The AMD Radeon VII Review: An Unexpected Shot At The High-End
by Nate Oh on February 7, 2019 9:00 AM ESTRadeon VII & Radeon RX Vega 64 Clock-for-Clock Performance
With the variety of changes from the Vega 10 powered RX Vega 64 to the new Radeon VII and its Vega 20 GPU, we wanted to take a look at performance and compute while controlling for clockspeeds. In this way, we can peek at any substantial improvements or differences in pseudo-IPC. There's a couple caveats here; obviously, because the RX Vega 64 has 64 CUs while the Radeon VII has only 60 CUs, the comparison is already not exact. The other thing is that "IPC" is not the exact metric measured here, but more so how much graphics/compute work is done per clock cycle and how that might translate to performance. Isoclock GPU comparisons tend to be less useful when comparing across generations and architectures, as like in Vega designers often design to add pipeline stages to enable higher clockspeeds, but at the cost of reducing work done per cycle and usually also increasing latency.
For our purposes, the incremental nature of 2nd generation Vega allays some of those concerns, though unfortunately, Wattman was unable to downclock memory at this time, so we couldn't get a set of datapoints for when both cards are configured for comparable memory bandwidth. While the Vega GPU boost mechanics means there's not a static pinned clockspeed, both cards were set to 1500MHz, and both fluctuated from 1490 to 1500MHZ depending on workload. All combined, this means that these results should be taken as approximations and lacking granularity, but are useful in spotting significant increases or decreases. This also means that interpreting the results is trickier, but at a high level, if the Radeon VII outperforms the RX Vega 64 at a given non-memory bound workload, then we can assume meaningful 'work per cycle' enhancements relatively decoupled from CU count.
As mentioned above, we were not able to control for the doubled memory bandwidth. But in terms of gaming, the only unexpected result is with GTA V. As an outlier, it's less likely to be an indication of increased gaming 'work per cycle,' and more likely to be related to driver optimization and memory bandwidth increases. GTA V has historically been a title where AMD hardware don't reach the expected level of performance, so regardless there's been room for driver improvement.
SPECviewperf is a slightly different story, though.
289 Comments
View All Comments
just4U - Friday, February 8, 2019 - link
It's not a turkey at all.. it beats a Vega64 for around 30% ads 2x the ram (which is not really utilized yet) has a 3 fan design with Amd's top end shroud/block takes less power, runs cooler, and has the same characteristics which means Amd was generous on power so undervolting it without appreciable performance losses will be easy enough to do as will overclocking.For me that's a winner. I have blower 1080s and their very loud if I let them or run things at stock (i undervolt there to..) and I've seen how loud the Vega56/64 blowers can be.. this with the 3 fans? pfft.. way quieter.
Oxford Guy - Friday, February 8, 2019 - link
I think you should look at the data in this review because your analysis is way off.ballsystemlord - Thursday, February 7, 2019 - link
They are sold out! All the online retailers I checked have no Radeon VIIs! Unless you go to ebay and pay way too much.Oxford Guy - Thursday, February 7, 2019 - link
Overpriced underbaked vaporware? Never-coulda-happen.It's an ugly time to be a "serious" PC gamer.
ballsystemlord - Friday, February 15, 2019 - link
Well, it's been a week. They came into stock for about 5min.LogitechFan - Friday, February 8, 2019 - link
amdumb defense force in full denial mode, sorry, we can't hear you over the 55db noise level of the radeon VII ;)))))))))rukufe - Friday, February 8, 2019 - link
If you want to play with AI, you need tensorflow, and for a "server" card, at this price, it doesn't not makes sense to not support tensorflow. AI is everywhere today. this card is obsolete.gsalkin - Friday, February 8, 2019 - link
So is this too little too late? I'm bewildered that even at 7nm this card is pulling 300W of power and generating insane noise.It's also unfortunate that the rumor of 128 ROPs was bunk. These cards definitely have an imbalance in the CU to ROP ratio. Nvidia Titan Xp had 96 ROPs strapped to 3840 SPs but AMD is shipping a max of 64?
Oxford Guy - Friday, February 8, 2019 - link
"It's also unfortunate that the rumor of 128 ROPs was bunk."That rumor typifies the irrational thinking that plagues the gaming community. AMD isn't going to make the effort of changing the Instinct GPU to better suit gamers. It isn't and it hasn't.
dr.denton - Saturday, February 9, 2019 - link
I wonder, do people actually read and comprehend these articles? By now it should be obvious to everyone, that VII is not and was never supposed to be AMD's next generation of GPU. In fact, they always denied that Vega 7nm would make it into the consumer market - and for very good reason: they had Navi for that. Now that Navi is delayed, they need something for people to talk about - and talk about it we do.