Vega/Navi Rumors (Updated)

[Krteq]

I wouldn't put too much faith in those specs, like mentioned the TFLops is wrong and it says it doesn't support Freesync

What exactly is wrong on those TFLOPS numbers? And where is said that it doesn't support FreeSync?

//Those specs are in line with records in CompuBench DB

 

[Magic Hate Ball]

Or they had to push the clocks up as high as they could to compete with 1080Ti, wrecking their perf/watt in the process. It wouldn't be the first time.

Eh, it's only a 4.5% increase. I wouldn't read too much into it.

 

[Snarf Snarf]

Or they had to push the clocks up as high as they could to compete with 1080Ti, wrecking their perf/watt in the process. It wouldn't be the first time.

This is a professional card, I doubt that AMD clocked this thing to its ceiling to compete with a gaming card. 1600 MHz is likely well within the efficiency range of the card due to the fact that it's being cooled with a blower style cooler which we know to be less than ideal for max clocks. If it was AIO only at those clocks, then I'd be more inclined to think that was the max clock the card could hit like Fury X.

 

[Bacon1]

What exactly is wrong on those TFLOPS numbers? And where is said that it doesn't support FreeSync?

https://www.scan.co.uk/products/16g...96-streams-262-fp16-tflops-131-fp32-tflops-3x

Has an X for Freesync support

I misread the HBM speed as core speed

 

[LaikaSpaceCat]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

 

[cytg111]

So .. ~twice as fast as rx480.. That doesnt really catch a TI does it?

 

[zinfamous]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

I don't like that. I hope there's a more efficient nano version.

 

[cytg111]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

As long as it doesnt melt my mobo I dont care... If anything it would be the idle wattage.

 

[thilanliyan]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

Yeah that's disappointing if true.

 

[Dygaza]

For pure performance, I just hope Vega cards will have atleast some overclocking capacity left. Polaris cards atleast have some room.

 

[Bacon1]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

https://exxactcorp.com/index.php/product/prod_detail/2566

Is the WC version

But why would they have the exact same specs but use 75 watts more?

Says the perf is for the air cooled, which means the WC should have higher TFLops.

 

[sandorski]

https://exxactcorp.com/index.php/product/prod_detail/2565

300W for air cooled, 375W for water cooled.

amd
wat r u doing
amd
stahp

More Performance with a relatively small increase in Power consumption. Not the end of the world.

 

[LaikaSpaceCat]

More Performance with a relatively small increase in Power consumption. Not the end of the world.

That's not more perfromance. The Titan Xp doesn't have Quadro drivers. When compared against Quadro, that Vega is slower than GP104.

 

[jrphoenix]

I've been waiting on Vega for months with my crossfire 5850. So, based on the FE links should I keep waiting on,my new ryzen system or go the 1080 / 1080 ti route?

 

[itsmydamnation]

That's not more perfromance. The Titan Xp doesn't have Quadro drivers. When compared against Quadro, that Vega is slower than GP104.

Not that it matters, these benchmarks are geometry setup limited, So What that says is that in geometry performance is above GeForce lines but below Quadro lines (not that far really either) unless you think AMD "IPC" regressed ~40% from hawii to vega and so did NV's from Maxwell to Pascel.

 

[TerionX6]

While we shouldn't be taking these results as a measure of expected real world performance, they show a 20% increase in "typical scenario" power brings a 32% increase in performance within these particular benchmarks.
I suspect it will still be a week or more before any usable leaks show us what's what though. Four programs of a similar workload are practically meaningless.

 

[raghu78]

If those power numbers are true AMD is going to have a very difficult time. Vega needed to be more efficient than GP102 given that it uses HBM2 instead of GDDR5X. Whats even worse is that Volta with GDDR6 will bring roughly 40-50% higher perf/watt. Looks like Nvidia's domination in performance and efficiency will continue.

 

[geoxile]

They overclocked it. It was 225W @12TF and they're pushing over 13TF

 

[Bacon1]

Someone made that comparison of 580 vs Vega, pretty huge boost.

 

[crisium]

Maybe? it's just basing max power consumption on pins.

PCIe + 8-pin + 8-pin = 375w water
PCIe + 8pin + 6-pin = 300w air

Then again the image below has air with 8+8 pin maybe not.

$1200/$1800 US? https://www.sabrepc.com/catalogsearch/result/?q=radeon+vega&order=relevance&dir=desc

 

[Bouowmx]

I've been waiting on Vega for months with my crossfire 5850. So, based on the FE links should I keep waiting on,my new ryzen system or go the 1080 / 1080 ti route?

Wait for Volta

Some sites may get hands on AMD Vega Frontier Edition and release gaming results. Then, people will say wait for Vega Gaming edition (near end of July).

 

[crisium]

Other possibility for water using less watts with same clocks. From the Anandtech Fury X review:

"a side note, as we mentioned in our architectural breakdown, the amount of power this card draws will depend on its temperature. 408W at the wall at 65C is only 388W at the wall at 40C, as current leakage scales with GPU temperature. Ultimately the R9 Fury X will trend towards 65C, but it means that early readings can be a bit misleading."

Lower temps is lower power consumption. While that's only 20W, consider that's only 40C vs 65C. Vega blower could run at over 90C, as Hawaii was 95C IIRC.

 

[Bacon1]

Other possibility for water using less watts with same clocks. From the Anandtech Fury X review:

"a side note, as we mentioned in our architectural breakdown, the amount of power this card draws will depend on its temperature. 408W at the wall at 65C is only 388W at the wall at 40C, as current leakage scales with GPU temperature. Ultimately the R9 Fury X will trend towards 65C, but it means that early readings can be a bit misleading."

Lower temps is lower power consumption. While that's only 20W, consider that's only 40C vs 65C. Vega blower could run at over 90C, as Hawaii was 95C IIRC.

Yeah its marked as TDP not real world watt used and obviously a WC can handle more power / heat than air blower, but still very odd to have them marked with the same clocks and specs. It just makes it look like the WC version uses 75w more power since people think TDP = Power Used and not Thermal Design Power (Max power)

 

[IllogicalGlory]

Someone made that comparison of 580 vs Vega, pretty huge boost.

The chart is misleading. Vega's advantage is actually 100% lower than those numbers.

Here's a corrected version:

Vega may have a great advantage over the 580, we'll have to wait and see, but the raw numbers aren't really where that is to be found. It'll be under the hood.

 

[tamz_msc]

Pixel fill-rate seems to indicate that it's a Fury X @1.4GHz. Is the new geometry pipeline not working as intended?