News NVIDIA GeForce GTX 1650 Launching April 2019

https://wccftech.com/nvidia-geforce-gtx-1650-turing-gpu-confirmed-launch-next-month/

Pray to the gods of low profile dual slot cards, for a 1050ti replacement !!!!

 

[VegetableStu]

oh didn't know there were spec rumors o_o but it's now kinda closer to the 1660ti

 

[CC Ricers]

EVGA still is using a triple slot cooler for a <120W card... wat

 

[NateDawg72]

EVGA still is using a triple slot cooler for a <120W card... wat

and their triple slot cooler is pretty lackluster when it comes to noise & temperatures... double wat

https://www.techpowerup.com/reviews/EVGA/GTX_1660_Ti_XC_Black/32.html

 

[smitty2k1]

Anandtech's 1660 review is out. Confirms it is a 120w part. It does run really power efficient though.
https://www.anandtech.com/show/14071/nvidia-gtx-1660-review-feat-evga-xc-gaming

 

[Legion]

https://www.techpowerup.com/253767/nvidia-geforce-gtx-1650-details-leak-thanks-to-eec-filing

Still no news on core count, pff !!!

 

[Legion]

First benchmark leak!




Source: https://hexus.net/tech/news/graphic...raphics-nears-amd-vega-8-perf-aots-benchmark/

 

[smitty2k1]

First benchmark leak!




Source: https://hexus.net/tech/news/graphic...raphics-nears-amd-vega-8-perf-aots-benchmark/

Nice find. I hope 1650 ti's launch at the same time... I want an upgrade from my 1050 ti

 

[VegetableStu]

1650 about the same as a 1050ti ._. if there'd be a 1650ti I'm roughly guessing it'll ballpark the 1060 3GB

my only hope is LP 1650ti

 

[NateDawg72]

There hasn't been any information on a 1650ti, so I just hope the 1650 isn't the full uncut die for TU117. If that's the full performance of the chip then it'll be disappointing compared to the 1660/ti which I've actually been impressed with. Too early to know though.

 

[smitty2k1]

There hasn't been any information on a 1650ti, so I just hope the 1650 isn't the full uncut die for TU117. If that's the full performance of the chip then it'll be disappointing compared to the 1660/ti which I've actually been impressed with. Too early to know though.

I guess I wouldn't be surprised - the 16xx series got a 1660ti, which was a model that didn't exist in the 10xx series. Could be Nvidia felt that performance gap was more important than the 1x50-1x60 gap this go-around. Here's hoping though!

 

[annasoh323]

Hopefully, this one data point isn't indicative of overall performance and this is on the low-end. It, unfortunately, wouldn't be unreasonable to expect that this is right about where the 1650 would land performance-wise. This generation has been more about moving each performance bracket one price bracket down (or up, as you may see it).

Rumors say Ampere will be unveiled later this year. Makes me wonder, though: If they've taken this long and are now just releasing the budget chips, I doubt we'll see anything in the low-power spectrum till much later. Doesn't bode well for my imagined mini-build for my brother. Oh well <wallet breathes a sigh of relief>

 

[VegetableStu]

same recent rumors of Ampere says it's a Volta replacement ._. so I'm dialing down my dialed-down expectations of the next GPU hardware release from Nvidia

 

[annasoh323]

same recent rumors of Ampere says it's a Volta replacement ._. so I'm dialing down my dialed-down expectations of the next GPU hardware release from Nvidia

Ah, pro market you mean? That makes sense. I didn't read too much into things - probably just a few moments scrolling past a headline before bed. Still, didn't Turing borrow from Volta to some degree? The way nVidia likes to curbstomp their competition into submission (this statement is not to be taken as a sign of admiration), I can't see them wanting AMD to be on 7nm by themselves for long.

 

[VegetableStu]

actually I just had a passing thought: how much of AMD's GPU predicament had to do with that weirdo contract with GoFlo? they had something about requiring to make X amount of stuff on their process tiers for a super long period of time

even at RX500 series they were making previous GCN stuff at 28nm, albeit for OEM purposes

 

[NateDawg72]

I think AMDs predicament is entirely down to not doing much R&D for architecture improvements or a new architecture. Most of their budget and talent seems to have gone to Zen. Nvidia has shown us several times in the last few years that they can improve power efficiency without a new process node, yet AMD has barely improved theirs. They have a 7nm GPU and it's still less efficient than their competitors ~3 year old gpus (gtx 1080 may 2016), it's a sad situation

 

[VegetableStu]

if it's just a R&D lull thing, I'm hoping Navi would be their breakout project ._.

 

[Legion]

It's the GCN arch they need to move away from, it's no longer scaling.
Arcturus (their next design architecture) is a shift away from this.

Navi is still GCN based and won't be anything significant to threaten Nvidia with, best we can hope for is more power efficiency and they will only be low to midrange parts (initially)

They did indeed focus all their R&D budget on Zen, now that is paying off for them.
We can only hope some of those profits go into the RTG design teams.
It's highly unlikely though we'll ever see AMD compete at top end of the market (again), I think they are happy selling mid range parts as that's where the profits are.
Top end gfx cards are less than 1% of gfx card sales!

 

[NateDawg72]

yeah, its true mid range has the most sales but as you said they need to move away from GCN - I don't believe they are making much profit with their current cards any longer. Their graphics revenue has been looking pretty weak iirc from their last quarterly report.

 

[VegetableStu]

oh dang, wasn't aware Navi was still on GCN o_o gonna adjust my notes a bit. thanks!

 

[Legion]

Another leak:

https://wccftech.com/nvidia-geforce-gtx-1650-gaming-benchmark-leak-launch-22-april/