I don't understand people saying that the 1080ti is overpriced for what it is. It's one of the fastest cards on the market and from what I've heard on these forums people are more than willing to pay what's being asked. After all something's only worth what people are willing to pay. Back on topic though, I'm really looking forward to seeing how Vega will perform, I'm due to upgrade my 290x
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD VEGA confirmed for 2017 H1
- Thread starter Doom112
- Start date
- Status
And that is where nvidia are genius with their pricing of the halo product
If there was no titan product, I be willing to bet that most people would call the ti a rip off, if the titan was priced even higher, people would think that the ti is an even greater bang per buck buy....
Caporegime
And that is where nvidia are genius with their pricing of the halo product
If there was no titan product, I be willing to bet that most people would call the ti a rip off, if the titan was priced even higher, people would think that the ti is an even greater bang per buck buy....
+1
I would guess most people conclude that because it has no competition it is more expensive than it would normally be in a functioning market. It will be interesting to see what happens to the price if Vega comes near it and manages to be sold for less.
Last time I read TomsHardware as the main source of information was in fact release of first Pentium 4 3.06 HyperThreading CPUDon't trust a thing Toms Hardware say or even do, they are fakes, shills.
After that other review sites provided a better picture I guess.
Permabanned
Navi will be scaleable, it's the way forward. Called it ages ago. https://forums.overclockers.co.uk/t...arly-as-2h-2017.18742003/page-5#post-29829887
Permabanned
Lol. At least you are honest
I would tell you that's my middle name but that'd be a lie.
Last edited:
AFAIK or understood. Navi wont work as your standard xfire setup. It will be like crossfire but also very different. Where xfire today requires sending information over the PCI_E Bus to the other card/GPU which in essence requires the CPU to be involved to send this data where Navi wont. Navi won't work in the same way. Currently it requires Devs to program for it but Devs will only see Navi as one GPU making their life non the harder. AMD will have to do all the work to ensure Navi scales well and works well. Not the devs. The interconnect aka Infinity Fabric replaces the PCI_E Bus which also cuts out needing the CPU to send data to the other GPU though Navi won't really be like 2 gpus but clusters. The clusters will share data over a cache... ahhh who the flippin hell am i kidding. I don't totally understand it all tbh lol. But from What Raja said (leading GPU Architect) it will act as one GPU. So as long as AMD don't drop the ball with their drivers and micro code/bios then Navi will be a really cost effective way for AMD to make a big GPU(with clusters) that acts just like a single die GPU.
However we might end up going down the same road. On paper Navi looks great for example could hit over 20Tflops of single precision compute but in reality performance does not seem to compare because drivers need to catch up. Fiji / Furyx is a cracker of a card which was limited to 4Gb of memory but early days seemed like a noob but now hangs with the big boys now drivers are there.
However we might end up going down the same road. On paper Navi looks great for example could hit over 20Tflops of single precision compute but in reality performance does not seem to compare because drivers need to catch up. Fiji / Furyx is a cracker of a card which was limited to 4Gb of memory but early days seemed like a noob but now hangs with the big boys now drivers are there.
And that is where nvidia are genius with their pricing of the halo product
If there was no titan product, I be willing to bet that most people would call the ti a rip off, if the titan was priced even higher, people would think that the ti is an even greater bang per buck buy....
Yep and why nVidia release the Titan first. Ohh look Titan the single most fastest card. Dammm that price though. Blumin rip off. GTX xx80Ti comes along and at half the price or less. Ohh look it's same performance or slightly faster than the Titan but costs a lot less than a Titan. I can get Titan performance for a lot less. I'm getin that card
Reality is nVidia never intended you to buy the Titan/same card for the higher price but intended the titan to act as a way to hike up the price to make the 80Ti look more appealing. And it works
TBH i hate my self for buying the 1080Ti but just looking at what sort of performance i could get at 1440p just grabbed me and i found it hard to wait around. Even after buying one i think it's overpriced and have buyers remorse but never the less love the performance i get in games.
Soldato
It's only overpriced if you can get better for less but from a historical sense I know why people would say such a thing as a lot of enthusiasts know that the 1080ti should only be a mid/upper card would sell in the 300-400 range. But due to the lack of competition as a result AMD's never ending ability to make the wrong calls on which way he industry is heading (with technology such as GNC and Bulldozer, all ways easier with hindsight but still it shows just how costly it can be when this goes wrong) Nvidia can take all their products and shift them all up into the next price bracket.
I'm really skeptical about Vega but I feel that is the right approach as AMD have underwhelmed me with every GPU product lunch since the 7970 so I don't know why I should expect anything different this time around.
Last edited:
Caporegime
I think Loadsamoney nearly fell for the same trap, but realised he would end up hating himself just in time to cancel his 1080Ti "pre-order"
Soldato
- Joined
- 30 Nov 2011
- Posts
- 11,376
And that is where nvidia are genius with their pricing of the halo product
If there was no titan product, I be willing to bet that most people would call the ti a rip off, if the titan was priced even higher, people would think that the ti is an even greater bang per buck buy....
The gtx8800 was $650, 10 years ago, adjusted for inflation thats $766
The 8800ultra launched at $830, $978 adjusted for inflation
Doesnt look to have changed all that much to me.
Even the geforce 2 ultra was $499, $708 adjusted, but look at it and what you get for the money now.
One more thing. The upgrade to the current gen consoles both come with dual GPU configurations. So i think we will see more game engines that are better at using multiple graphics cards.
GCN wasn't a wrong call, it was too early and under utilised by AMD.
GCN wasn't a wrong call, it was too early and under utilised by AMD.
So just like everything else they do then.
AMD are being taken for mugs by Intel and nVidia; they let AMD invent all the actually interesting and forward-looking tech (because AMD has no choice; it can't compete on the "current" playing field), then wait for people to actually start using it in 3-5 years. In the mean time they just keep going on their relatively cheap incremental upgrades and when the new tech is sufficiently widespread they implement their own version of it when it's much cheaper to do so. AMD, to their credit, at least open source stuff they invent but, again, this is pretty much out of necessity. DX12/Vulkan would not be where it is today if AMD had kept it proprietary because no-one would be buying their GPUs to use a feature not implemented by any devs. nVidia and Intel of course don't have this problem: they have the market share so people will use what they put out no matter what, hence G-Sync and things like restricting 6+ core CPUs and overclockable CPUs with VT-d to a more expensive platform.
Soldato
Your not comparing like for like. Nvidia's top end consumer GPU retails at $1,200.
Soldato
How was it not the wrong call if they released it to early? They messed up and their so far behind the curve now I can't see how they will ever catch up.
Soldato
- Joined
- 28 May 2007
- Posts
- 10,067
Vega will put them right back where they need to be on the Curve. It's not very hard to see that. Look at Ryzen. Did you ever think AMD would catch back up with Intel like they have. They are doing that with a very small budget in comparison using an inferior process to manufacture it on. The only reason it looks like AMD are so far behind is the stupidly long wait for Vega and it looks like that's mainly down to HBM2. With what AMD have on the market now yea it looks like they are way behind but it's not really the reality as as soon as Vega releases they have caught right back up. Volta might give Nvidia a leg up again in 2018 but AMD also have stuff coming out then as well.
Unfortunately Yes, hopefully with a dedicated radeon group they will get better utilisation of there technology.
I read your post as the technology itself being a wrong call/bad. Bulldozer was bad, GCN wasn't and my post is wrong it wasn't early it was under utilised by AMD. Imagine if AMD had released vulkan back when the first GCN cards came out. They potentially would have been in a much stronger position than now. In fact I think if AMD had the challenged CUDA with some sort of software built on OpenCL that fully utilised GCN they potentially could have beaten Nvidia in the HPC market.
As above, AMDs problem is the lack of staff they have to get there technology working in the real world which is down to having no money.
And wasn't Intel's lead unreachable?
Intel NetBurst/Pentium 4 vs Athlon 64s...
With stock cooler some Pentium 4s were thermal throttling even at idle!
"Dust buster/leaf blower" but still hot GeForce FX vs. Radeon 9700 Pro running original Far Cry intended to be Nvidia's DX9 showcase lot better?
Heating Fermistor vs HD5870?
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_480_Fermi/33.html
- Status