Proesterchen
-
Posts
27 -
Joined
-
Last visited
Proesterchen's Achievements
-
All of them. OK, I lied. You'll get three.
-
Do you remove the sponsored stickers during testing of the devices?
-
I'm not even sure RST supports caching a RAID-array, but even if it does, the usefulness of this module would highly depend on your use case. Do the video files in use at any one time fit mostly or entirely on the Optane module? (16GB / 32GB) How is your server connected to your workstations? (if it's just GbE, forget about it) Also, a 20TB, 5 disk RAID 5 sounds like a terrible idea, I hope you have a proper backup strategy. (and adhere to it!)
-
If there is anything I would have liked to see added to this video, it would be a straight-up test of the $80 module vs an $80 SSD. (And yes, I can guess Intel wouldn't particularly like that, but if the question is how to best invest $80 of your PC budget, this is what it comes down to.)
-
Because even a single channel DDR3 interface is massive compared to all the rest that currently makes up the "chipset", i.e. PCH, and would add significantly to its die size, package size, and number of balls. (and no, in this case, that's not a good thing) Also, the number of people upgrading PCs and gutting their old stuff for re-use, rather than selling them off or just giving them away as a whole, is basically a rounding error when compared to the roughly 200 million Intel-based PCs that will ship this year. There is simply no market for the technology you envision,
-
Intel Optane Exclusive Hands On - SSD Crushing Performance
Proesterchen replied to CPotter's topic in LTT Releases
Thanks! Yes, the files themselves don't go away because Optane uses non-volatile memory (i.e. doesn't need constant energy to retain data, like a regular SSD/HDD), but the software controlling what files are the most important at this very moment, and therefor cached on the Optane module, will have to discard data used less recently when fresh, more important data are identified and hence move onto the module. The problem here is the Optane module's very limited size (much like it was with the NAND-based caching solution Intel tried to sell the world on 6 years ago), which means that unless you use very few programs beyond the OS itself, data will need to be shuffled in and out based on whatever caching strategy the software implements. In my example, when you've been using your PC for a while before the monthly update-induced reboot, chances are that most of the OS files cached after your last reboot have been purged from the Optane module to make way for files from more recently used applications. That is why there is no speed-up in this scenario, the data are simply not present in the quick cache because they haven't been used in a long while. Unfortunately, Optane's strategy to boost second-time-round reboots means that OS files will have been cached after the first reboot, throwing out (now useful again) apps, games, or files. This is what results in the need to re-cache (i.e. read from the slow HDD) parts, most, maybe all of the apps you regularly use the first time around. IMHO this whole idea was flawed in 2011 and is still flawed in 2017, regardless of the use of X-point or NAND as the actual file storage technology. It has not taken off in the market in the past 6 years, and with regular SSD pricing hopefully not exploding as much as new capacity comes online in the rest of the year, the business case for Optane doesn't look particularly enticing. -
Intel Optane Exclusive Hands On - SSD Crushing Performance
Proesterchen replied to CPotter's topic in LTT Releases
There's not benefit from separating the OS from the rest, the benefit demonstrated in the video is solely from putting the OS files needed during boot-up on much faster media, hence reducing the time the PC is waiting for them to be read. Which is what brings me to the complete lunacy of this technology: It's great for doing things a second time. Reboot after having run your PC for days, weeks, or the full month? Prepare to be waiting for a long time, because few of the above-mentioned boot files should still be cached on the quick storage (here: Optane, but the same was true for the old NAND-based version of the same caching technology), because they should be replaced with applications and data that you have used more recently. But boy howdy, after that first reboot, your next one is going to be, well, quicker. Why you would reboot again after already waiting to get back to work? No idea. Probably to show that this product has any merit. So lets not reboot, and instead go to work, or "work" as I call it. Open your applications of choice (game), open your files (fav list of servers), and open your browser if you desire contact with the world at large. Will any of that significantly benefit from this brilliant technology? Of course not, because most of the data once present in the caching module was just replaced with OS boot files, you know, to make that coveted second reboot that much quicker! But soon everything will be great again, because the software has now recognized your desire to hasten that particular application and will do its best to cache important parts of it on the tiny module. Of course, that will again slow down reboots, should your overclock crap out and lock-up your PC. And what if, pray tell, you'd like to play another game? Well, I see what you did there my friend, and back to HDD-ville just for that! You are a savage! You are not worthy of caching technology! You might indeed be the use-case for an already established, non-proprietary, always-performing technology! Let me introduce you to the magic of SSDs ... -
Which I'd be perfectly fine with, given the value of reviewing stickered-up products. Just to be clear, if this company wants to show off the terrible quality of their wares, by all means, have them product-place the living daylights out of LTT content, but don't compromise product reviews by having said product covered by stickers.
-
The stickers change the surface(s) and the looks of the device, both of which are part of a proper evaluation. I'm sure you can argue that the stickers in this particular video didn't do much other than showing the lackluster quality and hilarious lack of design sense (sure, lets put a patterned sticker on a curved surface, that's going to look great! ) of the company peddling them, but we wouldn't know that unless they get to test the product without them applied. I'm explaining my downvote because voting up/down on Youtube has no way to relay my reason, and IMHO the actual test content did not warrant said downvote. I think I've posted on this topic before, you can easily check this. It's just that I've come to the conclusion that the use of stickers in a product review will equal a downvote from me going forward, because it invalidates at least part of the test criteria. (not that big a part here, major part in phone reviews, hilariously important part in notebook reviews with stickered-over touchpads.)
-
Just commenting here to say I downvoted the video on Youtube because of the ugly stickers, not the actual content or test. Also, I will downvote futures videos where products are tested covered with these stickers, because that is not a valid test of the actual product.
-
Man, these ugly stickers rub me the wrong way in every video. I really hope they're only applied for sponsor-friendly B-roll and aren't part of the review proper.
-
As an owner of both an original Cathar/Little River White Water and Storm, this was the singularly most interesting and useful video LMG has probably ever produced. Though, to be honest, it has me contemplating fixing the Storm up with a modern mounting plate (ghetto-mounting it works on 115x boards) rather than investing in a new block despite the performance advances. Ah, watercooling, so useless, so fun, so annoyingly noisy.
-
I'm running a 9800 GTX+ right now and can't even play Rocket League with it cause it simply isn't 3D-stable anymore. (just age, never overclocked it) I'd like a Polaris card because I've heared being able to go longer than 40s in a game without crashing the PC enhances the experience. Not sure if true or hyperbole.
-
Which version of the A9 is in your Iphone 6S sample? I suppose with the Samsung A9 suffering significantly in terms of battery life running Geekbench, it might be more susceptible to thermal throttling. Unless of course Apple designed their thermal solution to deal with the entirety of the Samsung A9's load profile, which they very well may have done.
-
I don't think the sponsor stickers look any good on this. In fact, while I understand they provide your review units, I truly hope you test the devices without the stickers applied because it's simply not fair to the manufacturer to mess with its original surfaces and design and then have that be part of the evaluation.
.png)
