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
schizoide - Thursday, February 7, 2019 - link
Sure it does, at the bottom-end. It basically IS an instinct mi50 on the cheap.GreenReaper - Thursday, February 7, 2019 - link
Maybe they weren't selling so well so they decided to repurpose before Navi comes out and makes it largely redundant.schizoide - Thursday, February 7, 2019 - link
IMO, what happened is pretty simple. Nvidia's extremely high prices allowed AMD to compete with a workstation-class card. So they took a swing.eva02langley - Friday, February 8, 2019 - link
My take to. This card was never intended to be released. It just happened because the RTX 2080 is at 700+$.In Canada, the RVII is 40$ less than the cheapest 2080 RTX, making it the better deal.
Manch - Thursday, February 7, 2019 - link
It is but its slightly gimped perf wise to justify the price diff.sing_electric - Thursday, February 7, 2019 - link
Anyone else think that the Mac Pro is lurking behind the Radeon VII release? Apple traditionally does a March 2019 event where they launch new products, so the timing fits (especially since there's little reason to think the Pro would need to be launched in time for the Q4 holiday season).-If Navi is "gamer-focused" as Su has hinted, that may well mean GDDR6 (and rays?), so wouldn't be of much/any benefit to a "pro" workload
-This way Apple can release the Pro with the GPU as a known quantity (though it may well come in a "Pro" variant w/say, ECC and other features enabled)
-Maybe the timing was moved up, and separated from the Apple launch, in part to "strike back" at the 2080 and insert AMD into the GPU conversation more for 2019.
The timeline and available facts seem to fit pretty well here...
tipoo - Thursday, February 7, 2019 - link
I was thinking a better binned die like VII for the iMac Pro.Tbh the Mac Pro really needs to support CUDA/Nvidia if it's going to be a serious contendor for scientific compute.
sing_electric - Thursday, February 7, 2019 - link
I mean, sure? but I'm not sure WHAT market Apple is going after with the Mac Pro anyways... I mean, would YOU switch platforms (since anyone who seriously needs the performance necessary to justify the price tag in a compute-heavy workload has almost certainly moved on from their 2013 Mac Pro) with the risk that Apple might leave the Pro to languish again?There's certainly A market for it, I'm just not sure what the market is.
repoman27 - Thursday, February 7, 2019 - link
The Radeon VII does seem to be one piece of the puzzle, as far as the new Mac Pro goes. On the CPU side Apple still needs to wait for Cascade Lake Xeon W if they want to do anything more than release a modular iMac Pro though. I can't imagine Apple will ever release another dual-socket Mac, and I'd be very surprised if they switched to AMD Threadripper at this point. But even still, they would need XCC based Xeon W chips to beat the iMac Pro in terms of core count. Intel did release just such a thing with the Xeon W 3175X, but I'm seriously hoping for Cascade Lake over Skylake Refresh for the new Mac Pro. That would push the release timeline out to Q3 or Q4 though.The Radeon VII also appears to lack DisplayPort DSC, which means single cable 8K external displays would be a no-go. A new Mac Pro that could only support Thunderbolt 3 displays up to 5120 x 2880, 10 bits per color, at 60 Hz would almost seem like a bit of a letdown at this point. Apple is in a bit of an awkward position here anyway, as ICL-U will have integrated Thunderbolt 3 and an iGPU that supports DP 1.4a with HBR 3 and DSC when it arrives, also around the Q3 2019 timeframe. I'm not sure Intel even has any plans for discrete Thunderbolt controllers after Titan Ridge, but with no PCIe 4.0 on Cascade Lake, there's not much they can even do to improve on it anyway.
So maybe the new Mac Pro is a Q4 2019 product and will have Cascade Lake Xeon W and a more pro-oriented yet Navi-based GPU?
sing_electric - Thursday, February 7, 2019 - link
Possibly, but I'm not 100% sure that they need to be at the iMac Pro on core count to have a product. More RAM (with a lot of slots that a user can get to) and a socketed CPU with better thermals than you can get on the back of a display might do it. I'd tend to think that moving to Threadripper (or EPYC) is a pipe dream, partly because of Thunderbolt support (which I guess, now that it's open, Apple could THEORETICALLY add, but it just seems unlikely at this point, particularly since there'd be things where a Intel-based iMac Pro might beat a TR-based Mac Pro, and Apple doesn't generally like complexities like that).Also, I'd assumed that stuff like DSC support would be one of the changes between the consumer and Pro versions (and AMD's Radeon Pro WX 7100 already does DSC, so its not like they don't have the ability to add it to pro GPUs).