and 970) for 4k/HDR stuff. Not sure if its worth getting anything faster than 2133…..
PLanning GTX 1080 (gigabyte or MSI)
I think with skylake unless
I think with skylake unless you are overclocking there is not much benefit to going much beyond 2133 though LegitReviews found some performance improvements going to 2666. I would say that unless you are overclocking spend your money on the amount of RAM and less on clocks and then less on latency in that order. With that said I don't think going to DDR4-3200 is that big of a price jump so you might as well :). Your video editing apps might be able to take advantage of the bandwidth, I'm not sure though.
Memory timings is more
Memory timings is more important to me than frequency – especially since most high frequency memory kits seem to sell off of the “MOAR MHZ!!!!1!” marketing effect, while keeping costs low by having shitty timings.
I stick to the AnandTech way of calculating actual memory performance: frequency divided by CAS. Not a perfect measure, but a decent guideline, and one that illustrates why DDR4-2400 C13 might outperform DDR4-3300 C17.
On the other hand, even given great timings on high-frequency RAM, I’d go for a slower kit and rather spend my money on something that actually makes a difference in the workloads I run.