[SOLVED] NVIDIA is working on a GTX 1650 Ti GPU, for an October launch !! *new info*

Page 3 - 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.
Hello,

One leak has just been published on an Asian website.

Rumour has it that Nvidia has plans to add a new entry into its Turing graphics card lineup, this time adding to the company's lower-end GTX range of GPU hardware.

Nvidia's next chip is rumoured to be called the GTX 1650 Ti, a graphics card which will sit between the GTX 1650 and GTX 1660, acting as a mid-way point between the two. The chip is rumoured to feature a CUDA core count of between 896 and 1408, enabling performance levels which could match or exceed Nvidia's last-generation GTX 1060 graphics card.

At this time, little is known about Nvidia's GTX 1650 Ti, though it is rumored to be released sometime in late September or early October.


This graphics card will fill a major gap in Nvidia's low-end GPU lineup, a space which is being exploited by low-cost RX 570 and RX 580 graphics cards from AMD.
To sit between Nvidia's GTX 1650 and GTX 1660, Nvidia's GTX 1650 Ti would need to feature between 1024 and 1280 CUDA cores. Nvidia graphics cards tend to offer CUDA cores in collections of 128, making 1024, 1152 and 1280 CUDA core counts possible. Ultimately, it depends on kind of silicon Nvidia plans to use to create its GTX 1650 Ti.

AMD is rumoured to be working on a low-end Navi graphics card to compete with Nvidia's GTX Turing offerings, making the creation of a GTX 1650 Ti a solid move for the green team. The GTX 1650 Ti could act as a counter to Radeon's low-end Navi graphics cards, or at a minimum solidify Nvidia's position in the low-end GPU market.

http://www.fashaoyou.net/Article/1543/94385.html

Thanks for reading. And translate the above website.
 
Solution
I think people started buying 2 cheap GPUs instead of 1 high end one, so AMD and Nvidia disabled multigpu support on lower-end GPUs.

personally i think that might not the reason. if user end up buying two GPU nvidia and AMD still end up selling more anyway. and those that buy low end or mid range GPU probably never dream of spending on buying high end GPU in the first place either. at the very least i think this will be the case for majority of people. the problem lies on this: will spending the money for two GPU will going to worth the performance it will bring despite the drawback of such system? ideally we want two low end GPU to outperform mid range GPU while cost lest to buy those two. same rule with mid range GPU. for the...
You can "SLl" 1060 using DX12 mGPU. but the real problem is no game developer would care to do it. Many games that support CF or SLI in their DX11 but when ported to DX12 game developer most often ignores mGPU support. Even those did it seems they just "port" nvidia/amd existing DX11 CF/SLI profile to DX12 instead of doing true DX12 mGPU optimization.
 
multi GPU need to evolved. GPU maker need to make it where the game can recognize all the GPU inside the system as one instead needing specific drivers or profile to "hack" multi GPU to work. but i think one of the primary problem will be the communication between GPU to GPU. PCIE the way it was made creating latency for the two. the primary reason why nvidia develop nvilink not because PCIE lack in bandwidth but because of the latency issue it pose to the work load they done in HPC class system.

nvidia to certain extend have successfully done this with their NVSwitch inside DGX-2 but it won't be usable for our regular consumer PC because the way how those NVSwitch being integrated with their GPU. also heard about how intel want to make multi GPU easy with their Xe. just plug in the second GPU onto your PCIE slot and voila games will automatically see the two GPU as one without needing special software to make it work. although to make it happen they need to integrate something proprietary with existing PCIE spec. meaning it might end up being only useable with intel CPU and chipset. for AMD i don't haven't heard any rumor how they want to improve mGPU for gaming usage yet. the one they have right now (connecting two GPU in IF) is more for professional usage rather than gaming. but this will end up being a headache for consumer if all three going with their own unique implementation for next gen mGPU hahaha.
 
I think people started buying 2 cheap GPUs instead of 1 high end one, so AMD and Nvidia disabled multigpu support on lower-end GPUs.

personally i think that might not the reason. if user end up buying two GPU nvidia and AMD still end up selling more anyway. and those that buy low end or mid range GPU probably never dream of spending on buying high end GPU in the first place either. at the very least i think this will be the case for majority of people. the problem lies on this: will spending the money for two GPU will going to worth the performance it will bring despite the drawback of such system? ideally we want two low end GPU to outperform mid range GPU while cost lest to buy those two. same rule with mid range GPU. for the most part joining two low end GPU most often will not going to make it faster than a midrange GPU. for this reason it is almost no good reason to SLI or CF low end GPU. but for mid range GPU it is possible to make that "miracle" happen. GTX460 in SLI is one such example:

https://www.techpowerup.com/review/nvidia-geforce-gtx-460-sli/25.html

even without perfect scaling GTX460 in SLI still outperforming a single GTX480 with noticeable lead. GTX660 in SLI was also like this. faster than GTX680 while cost less. the appeal for multi GPU will be gone if they cannot do this. this is the case with GTX960 SLI.

https://www.techpowerup.com/review/nvidia-geforce-gtx-960-sli/23.html

despite being cheaper it cannot beat GTX980. GTX960 can't beat GTX980 in SLI because of it's CUDA core configuration. just to match GTX980 performance GTX960 in SLI will need 100% scaling which is pretty much impossible to happen let alone beat GTX980 in performance. they say nvidia disabled SLI on GTX1060 because nvidia afraid GTX1060 SLI will end up eating into GTX1080 sales. for me that will never happen. because the situation on GTX1060 is exactly similar to GTX960. it needs 100% scaling just to match GTX 1080 performance. many reviewer end up not recommending GTX960 because of it's scaling performance. nvidia know this should also happen with GTX1060 hence they cut support for SLI on 1060 to further save the budget.

alternately you can also said nvidia configured their mid range GPU in such way so it will not going to eat into their high end GPU sales when two mid range GPU being paired together. but i still stands behind the notion that SLI being killed in the mid range (for nvidia case) not because it will eat into high end GPU sales but more because of such setup will not going to be popular to begin with. not to mention with 8th gen generation console many games has updated their game engine with new rendering method that most often did not like how AFR work. this also slowly killing multi GPU even if mid range GPU able to out perform high end GPU. in the end the investment does not worth it when SLI/CF simply not working in majority of new games even if they can beat high end GPU in certain games. TPU when doing the test for RX480 in CF high light that among the 16 games that being tested CF only working in 6 of them! from TPU RX480 CF conclusion:

We averaged games that do take advantage of CrossFire in a separate relative performance data-point than the overall relative-performance. The findings are interesting; when averaged among games that do scale, the RX 480 CrossFire is about 5-10% faster than a GTX 1070. However, only 6 out of 16 tests are taking advantage of the second card.

If you only have money for a single card now, and you want to buy a second card later, you're in for more than playable frame-rates at 2560x1440 resolution in games that do scale, and even 4K performance with eye-candy watered down a little. RX 480 CrossFire loses out on its overall relative performance big time due to the number of games that don't scale well (6 out of 16).
 
  • Like
Reactions: Solidjake
Solution
Had they supported the feature, 1060 SLI could have potentially been an option some might have found worth considering when the cards first launched, when a 1080 cost nearly three times as much as a 1060, which could have allowed the scaling issues in many games to be more easily overlooked. That option would have made little sense as soon as Nvidia launched the 1080 Ti and dropped the MSRP of the 1080 by a couple hundred dollars though. And at this point, you might as well just go for a single 2060 for a roughly similar level of performance with additional hardware features.
 
that might be interesting value and performance wise but the primary speculation is nvidia disabled SLI on GTX 1060 because they afraid GTX1060 SLI would easily beat GTX1080 performance (like many x60 cards in the past). but in the end i still think many would avoid 1060SLI due to many modern games did not like how AFR work. nvidia should be aware about this hence they slowly take away SLI support. those that still hoping multi GPU to keep on living probably banking the future of multi GPU on AMD shoulders seeing how only AMD still care to make dual GPU cards. but in the end AMD decided to kill CF even faster than nvidia and hope game developer will carry the mantle instead by only supporting DX12 mGPU going forward with their newest generation of GPU.
 
If I had to guess, AMD will likely announce a faster GPU early next year. The Vega-based Radeon VII was made kind of redundant (at least for gaming) by the 5700 XT performing not far behind for a few hundred dollars less, and I would expect them to have a Navi-based successor to it before long. I wouldn't be surprised if a 300 watt Navi card were to perform roughly similar to a 2080 Ti. Supposedly, AMD also has raytracing acceleration in the works, which could potentially also make an appearance on such a card.
 
most likely to fend of AMD RX5500 and RX5600. this segment is very crowded. instead of introducing new cards nvidia should just drop existing price. but for nvidia to drop price....they would rather coming out with new variant of cards with cheaper price than cut existing product price because doing direct price cut shows that they admit the heating competition to the public.

TBH at this point i just want to hear about Turing successor.
 

WildCard999

Titan
Moderator
Two MSI GTX 1660 Super cards have been spotted.
https://www.tomshardware.com/news/two-of-msis-gtx-1660-super-spotted-in-the-wild

If the pricing ends up around $300 I may need to consider upgrading to that to replace my RX 580 4gb. I've considered the 5700 however drivers have been a bit wonky and can't take risk with an old game I've been playing a lot (City of Heroes: Homecoming, there's no new drivers for it past 2011).
 
If the pricing ends up around $300 I may need to consider upgrading to that to replace my RX 580 4gb.
If the leaked specs are to be believed, I get the distinct impression that the 1660 Super will be a replacement for the current 1660, and will only perform between that and the 1660 Ti, probably with with an MSRP in the $230-$250 range. If you want something faster than a 1660 Ti, you will probably need to wait longer for a possible 1660 Ti Super, or just go with an RTX 2060. Or whatever AMD launches around this performance range, assuming they get around to that sometime in the near-future.
 
no pricing yet but nvidia most likely want to force AMD to launch their RX5500 hence late november launch instead of launching together with 1660 super like how it was expected before. i think AMD also waiting for 1650 Super before they set final price for RX5500 hence they never talk about price until now. now nvidia wiling to delay 1650 until late november. will AMD wiling to wait that long or just release their RX5500 and just get the sales? if crazy things end up happening this might end up being drag to december haha
 
Status
Not open for further replies.