Cast your vote Do you agree or disagree with GPUBoss?

Thanks for adding your opinion. Follow us on Facebook to stay up to date with the latest news!

Differences What are the advantages of each

Front view of Radeon RX Vega 64

Reasons to consider the
AMD Radeon RX Vega 64

Report a correction
Better floating-point performance 12,665 GFLOPS vs 11,340 GFLOPS More than 10% better floating-point performance
Much wider memory bus 2,048 bit vs 352 bit More than 5.8x wider memory bus
Higher texture rate 395.8 GTexel/s vs 354.4 GTexel/s More than 10% higher texture rate
More shading units 4,096 vs 3,584 512 more shading units
More texture mapping units 256 vs 224 32 more texture mapping units
Front view of GeForce GTX 1080 Ti

Reasons to consider the
Nvidia GeForce GTX 1080 Ti

Report a correction
Much higher effective memory clock speed 11,008 MHz vs 1,890 MHz More than 5.8x higher effective memory clock speed
Higher clock speed 1,480 MHz vs 1,247 MHz Around 20% higher clock speed
More memory 11,264 MB vs 8,192 MB Around 40% more memory
Higher pixel rate 139.2 GPixel/s vs 98.94 GPixel/s More than 40% higher pixel rate
More render output processors 88 vs 64 24 more render output processors
Higher memory clock speed 1,376 MHz vs 945 MHz More than 45% higher memory clock speed
Lower TDP 250W vs 295W More than 15% lower TDP

Features Key features of the Radeon RX Vega 64  vs GeForce GTX 1080 Ti 

memory bandwidth Rate at which data can be read from or stored in onboard memory

Radeon RX Vega 64
483.8 GB/s
GeForce GTX 1080 Ti
484.4 GB/s

pixel rate Number of pixels a graphics card can render to the screen every second

Radeon RX Vega 64
98.94 GPixel/s
GeForce GTX 1080 Ti
139.2 GPixel/s

texture rate Speed at which a graphics card can perform texture mapping

Radeon RX Vega 64
395.8 GTexel/s
GeForce GTX 1080 Ti
354.4 GTexel/s

floating point performance How fast the gpu can crunch numbers

Radeon RX Vega 64
12,665 GFLOPS
GeForce GTX 1080 Ti
11,340 GFLOPS

shading units Subcomponents of the gpu, these run in parallel to enable fast pixel shading

texture mapping units Built into each gpu, these resize and rotate bitmaps for texturing scenes

Specifications Full list of technical specs


Radeon RX Vega 64  vs
GeForce GTX 1080 Ti 
GPU brand AMD Nvidia
GPU name Vega 10 GP102
Clock speed 1,247 MHz 1,480 MHz
Turbo clock speed 1,546 MHz 1,582 MHz
Is dual GPU No No
Reference card None None

raw performance

Shading units 4,096 3,584
Texture mapping units 256 224
Render output processors 64 88
Pixel rate 98.94 GPixel/s 139.2 GPixel/s
Texture rate 395.8 GTexel/s 354.4 GTexel/s
Floating-point performance 12,665 GFLOPS 11,340 GFLOPS


Radeon RX Vega 64  vs
GeForce GTX 1080 Ti 
Memory clock speed 945 MHz 1,376 MHz
Effective memory clock speed 1,890 MHz 11,008 MHz
Memory bus 2,048 bit 352 bit
Memory 8,192 MB 11,264 MB
Memory bandwidth 483.8 GB/s 484.4 GB/s

noise and power

TDP 295W 250W


Showing 24 comments.
Really? Find any cheap RX580s recently?
Not if you can't buy one for 6 months. 1080ti arriving tomorrow.
it is 4-6% inferior to gtx 1080, and about 23% inferior to gtx 1080ti. it just came out and someone has already did the test, WHILE BEING MORE EXPENSIVE THAN BOTH.
not a typo. rx vega got smashed. it is inferior to gtx 1080ti while being more expensive. 23% inferior performance is too much tbh.
rx vega 64 just got smashed. it is just about equal to gtx 1080, while being so much more expensive. GTX 1080 TI is about 23% better than this shit vega while being even cheaper.
your rx vega 64 got smashed. it is just about equal to gtx 1080, while being so much more expensive
You want AMD For Rendering.
cough mining cough khe khe
Why don't you wanna see AMD win? They have cheap GPUs that perform well. They're the good guys here.
Tks my friend. For the price, then compensatory a Vega.
Ever since the 3Dfx Voodoo 2 we know that you get the best results with a ROP for every 2 TMUs, not 4 TMUs. If only AMD released a Vega with 128 ROPs, that would be great. Is it too much to ask? Sure, compute performance is definitely there even today, but the gaming performance depends on ROPs with pipelines as short as possible, but able to handle the clock speed.
Which is bs because more modules improves the throughput, not the latency. It's more like "effective bus-width" that increases with more modules.
That bus though! I'm definitely looking forward to the Rx Vega 64
The technology uses everything so it can push more power. Its like a 8GB HBM2 has a similiar power as a 16GB DDR5 module. No gaming trought because this GPU should be used for professional use as rendering of videos... I mean for business use.
AMD Vega is made to look too good here, wait for benchmarks to come out. It is slightly faster than a 1080 but slower than a 1080 Ti.
Can anyone translate me this data up to me: What is best to do the rendering in vray sketchup?
AMD will come with a cheaper price, as always. So the competition is still on
So what, AMD finally came close to making something similar to the 1080 TI and its still barely better? If i remember correctly, when GeForce release the 1080 it was times better than the old products, which was a huge step-up, meanwhile AMD is barely able to keep up it seems.
It's probably from all memory modules combined ;) just like my old hd7850, memory clock 1250mhz but it has four memory modules so effective memory speed is 5000mhz
I doubt it's got clock speed of 11,008 MHz Definitely a typo... ;)
HBM2 use a different technology but think about raw data and see this memory bus AMD 2,048 bit vs Nvidia 352 bit, but memory clock speed AMD 1,890 Mhz vs Nvidia11,008 MHz simple math here: AMD 2048x1890=3,870,720 bits vs Nvidia 352x11008=3,874,816 bits that is a 0.1057% in data transfer different on paper and in theory HBM2 technology offers some better performance if used in the right way especially in case of sli/crossfire kind of scenario.
Vega uses new HBM2 memory thus the difference as it is different technology.
Yeah, what happened there?
The memory diference is so vast
comments powered by Disqus