kirankara78
Member
Does it beat Vega FE at prof workloads?
Does Vega FE cost same as a 1080ti?
Titan XP is similar however and is superior for gaming by far and is after latest update much improved in professional performance
Does it beat Vega FE at prof workloads?
Somewhat burying the flagship, bold strategy. Let's hope the 56 is actually competitive with the 1070...
Somewhat burying the flagship, bold strategy. Let's hope the 56 is actually competitive with the 1070...
This was the case with Fury and Fury X too. And possibly 470/480 to a certain extent.
It isn't much more expensive, but I"m not sure why it matters.Does Vega FE cost same as a 1080ti?
It isn't much more expensive, but I"m not sure why it matters.
RIP remark would make sense, in my opinion, if AMD would not be able to even remotely match the performance (in prof workloads). If it actually can and even is faster, then the comment is misplaced.
But then, I don't see why a gamer would go with ThreadRipper to begin with, perhaps, there are some reasons.
This is pretty good![]()
Damn rip Vega
Somewhat burying the flagship, bold strategy. Let's hope the 56 is actually competitive with the 1070...
There were fewer RX 470's in circulation too, even moreso the 470 8GB....RX 480 was a more popular card regardless....The 480 4GB was a fine card, and at $200 a bargain. Such a bargain AMD did their best to produce basically 0. Man what a flubbed launch. I don't remember them plugging the 470 at launch much compared to the "$200 for the 480!" which you couldn't really buy...
There were fewer RX 470's in circulation too, even moreso the 470 8GB....RX 480 was a more popular card regardless....
AMD might try to push Vega 56 and that may sell well as card that outperforms the 1070 at $400.00, but lots of people will buy the flagship 64 cards, especially the OC'd AIB ones...
Well, at least it will be nice to see some Vega 56 benches on the 14th and see how it does against the 1070....Also how much perf improvement we have with the 64 over the 56...Should be very interesting...
That's my dream, slightly better than the 1070 at a good price. I just want to unlock Freesync and have a stronger card to hit 144fps in more games (or at least 100-120).
1924mhz is not a stock clock on any 1080 that is close to msrp. You're comparing a premium 3rd party card or an overclocked card (yes I get that is a clock basically any 3rd party card can hit) to a blower style reference Vega and then claiming the Vegas overclock speeds are invalid. Compare stock to stock and overclock to overclock there's no reason to believe that 1750mhz won't be as normal of a clock for 3rd party Vega cards as 1924mhz is for 3rd party Nvidia cards.
Here's the clock graph for a 1080 FE:
![]()
Note that it basically starts at 1900 MHz - it's highly likely that even this "stock" FE would report 1925 MHz as its max boost clocks to the 3DMark monitoring module. Maybe a step lower, around 1911. In any case, 1925 is basically stock "max" clock for a GTX 1080.
1750 MHz max is for RX Vega 64 WC model so you could say that I'm actually comparing a much cheaper AC factory OC 1080 to a water cooled Vega there. And you're completely ignoring the lower range of Vega results in these graphs again which clearly hint at top results actually being for the OCed WC model.
A "normal" clock for a WCed OC GTX 1080 is ~2100-2150 MHz. Such card will likely be faster than any of the Vega results in these graphs.
The sole fact that AMD has moved forward Vega 56 launch and told the press to prioritize that over Vega 64 should already tell you all you need to know about Vega 64 performance - it simply won't beat its (more than a year old) competition to any significant degree at the same price.
Vega 56 should fare a lot better against the 1070 and is likely to end up being the only Vega card worth buying.
The red line below shows you the default clock speed for comparison. The lowest clock speed of the overclock spiked down to 1873MHz very briefly, but otherwise the average new clock speed is 2062MHz versus 1770MHz average clock speed at default operation.
A card running at an average clock speed of 1924 MHZ is running much higher than an average boost clock of 1770MHZ..
Looks like Vega 56 isn't launching till August 28th, two weeks after Vega 64.
http://www.gamersnexus.net/news-pc/3016-amd-moves-vega-56-embargo-forward-prioritizes-over-64
No, every Vega card is getting the AIB treatment.I thought there are only going to be reference Vega 64 cards and AIB partners are only getting Vega 56 chips to play with?
Gonna try and snare one on launch day as well. But from here in UK that might be impossible.
No, every Vega card is getting the AIB treatment.
https://evonews.com/tech-science/itc/2017/aug/10/amd-radeon-rx-vega-64-finally-has-aib-partners/
http://wccftech.com/asus-rog-strix-rx-vega-64-custom-aib-graphics-card-launch-september/
The article you pulled that from
Not accurate to say most 3rd party 1080s hit 2.1 - 2.15ghz...hitting 2.1 is pretty tough even with the high end ones. I have one and generally 2088 is best I do. 2.1 is the wall.
Okay, let's try this for the third and final time - 3DMark monitoring module logs _maximum clocks_ a card reports. This is why you see only 1630 and 1750 MHz for all RX Vega results despite said results being very different. 1925 MHz can easily be a maximum clock reported by a stock GTX 1080. A properly OCed GTX 1080 would report clocks starting from 2000 MHz there.
![]()
![]()
![]()
Availability
The new Vega 64 based SAPPHIRE graphics cards will be available to buy from August 14th.
For full product information, head to the SAPPHIRE Technology website.
Sapphire teasing some non-customized card release
https://sapphirenation.net/sapphire...etter&utm_medium=email&utm_campaign=YoNAugust
The first image I've seen before I think (wasn't it a slide from an AMD Vega conference?) but the second seems new. There's more at the link but it's mostly PR fluff
I think all of these numbers are from SIGGRAPH
So Vega has 70% higher clockspeed than Fury X, but only 30% more performance
I just don't get it
Are all Gpu launches this stressful?
I think all of these numbers are from SIGGRAPH
So Vega has 70% higher clockspeed than Fury X, but only 30% more performance
I just don't get it
I've been following the PC gaming scene for a few weeks after getting out of it awhile ago. I'm pretty much up to speed now on all the new tech and all the rumors of tech coming out. I got the green light from my wife to replace my 2014 MacBook Pro with a gaming pc, but there is so much happening right now it doesn't seem like a good time to jump in new.
Inflated graphics prices, inflated memory costs, new intel chips at the end of the month all make me want to wait til the beginning of 2018 when new Ryzen CPUs and possibly new nvidia cards will be out... I'm thinking of just getting a mechanical keyboard and make due with my Xbox one and PS4 until next year... or if I'm weak I may wait until Black Friday deals to pickup a monitor or possible my whole rig. Really excited to see the numbers on Vega on Monday.
This was the case with Fury and Fury X too. And possibly 470/480 to a certain extent.
To be honest, is a smart move. They will sell more of them as bigger market than for flagship cards. It seems 56 will deliver a far more competitive price/ performance ratio
I've never bought a GPU at launch before.
What is my best bet for buying a Vega 64 liquid at launch?
Can I just buy one on Amazon? When will it go up?
Any news about the Vega Nano? I'm looking to see what would be a reasonable upgrade (8-pin ITX GPU) from my "Fury" Nano be like. I was initially thinking that the 1070 itx would be the premium choice, but, it turns out the appearance of a 1080 itx changed that. A 1080 Ti would've made for a great choice, too, if not for the power draw (requires 2 8-pin connectors), heat (test show the card hitting thottling temps, unlike its non-Ti cousin), noise (unlike its non-Ti cousin), and price ($600~ is waaaay too much for me). I'm hoping Vega Nano would at least be competitive in the performance/$ ratio, even if it ends up in-between a 1070 and 1080.
Mistakes were made
Pcper accidental live testing
![]()
My 1080ti gaming x cost £650 and performs much better than FE which is £1000.
That's a big difference
But yes original comment was nonsense
Except these are huge die chips competing with NV mid sized dies. NV is killing amd on margins. That's why nvs shareholders are pretty happy and amds aren't. You can't really spin this as any sort of win for amd.
Unfortunately trying to increase clock speed potential can often be detrimental to per-clock performance. The most famous example of this is the NetBurst architecture in Pentium 4.
AdoredTV ran it with 750W in the most power consuming mode, claims any decent 650W should do.Jesus Christ, the liquid cooled version demands a minimum of a 1,000W power supply..
First leaks of RX Vega Air gaming benches (295W TDP):
![]()
https://www.reddit.com/r/Amd/comments/6tfzdm/vega_leaked_benchmarks_performance_seems_ok/
AdoredTV ran it with 750W in the most power consuming mode, claims any decent 650W should do.
First leaks of RX Vega Air gaming benches (295W TDP):
![]()
https://www.reddit.com/r/Amd/comments/6tfzdm/vega_leaked_benchmarks_performance_seems_ok/
should do.
Hard to take much from that graph without knowing what games/resolutions were being tested, but that would be crazy if the performance gap between the 56 and the 64 is really that slim.
First leaks of RX Vega Air gaming benches (295W TDP):
![]()
https://www.reddit.com/r/Amd/comments/6tfzdm/vega_leaked_benchmarks_performance_seems_ok/