Skip to main content

AMD Radeon VII Will Ship Without Double-Precision

AMD's soon-to-be-released Radeon VII graphics card will come with FP64 (double-precision floating-point) disabled, Sasa Marinkovic, Director of Product Marketing at AMD, confirmed to Techgage in a report this week.

(Image credit: AMD)

One of the first things you may notice about the AMD Radeon VII is its resemblance to the existing Radeon Instinct MI50. The former is a gaming graphics card, while the latter is marketed towards data centers and enterprises. However, the two are both based on the Graphics Core Next (GCN) 5.0 microarchitecture and employ the Vega 20 silicon. Not to mention that they are also product of TSMC's 7nm FinFET manufacturing process. But the similarities don't stop there.

AMD Radeon VIIAMD Radeon Instinct MI50AMD Radeon RX Vega 64
Architecture (GPU)Graphics Core Next 5.0 (Vega 20)Graphics Core Next 5.0 (Vega 20)Graphics Core Next 5.0 (Vega 10)
Shaders384038404096
Peak FP32 Compute13.8 TFLOPS13.4 TFLOPs12.5 TFLOPS
Texture Units240240256
Base Clock Rate1450 MHz1200 MHz1200 MHz
GPU Boost Rate1800 MHz1746 MHz1536 MHz
Memory Capacity16GB HBM216GB HBM28GB HBM2
Memory Clock2000 Mbps*2000 Mbps1890 Mbps
Memory Bus4096-bit4096-bit2048-bit
Memory Bandwidth1024 GB/s1024 GB/s483.8 GB/s
ROPs646464
TDP300W*300W295W
Transistor Count13.2 billion13.2 billion12.5 billion
Die Size331 mm²331 mm²487 mm²

*=unconfirmed

AMD equipped both the Radeon VII and Radeon Instinct MI50 with 3,840 stream processors, 240 TMUs (texture mapping units), 64 ROPs (render output units) and 16GB of HBM2 memory across a 4096-bit memory interface. Being a gaming-focused product, the Radeon VII comes with a higher base clock and boost clock. Therefore, it seems the chipmaker had to take something away from the Radeon VII so it doesn't negatively impact Radeon Instinct MI50 sales. That's where FP64 comes in.

The Radeon Instinct MI50 can deliver up to 6.7 TFLOPs of FP64 peak performance. By gimping the Radeon VII's FP64 performance, AMD effectively creates a huge gap between the two models. Despite all this, the Radeon VII should be capable of pumping out 1,680 GFLOPs of FP64 performance and stay ahead of the Nvidia Titan RTX (509 GFLOPs) and GeForce RTX 2080 Ti (420 GFLOPs).

In a likely move to further segment the Radeon VII from the Radeon Instinct MI50, AMD has disabled PCIe 4.0 on the former. The reasoning is that AI workloads in the data center can saturate the PCIe 3.0 interface with relative ease, and limiting the available bandwidth would dissuade data center operators from repurposing cheaper consumer gear for those types of workloads. So, AMD has left the feature active only for the Radeon Instinct MI50.

For the average consumer, is it a dealbreaker that the Radeon VII's FP64 performance isn't on par with the Radeon Instinct MI50? Not really, since gamers can surely live without it. But there are professional users who are also gamers, and while they would have loved to have a graphics card that brings both premium FP64 and gaming performance, AMD isn't letting that happen with the Radeon VII.

  • JamesSneed
    Can we assume this was done to make it less attractive to miners?
    Reply
  • redgarl
    ^ Simply... no...

    I kind of understand AMD on this, if they didn't, the MI50 Instinct would have been outsold by the VII on price alone by lower quality silicon.

    Still, I believe AMD could have left it there if they could. It would have been a superb selling point for professional.
    Reply
  • thernn8891
    They should have left it intact and marketed it as a professional card that can also game. Now the only option for gamers is the Titan V or some ancient hardware off of EBay. There are plenty of ways to segregate the VII from the instincts without lopping off the FP64. There are many amateur professionals and students for which the Instincts are not a good fit beyond just price considerations and that has left an unfulfilled market which is hungry for a FP64 capable card that can also game but just not at the outrageous prices that Nvidia wants for the Titan V.

    Personally, FP64 was the primary driver of my interest in the card but now I couldn't care less about it. I knew many people that were excited also but now not one of them plan to get one. The extra ram is simply not enough to make it a compelling purchase.
    Reply
  • arielvtpma
    21677571 said:
    Can we assume this was done to make it less attractive to miners?

    Yes.
    Reply
  • thernn8891
    They could make a Radeon VII Pro with FP64 and sell it for $999 and it would sell like hotcakes in the community.

    Market it like we know our customers need to work but they also like to relax. Now you can do both with one card! The Radeon VII pro provides a 4k gaming experience and a similar compute value compared to a Titan V for a third of the price. Buy two and double your compute and it will still cost you less!

    Instant hit.
    Reply
  • shabbo
    If you're into FP64 compute then you should stick with the dedicated hardware and software that supports it. Most likely you'll be needing more than one GPU for those kind of tasks including ECC memory, and also affiliated with a university or a scientific organization that affords it. There's no reason for AMD to include FP64 and then have to support it at a higher-cost for everyone. It's perfectly fine the way it is.

    Maybe AMD can release drivers/software for $300 that you guys can purchase?
    Reply
  • rabbit4me1
    sorry AMD...crash and burner..and away you go :(
    Reply
  • derekullo
    21677679 said:
    21677571 said:
    Can we assume this was done to make it less attractive to miners?

    Yes.

    Miners don't even use the double precision feature.

    If every 100,000th calculation was wrong for a miner they wouldn't care. (Although they should probably dial back their overclock slightly.)

    They would just keep going with more calculations.

    Double precision is literally doing every calculation twice to make sure the math was done right.

    The chart on

    https://www.anandtech.com/show/12673/titan-v-deep-learning-deep-dive

    illustrates this beautifully.
    Reply
  • TJ Hooker
    21679060 said:
    Double precision is literally doing every calculation twice to make sure the math was done right.
    No, not at all. It refers to how many bits of precision you have. Storing an arbitrarily long number in a fixed number of bits results in rounding/truncation error, by doubling the number of bits you reduce this error. Double precision (FP64) FLOPS refers to the rate at which operations involving 64 bit floating point numbers can be performed.

    You're right that miners probably don't care about FP64 performance though. But that's because most popular algorithms for GPU mining these days are more memory-bound than compute-bound.
    Reply
  • mihen
    The AMD Vega Frontier Edition still exists if you wanted something to compete with the Titan V. It will get similar performance to this Radeon VII but consume more power.
    Reply