_mockingbird
[H]ard|Gawd
- Joined
- Feb 20, 2017
- Messages
- 1,066
It's laughable at best to think that these "optimizations" that Radeon RX Vega supposedly have is going to turn the table.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Vega Rumors
- Thread starter grtitan
- Start date
It's laughable at best to think that these "optimizations" that Radeon RX Vega supposedly have is going to turn the table.
shad0w4life
Gawd
- Joined
- Jun 30, 2008
- Messages
- 689
Hey, it beats having to try and get Nvidias drivers working properly and then having to use their piss poor interface, and try and fix error code 43 in device manager.
I own 4x 1070, a 1060 and a 1050ti
Not sure why the freak out about the 500w total system power consumption 1080FE was around 330W Stock OC'ed 388W in Crysis 3
talking about system or card?
Cause 1080 fe can't do 330watts without voltage mods, just not possible. Not only that doesn't have the pci-e connectors to do that.
With Vega FE we are talking about drawing 500 watts for the card.
Wow that is atrocious, 400 watts from the two 8 pins connectors.
lolfail9001
[H]ard|Gawd
- Joined
- May 31, 2016
- Messages
- 1,496
System, i think.
Is it the highest for single GPU on ambient? I don't think i remember numbers that high ever.
The last time something got that hot with two big wires plugged into it your mother was involved and it sold 25,000 copies.
No I don't recall a single GPU card that ate that much juice either. Even the dual radeons mostly behaved.
We should just call it Fermi II.
Magic Hate Ball
Limp Gawd
- Joined
- Jul 28, 2016
- Messages
- 366
Vermi?
http://imgur.com/a/ImPz1
some stuff worth noting here the HMB stacks are OVERVOLTED on the stock Vega FE card... seems Vega FE was clocked as high as AMD could get it to start with with out pulling stupid power
some stuff worth noting here the HMB stacks are OVERVOLTED on the stock Vega FE card... seems Vega FE was clocked as high as AMD could get it to start with with out pulling stupid power
- Joined
- Jun 19, 2003
- Messages
- 9,100
All I know is if anyone who complains about AMD new vega using a ton of power (and no joke about it, it does), better not be using a new X299 Skylake-x or Kaby Lake-x.....lol
Dayaks
[H]F Junkie
- Joined
- Feb 22, 2012
- Messages
- 10,289
That's different since the extra power creates useful output above and beyond what is available from the competitor. Like the 7900x uses 50% more power but is 40% faster vs. a 1800x (for cinebench that PCPER tested).... Vega is 0% faster for 200% more power. See the difference?
- Joined
- Jun 19, 2003
- Messages
- 9,100
of course I know the difference. I am talking about Watt's at the wall. Not what kind of performance there is compare to the watts used. AMD Vega's are power hungry just like Intel's new CPU's. Not once did I mention performance.
http://www.gamersnexus.net/guides/2986-amd-vega-hybrid-mod-results-overclocking-liquid-vs-air
Now I understand why BZ said GN hit the silicon lottery, air cooled they were able to get their card to 1660 mhz, which is 30 mhz higher then what BZ or anyone else seemed to have gotten with the FE.
Interesting that the card looks to be thermal limited at 85 c.
Good example of 30 degrees cooler 30 less watts being used too, directly correlating with leakage amounts as well.
Now I understand why BZ said GN hit the silicon lottery, air cooled they were able to get their card to 1660 mhz, which is 30 mhz higher then what BZ or anyone else seemed to have gotten with the FE.
Interesting that the card looks to be thermal limited at 85 c.
Good example of 30 degrees cooler 30 less watts being used too, directly correlating with leakage amounts as well.
Dayaks
[H]F Junkie
- Joined
- Feb 22, 2012
- Messages
- 10,289
.... I don't get that logic to be honest. At least compare apples to apples. Like a 7700k vs a 1800x or something.
We're talking about wattage at certain performance levels and how AMD is not even close. You could pull 400 watts each for Titan X Maxwell or a 5960x ... but the performance was there.
Last edited:
Gideon
Supreme [H]ardness
- Joined
- Apr 13, 2006
- Messages
- 4,219
Not really surprising Pascal reacts pretty much the same. Cooler you can keep it the better you can overclock it. Pascal tends to become unstable the hotter it gets, I had to tweak some stuff before I got mine stable at 2100 MHz. Just that Pascal starts at a lower wattage then Vega does right now. Now Pascal wont let you overclock beyond a certain TDP and that is the main reason water cooling it is kind of a waste so Nvidia put a hard limit on it. Couple more weeks and will see the gaming version and will have a much better picture of what Vega can do.
Does sound like the FE is one stout design for power or a well built card if it can handle over 400w. Gaming performance for that power usage is terrible. Other usage maybe it is OK, now who here will be using the FP16 capability? That might be the only area where it is untouched but in the scheme of things not used too much in general.
- Joined
- Dec 2, 2004
- Messages
- 32,855
I don't! I find the egregious spelling and grammar mistakes to be really sad.
Anarchist4000
[H]ard|Gawd
- Joined
- Jun 10, 2001
- Messages
- 1,659
FP16 and even INT8 look like they may be seeing a lot of use. Still need confirmation, but they may be the programmable texture filtering and blending often buried in TMUs and ROPs. So any HDR/Freesync2 title should be using it. That wouldn't account for particles, physics, sound, etc.
- Joined
- Dec 9, 2000
- Messages
- 16,282
Anarchist4000
[H]ard|Gawd
- Joined
- Jun 10, 2001
- Messages
- 1,659
Freesync2 was the HDR version that incorporated tone mapping for less latency. So quick FP16 conversions to the displays native color space.What does freesync have to do with any of that?
We all know vega fe uses a lot of power. We have to see where RX vega lands. Watercooled seems to be the one at 375w and pushed to limits. RX Vega XTX, other two are rated at 285 or 275 I believe of total board power and 220 for the chip itself. So AMD seems to have tweaked the gaming cards little better then FE. Now we just have to see where performance lands on rx vega.
its the same die for RX... im not sure what you are expecting from it
and i 100% can guaranty you that the lower TDP RX cards are ether lower clocked or have disabled compute cores or both
prediction right now
XTX = to FE or slightly slower
XT = or lower then 1070
XL = between 1060 and 1070
and i 100% can guaranty you that the lower TDP RX cards are ether lower clocked or have disabled compute cores or both
prediction right now
XTX = to FE or slightly slower
XT = or lower then 1070
XL = between 1060 and 1070
Dude you know everything. Like you knew about it using 500 watts not knowing that it only uses 25 from PCI-e and then stopped responding to my post when I gave you a source proving it. Don't ever guarantee anything 100% if you don't have facts to back it up. I am going by AMDs word that rx vega will be more optimized and I expect it to not blow the world out of water. Lets say its 10% better worst case it will be faster then FE, anything else? You keep assuming I expect vega to be a miracle. I don't expect shit from it. I am giving benefit of the doubt to AMD who repeatedly said it rx vega will be much better optimzed for latest titles and will be faster then FE. Just look at your conclusion.
I think it will likely be this.
XTX in between 1080 - 1080 ti
XT equivalent to 1080
xl equivalent to 1070
Looking at the power usage it looks like XTX will be highly clocked and if it is little better optimized for games even 10% you are looking at faster than 1080. No I don't guarantee this 100% because I don't have the cards and I could be wrong. But I suspect this will be case.
funkydmunky
2[H]4U
- Joined
- Aug 28, 2008
- Messages
- 4,016
It is beyond obvious where they will land. 1070, 1080, and slightly above 1080, all at better price points and higher power usage then Nvidia.
Kinda how it has been for how many generations now?
I mean their own marketing slides show 2016 Polaris for Mainstream. 2017 Vega for Enthusiast. 2018 Navi for Flag-ship.
AMD did not have the $ for an all out GPU assault, especially with CPU being so important to them. Unlike Polaris, Vega is brand new. Their first crack at this architecture and it will have teething issues for sure. Will it improve over time? I would bet on it. Will it be fan-boi super sauce for the win? Nope.
Will it be competitive for the price? Of course.
Kinda how it has been for how many generations now?
I mean their own marketing slides show 2016 Polaris for Mainstream. 2017 Vega for Enthusiast. 2018 Navi for Flag-ship.
AMD did not have the $ for an all out GPU assault, especially with CPU being so important to them. Unlike Polaris, Vega is brand new. Their first crack at this architecture and it will have teething issues for sure. Will it improve over time? I would bet on it. Will it be fan-boi super sauce for the win? Nope.
Will it be competitive for the price? Of course.
Couldn't have said it better. AMD will use more watts for now. They hit the nail on the head with ryzen and its power efficiency. Its clear where the R&D went for last few years and GPUs took the back seat. Navi will probably be much more power efficient. Vega will be more power hungry but will likely have pretty decent price/performance ratio but will use more power. Thats RTG for you now.
- Joined
- Aug 5, 2013
- Messages
- 13,461
idk about that. Fury X was an abject failure. If RX Vega is a repeat of that (vs the 1080 or 1080 Ti) it will also be a failure but with even higher power usage, apparently.
If miners don't buy these cards then they might be competitive in the current market, since everything else is jacked up.
funkydmunky
2[H]4U
- Joined
- Aug 28, 2008
- Messages
- 4,016
"IF"
Good argument.
- Joined
- Aug 5, 2013
- Messages
- 13,461
"How it's been for how many generations now"... meaning not Fury X? It was their previous flagship from the most recent generation. Vega shares a lot of similarities with Fiji architecturally, and HBM. If you're looking for a gen to compare Vega with, Fiji is the best choice and that's not good news.
Otherwise you should be more specific about which generation you expect Vega to take after. AKA cherry pick.
Last edited:
I don't care too much about power consumption per se but higher power consumption = higher heat output, which I don't want. My study gets hot enough in the summer already.
If Vega IS indeed 1080ish in performance, and Fury X can be considered a 'failure', then Vega is a disaster.
Fury X at least had no problems beating 980 and was up there with 980ti, and was released within weeks of 980ti.
Vega is 'down' there with 1080, and is released months behind 1080ti, with Volta looming on the horizon.
EDIT: Even though Vega hasn't been truly released to the gamers yet, I wonder if it'd be better if AMD just completely scrap Vega and speed up Navi's production instead.
Fury X at least had no problems beating 980 and was up there with 980ti, and was released within weeks of 980ti.
Vega is 'down' there with 1080, and is released months behind 1080ti, with Volta looming on the horizon.
EDIT: Even though Vega hasn't been truly released to the gamers yet, I wonder if it'd be better if AMD just completely scrap Vega and speed up Navi's production instead.
its not stock. I don't understand why you are all are going crazy over a modded card that is being pushed to its limits. 400 watts + 25 from PCI-e so 425 but stock it does use less than 300. Still its obvious AMD will lag behind nvidia in power/performance and will do for the near future. No way they match nvidia design and r&d for now. May be with Navi, but even if they don't I wouldn't be surprised. Costs money to build a good design which the GPU division is lacking with AMDs heavy focus on zen.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
All Furys sold out, so not a failure at all. May not have out performed Nvidias top performer but it did hang with it.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
Saw this, if only vega would make this accurate...
- Joined
- Feb 11, 2013
- Messages
- 6,478
Vega isn't new design as it wasn't polaris, as it wasn't Fiji, as it wasn't Tonga as it wasn't hawaii, the base architecture it's basically the same 5.5years old GCN. AMD launched "Mainstream Tonga" before Fiji to have a time to refine drivers, same with "Mainstream Polaris" to VEGA.. Fiji was already a very refined GCN part with a lot of their weakness against nvidia treated, limited by that consumer useless HBM, most of the performance gained in Fiji vs Hawaii was due brute forcing more Cores going from 2816 to 4096, VEGA it's a refinement of that with more efficient geometry performance, and as they can't add more cores all the gains comes from the Extra Core speed..
technically, yes. boost clock is never guaranteed, but now AMD is playing a new card to avoid their impossibility to reach boost clock calling it "Peak Engine clock" and "Typical Engine Clock", nice marketing move but useless.. so, base clock 1382mhz and max boost 1600mhz (unreachable typically according marketing words)..
This launch is much worse than Fury X, we are talking about a power differential to the ti model of 125watts which is 50% differential vs 10% differential and the performance is behind by much more.
PS this is the air cooled top end version, the water cooling version won't give much more performance either its more like a trophy wife that isn't even a trophy.
It was a failure, they didn't make money on the Fury line.
JustReason
razor1 is my Lover
- Joined
- Oct 31, 2015
- Messages
- 2,483
proof?