The AMD Radeon VII Review: An Unexpected Shot At The High-End
by Nate Oh on February 7, 2019 9:00 AM ESTFP64 Performance and Separating Radeon VII from Radeon Instinct MI50
One of the interesting and amusing consequences of the Radeon VII launch is that for the first time in quite a while, AMD has needed to seriously think about how they’re going to differentiate their consumer products from their workstation/server products. While AMD has continued to offer workstation and server hardware via the Radeon Pro and Radeon Instinct series, the Vega 20 GPU is AMD’s first real server-grade GPU in far too long. So, while those products were largely differentiated by the software features added to their underlying consumer-grade GPUs, Radeon VII brings some new features that aren’t strictly necessary for consumers.
It may sound like a trivial matter – clearly AMD should just leave everything enabled – but as the company is trying to push into the higher margin server business, prosumer products like the Radeon VII are in fact a tricky proposition. AMD needs to lock away enough of the server functionality of the Vega 20 GPU that they aren’t selling the equivalent of a Radeon Instinct MI50 for a fraction of the price. On the other hand, it’s in their interest to expose some of these features in order to make the Radeon VII a valuable card in its own right (one that can justify a $699 price tag), and to give developers a taste of what AMD’s server hardware can do.
Case in point is the matter of FP64 performance. As we noted in our look at the Vega 20 GPU, Vega 20’s FP64 performance is very fast: it’s one-half the FP32 rate, or 6.9 TFLOPS. This is one of the premium features of Vega 20, and since Radeon VII was first announced back at CES, the company has been struggling a bit to decide how much of that performance to actually make available to the Radeon VII. At the time of its announcement, we were told that the Radeon VII would have unrestricted (1/2) FP64 performance, only to later be told that it would be 1/8. Now, with the actual launch of the card upon us, AMD has made their decision: they’ve split it down the middle and are doing a 1/4 rate.
Looking to clear things up, AMD put out a statement:
The Radeon VII graphics card was created for gamers and creators, enthusiasts and early adopters. Given the broader market Radeon VII is targeting, we were considering different levels of FP64 performance. We previously communicated that Radeon VII provides 0.88 TFLOPS (DP=1/16 SP). However based on customer interest and feedback we wanted to let you know that we have decided to increase double precision compute performance to 3.52 3.46 TFLOPS (DP=1/4SP).
If you looked at FP64 performance in your testing, you may have seen this performance increase as the VBIOS and press drivers we shared with reviewers were pre-release test drivers that had these values already set. In addition, we have updated other numbers to reflect the achievable peak frequency in calculating Radeon VII performance as noted in the [charts].
The end result is that while the Radeon VII won’t be as fast as the MI60/MI50 when it comes to FP64 compute, AMD is going to offer the next best thing, just one step down from those cards.
At 3.5 TLFLOPS of theoretical FP64 performance, the Radeon VII is in a league of its own for the price. There simply aren’t any other current-generation cards priced below $2000 that even attempt to address the matter. All of NVIDIA’s GeForce cards and all of AMD’s other Radeon cards straight-up lack the necessary hardware for fast FP64. The next closest competitor to the Radeon VII in this regard is NVIDIA’s Titan V, at more than 4x the price.
It’s admittedly a bit of a niche market, especially when so much of the broader industry focus is on AI and neural network performance. But there’s none the less going to be some very happy data scientists out there, especially among academics.
| AMD Server Accelerator Specification Comparison | ||||
| Radeon VII | Radeon Instinct MI50 |
Radeon Instinct MI25 |
FirePro S9170 | |
| Stream Processors | 3840 (60 CUs) |
3840 (60 CUs) |
4096 (64 CUs) |
2816 (44 CUs) |
| ROPs | 64 | 64 | 64 | 64 |
| Base Clock | 1450MHz | 1450MHz | 1400MHz | - |
| Boost Clock | 1750MHz | 1746MHz | 1500MHz | 930MHz |
| Memory Clock | 2.0Gbps HBM2 | 2.0Gbps HBM2 | 1.89Gbps HBM2 | 5Gbps GDDR5 |
| Memory Bus Width | 4096-bit | 4096-bit | 2048-bit | 512-bit |
| Half Precision | 27.6 TFLOPS | 26.8 TFLOPS | 24.6 TFLOPS | 5.2 TFLOPS |
| Single Precision | 13.8 TFLOPS | 13.4 TFLOPS | 12.3 TFLOPS | 5.2 TFLOPS |
| Double Precision | 3.5 TFLOPS (1/4 rate) |
6.7 TFLOPS (1/2 rate) |
768 GFLOPS (1/16 rate) |
2.6 TFLOPS (1/2 rate) |
| DL Performance | ? | 53.6 TFLOPS | 12.3 TFLOPS | 5.2 TFLOPS |
| VRAM | 16GB | 16GB | 16GB | 32GB |
| ECC | No | Yes (full-chip) | Yes (DRAM) | Yes (DRAM) |
| Bus Interface | PCIe Gen 3 | PCIe Gen 4 | PCIe Gen 3 | PCIe Gen 3 |
| TDP | 300W | 300W | 300W | 275W |
| GPU | Vega 20 | Vega 20 | Vega 10 | Hawaii |
| Architecture | Vega (GCN 5) |
Vega (GCN 5) |
Vega (GCN 5) |
GCN 2 |
| Manufacturing Process | TSMC 7nm | TSMC 7nm | GloFo 14nm | TSMC 28nm |
| Launch Date | 02/07/2019 | 09/2018 | 06/2017 | 07/2015 |
| Launch Price (MSRP) | $699 | - | - | $3999 |
Speaking of AI, it should be noted that machine learning performance is another area where AMD is throttling the card. Unfortunately, more details aren’t available at this time. But given the unique needs of the ML market, I wouldn’t be surprised to find that INT8/INT4 performance is held back a bit on the Radeon VII. Or for that matter certain FP16 dot products.
Also on the chopping block is full-chip ECC support. Thanks to the innate functionality of HBM2, all Vega cards already have free ECC for their DRAM. However Vega 20 takes this one step further with ECC protection for its internal caches, and this is something that the Radeon VII doesn’t get access to.
Finally, Radeon VII also cuts back a bit on Vega 20’s off-chip I/O features. Though AMD hasn’t made a big deal of it up to now, Vega 20 is actually their first PCI-Express 4.0-capable GPU, and this functionality is enabled on the Radeon Instinct cards. However for Radeon VII, this isn’t being enabled, and the card is being limited to PCIe 3.0 speeds (so future Zen 2 buyers won’t quite have a PCIe 4.0 card to pair with their new CPU). Similarly, the external Infinity Fabric links for multi-GPU support have been disabled, so the Radeon VII will only be a solo act.
On the whole, there’s nothing very surprising about AMD’s choices here, especially given Radeon VII’s target market and target price. But these are notable exclusions that are going to matter to certain users. And if not to drive those users towards a Radeon Instinct, then they’re sure to drive those users towards the inevitable Vega 20-powered Radeon Pro.
Facebook
Twitter
RSS
289 Comments
View All Comments
repoman27 - Thursday, February 7, 2019 - link
The Radeon Pro WX 7100 is Polaris 10, which does not do DSC. DSC requires fixed function encoding blocks that are not present in any of the Polaris or Vega variants. They do support DisplayPort 1.3 / 1.4 and HBR3, but DSC is an optional feature of the DP spec. AFAIK, the only GPUs currently shipping that have DSC support are NVIDIA's Turing chips.The CPU in the iMac Pro is a normal, socketed Xeon W, and you can max the RAM out at 512 GB using LRDIMMs if you're willing to crack the sucker open and shell out the cash. So making those things user accessible would be the only benefit to a modular Mac Pro. CPU upgrades are highly unlikely for that platform though, and I doubt Apple will even provide two DIMM slots per channel in the new Mac Pro. However, if they have to go LGA3647 to get an XCC based Xeon W, then they'd go with six slots to populate all of the memory channels. And the back of a display that is also 440 square inches of aluminum radiator is not necessarily a bad place to be, thermally. Nothing is open about Thunderbolt yet, by the way, but of course Apple could still add existing Intel TB3 controllers to an AMD design if they wanted to.
So yeah, in order to have a product, they need to beat the iMac Pro in some meaningful way. And simply offering user accessible RAM and PCIe slots in a box that's separate from the display isn't really that, in the eyes of Apple at least. Especially since PCIe slots are far from guaranteed, if not unlikely.
halcyon - Friday, February 8, 2019 - link
Apple cannot ship Mac Pro with a vacuum cleaner. That 43 dBA is isane. Even if Apple downclocked and undervolted the bios, I doubt they could make it very quiet.Also, I doubt AMD is willing to sell the tons of them at a loss.
dark_light - Thursday, February 7, 2019 - link
Well written, balanced and comprehensive review that covers all the bases with just the rightamount of detail.
Thanks Nate Oh.
Anandtech is still arguably the best site for this content. Kudos guys.
drgigolo - Thursday, February 7, 2019 - link
So I got a 1080Ti at launch, because there was no other alternative at 4K. Finally we have an answer from AMD, unfortunately it's no faster than my almost 2 year old GPU, priced the same no less.I really think this would've benefitted from 128 rops, or 96.
If they had priced this at 500 dollars, it would've been a much better bargain.
I can't think of anyone who I would recommend this to.
sing_electric - Thursday, February 7, 2019 - link
To be fair, you could almost say the same thing about the 2080, "I got a 1080 Ti at launch and 2 years later, Nvidia released a GPU that barely performs better if you don't care about gimmicks like ray tracing."People who do gaming and compute might be very well tempted, people who don't like Nvidia (or just do like AMD) might be tempted.
Unfortunately, the cost of the RAM in this thing alone is probably nearly $350, so there's no way AMD could sell this thing for $500 (but it wouldn't surprise me if we see it selling a little under MSRP if there is plentiful supply and Nvidia can crank out enough 2080s).
eva02langley - Thursday, February 7, 2019 - link
That was the whole point of RTX. Beside the 2080 TI, there was nothing new. You were having the same performances for around the same price than the last generation. There was no price disruption.Oxford Guy - Thursday, February 7, 2019 - link
Poor AMD.We're supposed to buy a clearly inferior product (look at that noise) just so they can sell leftover and defective Instincts?
We're supposed to buy an inferior product because AMD's bad business moves have resulted in Nvidia being able to devalue the GPU market with Turing?
Nope. We're supposed to either buy the best product for the money or sit out and wait for something better. Personally, I would jump for joy if everyone would put their money into a crowdfunded company, with management that refuses to become eaten alive by a megacorp, to take on Nvidia and AMD in the pure gaming space. There was once space for three players and there is space today. I am not holding my breath for Intel to do anything particularly valuable.
Wouldn't it be nice to have a return to pure no-nonsense gaming designs, instead of this "you can buy our defective parts for high prices and feel like you're giving to charity" and "you can buy our white elephant feature well before its time has come and pay through the nose for it" situation.
Capitalism has had a bad showing for some time now in the tech space. Monopolies and duopolies reign supreme.
eva02langley - Friday, February 8, 2019 - link
Honestly, beside a RX570/580, no GPUs make sense right now.Funny that Polaris is still the best bang for the $ still today.
drgigolo - Saturday, February 9, 2019 - link
Well, at least you can buy a 2080Ti, eventhough the 2080 is of course at the same price point as the 1080Ti. But I won't buy a 2080Ti either, it's too expensive and the performance increase is too small.The last decent AMD card I had, was the R9 290X. Had that for a few years until the 1080 came out, and then, replaced that to a 1080Ti when I got a Acer Predator XB321HK.
I will wait until something better comes along. Would really like HDMI 2.1 output, so that I can use VRR on the upcoming LG OLED C9.
sing_electric - Thursday, February 7, 2019 - link
Oh, also, FWIW: The other way of looking at it is "damn, that 1080 Ti was a good buy. Here I am 2 years later and there's very little reason for me to upgrade."