1660Ti Max-Q vs 2060 Max-Q

Howdy.

I'm shopping for a new notebook. My current is an Alienware 15R4 with an i7 and 1060. It continues to be the most reliable, durable computer I've ever owned, and I could probably have continued using it for a good two or three years before I feel limited by its midrange-four-years-ago graphics.. but I've been seduced by newer, thinner form factors with tiny bezels, so I'm ready to retire it to media-server duty.

I know it's not a gaming laptop, but I'm interested in the HP Envy 15 for its premium-design/quality, subtlety, and reasonable capability with glowing reviews. My instinct is to opt for the 2060 Max-Q as it's the newest option available, but unfortunately it's only paired with the OLED screen (I'm NOT a fan Samsung's panel). The other option is a 1660Ti Max-Q. I'm having a hard time finding a straight answer on how they perform against each other as it pertains to laptops, both in Max-Q form.

All-else equal, what kind of performance difference could I expect between a 1660Ti Max-Q and a 2060 Max-Q (at 1080p without ray-tracing)?
Bonus questions: Am I remiss for shopping Turing-card laptops right now? When are the midrange 3000 cards expected to reach laptops? Do you think the performance gain for the price will be worth the wait for a moderate-gamer who was still satisfied with his 1060?

 

If they are still satisfied with their 1060, they can coast to mobile Ampere.

https://www.notebookcheck.net/NVIDIA-GeForce-RTX-3080-Mobile-GPU-Benchmarks-and-Specs.497450.0.html

 

The RTX 2060 Max-q is faster but not by much, I'll say about 5-10% over the GTX 1660Ti Max-q.

Mid range Ampere laptop GPUs will probably released in Q2 2021 or later if you can wait that long.

 

quick question?

I have a gtx 1060 that I run at half speed aka a 1050ti and it plays all games with decent settings and 1080p? any particular reason why you want to upgrade besides the fact the gpu is 4 years old? I say run it to the ground and hold out for the 3080 mobile or 3060 pending prices.

 

The performance gain will definitely be worth the wait. Turing is a comparatively minor bump over Pascal; Ampere is a huge jump over Pascal. I'm still playing on a 1060 myself and enjoying it; I'm holding out for a 3000-series upgrade. Laptops with 3000-series GPUs will most likely hit the market next year. If you're satisfied right now with your 1060, I strongly encourage you to stick with it until you can get a laptop with a 3060 or higher. You'll be getting a far better performance increase for the amount of money you'll be spending.

 

I would have to agree, stick it out till spring-summer 2021. I think somewhere between March and May sounds like a safe bet for Ampere laptop silicon.

 

Thank you all for your input.

I want to upgrade not because the GPU is older (as I mentioned, I'm actually still quite happy with it), but rather because gaming-oriented laptops have made strides lately with regards to weight and thinness.. and I'm a sucker for thin bezels. I also want a more modest appearance, something unsuspecting-but-classy in a public/office setting. The Alienware was pretty good for its day, but it's a chunker by today's standards. When I bought it, I actually didn't think it would become my daily device.

Something that I didn't mention is that for the past month, I've been working with a Razer Blade 17 Pro/2080 Super Max-Q. It's perfect for me, especially since I was able to get it during a good sale. I've been getting ready to return it, though, because it crashes every time it attempts to resume from sleep (only when I'm using my own 2x M.2s in RAID0). I bring this up now because I've actually made potentially big progress today that has me more hopeful than ever that it can be solved. In the event that I still can't get it working right, y'all have convinced me that the wait for Ampere is probably worthwhile.

 

POS anything razer

 

Totally off-topic, but why people do such things like RAID 0 two NVMe SSD? There is virtually zero performance gain in most of cases. You just double the risk of hardware failure.



Even the difference between super fast NVMe versus SATA III SSD is negligible in real world use (loading games, OS, apps, etc) :

 

My guess is that it's for bragging rights.

I'm looking forward to Ampere's implementation of I/O Directstorage, which may finally make NVMe drives worth it for gaming.

 

Actually, I couldn't care less about benchmarks. The reason I did it was because 2x 2tb SSDs cost about $400, whereas a single 4tb right now is closer to $700+. They're in RAID because I'm tired of working with two different drives. I love the simplicity of having just "one" drive again. And they're NVMe simply because one of the drive slots requires it.

 

That doesn't really make any sense..cheaper ok good point but running raid because you want one drive... illogical

 

Care to offer a better solution?

 

Tried this one?

https://www.windowscentral.com/how-create-one-large-volume-using-multiple-hard-drives-windows-10

 

Thank you for the link.

I'm not all that wise to computers, but I don't see how that's any better than what I'm doing currently, except I'm currently doing that at a lower level, which my intuition tells me should be more reliable/efficient.
Wouldn't that solution also require me to have yet another drive off which to boot Windows? It doesn't sound like it'd be safer to keep data on, either. I opted for RAID0 with the understanding that backups would be of increased importance, so I'm fully willing to take the risk with striping, if that's why this is so "illogical."

 

I don't see anything wrong with Raid 0 to keep things simple so long as you understand the risks and keep backups. It's better than storage spaces IMHO.

 

If you like having 1 drive..why would you buy two then say again you like one..no benefit storage or speed wise..you essentially spent two times the amount of money for no reason

 

The poster wants 4TB, buying 2x2TB drives instead of 1x4TB saved quite a bit of money. That is a benefit if you ask me.

 

hes running raid 0 and does not get 4tb of storage...he gets 2tb at faster speeds but ultimately at no benefit.

edit....

if applications benefited from the higher speeds i'd say its a wise even intelligent move but since there is no benefit vs sata 500mb/s speeds its a waste of storage and money

 

RAID 0 doesn't reduce the amount of usable space, unlike RAID 1.

 

nd raid

oh really whoops sorry guys I was under the impression raid 0 was like 1 drive...mixed up raid 1 and raid 0...my bad

but even to save this arguement it's still not worth it for OS


source https://www.thetechlounge.com/blog/raid-0-1-5-6-10/
PROS AND CONS
RAID 0 offers efficient performance because there is no overhead caused by parity controls, and also because all storage capacity is utilized. The technology is simple and easy to implement with different types of storage interfaces, and introducing new drives is a good way to increase performance without diminishing returns.


That being said, RAID 0 isn’t fault-tolerant. This means that if one drive fails, all the data in the RAID array will be lost. The more drives you have, the greater is the risk of data loss. For this reason, RAID 0 should not be used for OS and such. It is best for non-critical data that just needs to be read/written at high speeds.

 

But it fits that users need just fine. Saving money to get the drive size required albeit with a higher risk of failure.

I've used raid 0 in many machines for many years (as OS drives) and I never had an issue. I wouldn't use it for something that required high reliability though.

 

I think in this case duplicating the chance of losing ALL your data, just not to have an extra partition is a bit crazy man. The expected lifetime of SSD is lower than HDD, mostly because the TBW (Terabytes Written). I wouldn't actually store data on a SSD. SSD are good for games and OS files becase they are fast, but not really reliable for long term use. I suppose everyone should have the most sensitive data on a cloud storage for maximum safety.

 

Just have backups. I back up to my NAS every night using Macrium Reflect with an "incrementals forverer" model (merge after 10). I used it once, worked like a charm.

I do test it from time to time just to make sure though.

 

It does double your chance of losing your data, there is no arguing that. Having said that if it fits the users need and he/she backs up important data to multiple sources that is not a legitimate point (you should do so even if you have a single drive).

I don't quite worry about the SSD life vs HDD argument. HDDs in laptops have their issues, I was a tech and did a LOT of data recovery on laptop hard drives because spinning platters + write/read heads + movement = no bueno. The TBW is an estimate and really an underestimate for sure. I don't know what drives the OP is using but if it's something like a 970 Evo then the OP can write almost 700GB per day to each drive (1.4TB total) for 5 years and still be within warranty specs. I sure don't write that much on a daily, do you?

Also there have been lots of fun torture tests that shows that even TLC drives last much longer than the quoted TBW mark (like at least 3x). Then there is the thought that in 5-10 years will we really be using the same NVMe drives? I won't, but some will (but these are typically people who write less information anyway).

And no I wouldn't store data on SSDs long term, but that's only because the $/TB is not at a sweet spot for me to do so. When 15-20TB SSDs are ~$200 I will be switching over for my plex server that currently has 100TB of redundant storage. Enterprise servers are already making this switch for applications that need speed, but they have a wee bit more money than I.

 

@bierdc - Please ignore the confused and judgmental nature of this thread. Usually this forum is friendlier than this thread suggests. There is nothing wrong with your setup.

 

Yeah we are not attacking him I have run raid in the past its nice too have the extra speed but past nvme drives it doesnt matter...do what you like it should have been worded differently but wasn't...faster is faster....but was just giving a different angle..dude use raid and the likelyhood of failure is almost nhil...it was just a optional opinion that may have been not needed