Save Money on RAM.. with Optane?? - Intel Showcase

Linus Tech Tips ·Linus Tech Tips ·2019-05-06 · 1,034 words · ~5 min read
Floatplane YouTube

Transcript

JSON SRT VTT 102
0:02 all right time for a build 8700k
0:06 check z370 motherboard
0:10 check two RAM
0:15 one stick hold on we'll be back
0:24 where the devil is all the RAM
0:31 and that's actually a terrific question
0:35 dram has been in a global shortage
0:39 allegedly due to collusion between the major producers for months so
0:45 when Intel reached out asking for some
0:48 creative ideas for a sponsored video we pitched using their optane caching
0:54 modules as a way of topping up your system memory
1:00 and for some reason pig green lit this
1:03 so let's give it a shot
1:18 so obviously we're still gonna need some system memory but the plan really is to
1:24 take this and use it to replace some of
1:27 this and the rationale behind this is actually a lot less crazy than you'd
1:32 think in our testing we've seen that optane
1:36 performs more like RAM than it does like
1:39 an SSD which is to say that it's got ridiculously low response times compared
1:44 to this all while being much cheaper than this
1:49 and conveniently even though our optane module here plugs
1:53 into an m.2 slot so this is something that is designed for an SSD
1:59 Windows already has functionality that
2:03 lets us use a regular old
2:06 storage device as an extension of the
2:09 system memory so we will give up quite a
2:13 lot of our throughput compared to actual RAM
2:17 but for small transactions that might
2:21 not actually hurt us that much and it will certainly be faster than using a
2:26 traditional NVMe or SATA SSD or
2:31 heaven forbid a hard drive in this role
2:35 to put this theory to the test we're gonna do something a little bit strange
2:39 so we've actually installed two eight gig sticks of memory in our bench to
2:43 give us the benefit of dual channel but then we're gonna use Windows built-in ms
2:48 config utility to artificially
2:52 limit our memory to four gigs
2:58 without a plus sign in there
3:02 so our four gig entry actually gives us just shy of three gigs of usable memory
3:08 all right let's fire up some gaming benchmarks
3:12 these numbers actually start out showing
3:16 little concern for the lack of available RAM
3:19 however in the open world far cry 5 we
3:22 see a significant drop in 97th percentile frame rates as assets load in
3:28 on the fly as for unigine's superposition we did see a small dip but
3:32 that could just be due to run to run variation
3:36 moving on to productivity things get
3:40 painful adobe premiere took a whopping three
3:44 times longer than our native 16 gig
3:48 tests and blender is much the same story
3:51 taking over two hours to complete where
3:55 our non-limited testing gave us a result under 45 minutes
4:00 let's hope for Intel's sake that things go a little bit better
4:06 with optane
4:10 unsurprisingly our gaming results
4:13 don't really change much though we do gain back most of our 97th
4:18 percentile frames on far cry 5 compared to the 3 gig test and unigine
4:23 superposition did pick up a cool 33 points of probably run to run
4:28 variants moving on to productivity things look
4:33 well actually a lot better we shave about a minute off of adobe
4:39 premiere though we didn't get that close to true
4:42 RAM but as for blender wow
4:46 blender took a whole hour off the render
4:50 time i mean to be clear that still 25 and a
4:54 half minutes longer than if we had all just RAM
4:59 but considering that this render calls for a minimum of 12 gigs of RAM and
5:04 takes about 12 and a half we know that a
5:07 considerable amount of our working data was sitting in our optane module
5:13 but that's only part of the story let's take an even closer look at these
5:18 results what you're looking at now is a scatter
5:21 plot of frame times and the corresponding disk activity where we can
5:26 see that most of our spikes in
5:29 instantaneous frame times so these would be perceived as hitches
5:33 are actually caused by spikes in activity and this is especially evident
5:39 when we look at far cry 5. we can also look at how the relationship
5:43 works between drive activity and CPU
5:46 usage during our premiere and blender tests where we can see well mostly a
5:53 mess but but squint harder look at this
5:56 the dips in drive activity actually correspond to spikes in CPU usage
6:03 meaning our CPU is being kept fed so the
6:07 less a drive thrashes the more our CPU
6:12 gets to work nowhere is this more apparent than our
6:15 blender test where our non-optane result had the CPU working at below 50 capacity
6:22 most of the time so what can we take away here
6:27 well for one octane memory modules aren't going to be replacing system RAM
6:33 anytime soon but
6:36 think of the potential out of necessity it's common practice
6:40 for computing workloads to be coded around the availability of modern
6:45 hardware so we needed to create a pretty contrived experiment to show such a
6:50 clear benefit today but imagine if developers or creative
6:55 types or researchers were free to create
6:59 much larger data sets knowing that they'd have a cost-effective way to work
7:04 around bottlenecking their CPU we actually saw this in action in our
7:09 optane 900p video so for today we're still recommending
7:14 optane modules for accelerating slower storage devices like hard drives
7:20 but this was a really fun exercise in future
7:23 crafting and it really shows off the potential of the underlying optane
7:28 memory so thanks for watching guys if you disliked this video you can hit that
7:32 button but if you liked it hit like get subscribed maybe consider checking out where to buy the stuff we featured at
7:37 the link in the video description again to accelerate other storage not to
7:41 replace your system memory while you're down there you can check out our merch
7:44 store which has cool shirts like this one and our community forum it's kind of
7:48 a cool place to hang out and talk tech