Nvidia GeForce GTX 1050 & 1050 Ti Review

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

techy1966

Reputable
Jul 31, 2015
149
3
4,685
@InvalidError "Because Guru3D ran their tests at very high or ultra presets where both 1050s are under 40fps most of the time even at 1080p instead of aiming for detail levels that yield frame rates people would actually want to play at. Once you reduce details to achieve a more readily sustainable 60fps, the memory requirement drops, the 1050Ti's extra 2GB VRAM becomes mostly unnecessary and the performance gap with the 1050 gets that much narrower."

I am sorry but the way Guru3D tested is the proper way to do a review if you want to see the full performance of what each card has to give the end user and at the end of the day this lets the end user know just how well a card will perform. It does not matter how the end user ends up using the card in the games if he or she wants a steady 60 FPS they will set the game to get those frames but if you hand pick games and settings in a review like a lot of site are doing then you actually never really know which card is the fastest and most suited to your needs. All you get is part of the picture from reviews that set the cards to work at less than the game can produce. I say this because in the reviews I have seen for the 1050 series you see the 1050 non ti always beating out the AMD 460 card but in guru 3D you see the 1050 non ti getting it's butt handed to it by the AMD 460 so which is right is the 1050 faster or slower than the AMD 460.

From the Guru3D review I would say the 460 is faster most of the time when the cards are pushed to their limits. Crap in a few instances it even beat out the 1050Ti..which is kinda funny because in other reviews the 1050 Ti was a lot faster. For me I would want to know which card had more fuel left in the tank before I bought a new card not a card that looked faster because of tweaked settings from a review. It clearly shows the 1050 series are very under powered when pushed as hard as a 460 or 470 in the Guru3D review and that is what a review is meant to be a balls all out test to show which card or piece of hardware has that extra juice to pulled ahead of the others.

Sure you could also show what each card can do when the settings are tweaked for frame rates but you have to include both sets of data so a end user gets all the needed information before buying a new card.
 
If you read through the forums, people have been recommending the 750ti as recently as this morning. Not because it's the best card or the best value, but because it was the only upgrade possible for people with pre-builts. The 1050 series is targeted at all the people who were considering the 750 ti, as well as anyone who already has a 750 ti. The 1050 ti may not deliver on value over the 1050, but it's the only performance boost available in that segment so they can charge a premium.
 

Ne0Wolf7

Reputable
Jun 23, 2016
1,262
5
5,960
I know, I'm being that guy, but in the first sentance of the "Benchmarks" section, when the trsolution was named: "Nvidia’s GeForce GTX 1050 family targets the same 1920x01080 resolution and e-sports demographic as AMD’s Radeon RX 460..." there's a typo, 1920x01080 instead of 1920X1080 (extra zero before the 1080)
 

ravewulf

Distinguished
Oct 20, 2008
972
33
19,010
Looks like I'll probably go with the 1050 and not the 1050 Ti. Only thing that could change that is if Skyrim Special Edition + mods benefits a lot from having a 4 GB card instead of a 2 GB card.
 
It seems to me that $30 is not an unreasonable price for 2GB of VRAM. It may not be great for those who strongly favor FPS over visual quality, but for those going for quality, I know I'd consider it worthwhile myself.
 

Ne0Wolf7

Reputable
Jun 23, 2016
1,262
5
5,960


The 1050 doesn't need PCI power, so thats a must for some. Also, the good 1050 Ti is around $20 to $50 less than the RX 470, unless you get the referene card.
 

InvalidError

Titan
Moderator

I fundamentally disagree: artificially running GPUs into the ground with unrealistic settings nobody would ever play the games at using the GPUs being compared makes no sense since It over-emphasizes bottlenecks that wouldn't be there or at least be substantially less significant under a more GPU-appropriate detail level.

What is there to learn from a 1050(Ti) running a game at 1080p Ultra and less than 30fps aside from those GPUs not being playable at that detail and resolution combination? Nearly nobody wants to play games at 30fps, which renders the 1050Ti's 50% lead in worst-case scenarios completely moot. Once you lower details to an actually playable level, the performance difference between the two GPUs shrinks to 10-15% and for most budget-conscious buyers, that's the comparison that matters most.

Everything at ultra is only meaningful for higher-end GPUs that are actually capable of delivering playable frame rates at that level.
 
the two things i am most interested in seeing is these gpus in a low profile flavor and their mobile versions. we have been stuck with the 750ti in low profile for over 2 years now and i dont see any reason why board partners cannot squeeze these two pascal gp107 chips in.

but mostly i would like to see pricing structure for laptops with either one of the gp107 chips. its my understanding that the pascal mobility parts are sharing their desktop counterparts so something with 85% of the gtx1050s power in a laptop would be a game changer if it can be bought for $700 or so.
 


^^This. I just helped a friend upgrade his four year old Dell XPS with a 4GB RX 460 knowing full well he'd have to crank down graphic settings to get FPS up. He's always been a console gamer and this is his first attempt at getting into PC gaming without breaking the bank.

Further, it's the same reason why sites like Toms and Guru3D often do not show lower tiered cards in higher resolution benchmarks. There's absolutely no point in it.
 
I will never understand why big sites like this one cant just give us more than 1 freaking preset of quality in the review...
This is quite pointless.

The 1070 is done at ultra 1440p. The 1050 at medium 1080p.
Who are we supposed to gaze performance per dollar if they are NOT the same?!
 

payneg1

Distinguished
Nov 1, 2011
38
0
18,530
I game at 1600 * 900 (20"). I have many friends who game at 1366 * 768 (17 - 18"). There are way too many people who want to play games with the best possible settings in this resolution.

Check the steam survey on resolutions, you will know. For us ultra-setting is a must. Crysis Warhead looks better IMO at 1600 * 900 ultra vs 1920 * 1080 medium. For many of us, the Guru3D reviews make sense.
 

kinney

Distinguished
Sep 24, 2001
2,262
17
19,785


I was in your exact same boat. I used a 5870 from release date till just last month. It was all I needed. I decided once my machine died (the PSU finally did, 5870 would still be going strong) that due to my gaming habits and need for a whole new system to just move to a NUC. I'm using a Skull Canyon now, and it's slower than my 5870 but since I don't really play the really intensive stuff (FPS mainly) I'm fine.

I've considered a new mini-ITX build using Intel IGP or this 1050 but I don't think it makes sense for me. I skip over Samsung RAM and SSDs in favor of Micron RAM and Intel SSDs to get the best shot at reliability too even if they are slower in many cases (I went with the Intel 600P SSD and Crucial Ballistix RAM).

Consider a NUC if you have a wider system failure at some point in the future, saves on power too which I don't mind. I took the Thunderbolt3 port as an escape hatch in case I regretted the decision for an eventual eGPU case.
 

InvalidError

Titan
Moderator

If you are looking at the 1070, you aren't shopping anywhere near the same performance range as the 1050. Benchmarking a 1070 at settings meaningful to a 1050 shopper will produce absurd results on the 1070 while benchmarking at settings meaningful to a 1070 shopper will produce pointless slideshows on the 1050. The performance gap between the two is far too wide to produce meaningful comparison.


I wouldn't be so sure about that: 1366x768 (the second most common resolution on the Steam survey) is half the pixel count of 1920x1080. If Ultra benchmarks were re-done at that lower resolution, results could be drastically different due to all buffers being half as large, require half as much RAM bandwidth and half as much compute.


Although not in stock, the mini is listed at $157 including shipping:
http://www.newegg.ca/Product/Product.aspx?Item=N82E16814500412&cm_re=gtx1050-_-14-500-412-_-Product

Makes me nauseous when I look at BestBuy.ca which is still trying to sell people R5-250 and 750Ti for $120-180.
 

Joe Black

Honorable
Jul 3, 2013
88
0
10,640
Yep... Generally a big fan of AMD, but they "promised" value and they never delivered. So I actually jumped ship this gen. Once I found a good deal of course.

My advice if it were to be asked. Rather just put the cards out there, let the market decide about value. Clearly neither AMD nor NVidia have any control over the final prices of the cards.
 


First you are blaming AMD for the high prices by saying they broke their promise, and then you say AMD has no control over the prices.
 
Status
Not open for further replies.