Based on what you've seen and heard so far, what do you think is the most optimistic and most pessimistic plausible scenarios for AMD's upcoming GPU releases?
My speculation is as follows:
Most optimistic plausible scenario
Pretty much everyone assumed, for little reason other than that it's a round number, that Fiji had 4096 shaders. On 6/16, AMD confirms that this number is indeed correct... for Fiji PRO (Fury), the cut-down version. The water-cooled Fiji XT card (Fury X) has a whopping 5120 shaders, and it beats the pants off of Titan X, exceeding it in most benchmarks by 20% or more. AMD claims the performance crown by enough of a margin that even a hot-clocked, fully-enabled GM200 card won't be able to win it back.
Meanwhile, the rest of the lineup is not neglected. The rumored "rebrands" are actually full refreshes; Hawaii, Tonga, Pitcairn, and Bonaire are replaced with Grenada, Antigua, Trinidad, and Tobago, respectively. All the refreshed models are on the low-leakage GloFo 28nm SHP process, providing much better performance-per-watt than their TSMC counterparts. They all have GCN 1.2 support, and the newest UVD/VCE block so that they can handle HEVC decoding. HDMI 2.0 is also incorporated - maybe even DisplayPort 1.3. Only Grenada has a cut-down version; the others only come with the full chip, no salvage dice. It's a leaner, smaller lineup, and one that can finally compete with Maxwell across the board.
The crowning touch is a design win for Antigua and Grenada in the new Mac Pro, to be announced in mid-2015.
Most pessimistic plausible scenario
How bad can it get? Pretty bad. If AMD is unable to tame the challenges of first-generation HBM, then Fiji could prove to be more of a tech demo than a viable commercial product. We could be looking at Bulldozer 2.0, and that's a blow AMD may not recover from. Fiji is a massive, go-for-broke chip; if despite its technical advancements it fails to recover the performance crown (falling short of Titan X by 5%-10%), then it will be a failure. AMD won't be able to sell it for more than $499, and that price won't be sufficient to recoup R&D costs. Meanwhile, let's suppose that the pessimistic rumors are true and all the cards except Fiji are straight rebrands. No port to GloFo process, just better binning, higher clocks, and more RAM. This leaves AMD's entire lineup in an uncompetitive position; without bringing anything new to the table, they'll continue to have to compete on price with the legion of ex-mining cards out there, not to mention the cheap R9 200 series clearance sales. AMD continues to hemorrhage market share for 18 months or more, waiting for the FinFET+ process to arrive... assuming they survive that long. The only real hope at that point is a buyout offer from Samsung.
My speculation is as follows:
Most optimistic plausible scenario
Pretty much everyone assumed, for little reason other than that it's a round number, that Fiji had 4096 shaders. On 6/16, AMD confirms that this number is indeed correct... for Fiji PRO (Fury), the cut-down version. The water-cooled Fiji XT card (Fury X) has a whopping 5120 shaders, and it beats the pants off of Titan X, exceeding it in most benchmarks by 20% or more. AMD claims the performance crown by enough of a margin that even a hot-clocked, fully-enabled GM200 card won't be able to win it back.
Meanwhile, the rest of the lineup is not neglected. The rumored "rebrands" are actually full refreshes; Hawaii, Tonga, Pitcairn, and Bonaire are replaced with Grenada, Antigua, Trinidad, and Tobago, respectively. All the refreshed models are on the low-leakage GloFo 28nm SHP process, providing much better performance-per-watt than their TSMC counterparts. They all have GCN 1.2 support, and the newest UVD/VCE block so that they can handle HEVC decoding. HDMI 2.0 is also incorporated - maybe even DisplayPort 1.3. Only Grenada has a cut-down version; the others only come with the full chip, no salvage dice. It's a leaner, smaller lineup, and one that can finally compete with Maxwell across the board.
The crowning touch is a design win for Antigua and Grenada in the new Mac Pro, to be announced in mid-2015.
Most pessimistic plausible scenario
How bad can it get? Pretty bad. If AMD is unable to tame the challenges of first-generation HBM, then Fiji could prove to be more of a tech demo than a viable commercial product. We could be looking at Bulldozer 2.0, and that's a blow AMD may not recover from. Fiji is a massive, go-for-broke chip; if despite its technical advancements it fails to recover the performance crown (falling short of Titan X by 5%-10%), then it will be a failure. AMD won't be able to sell it for more than $499, and that price won't be sufficient to recoup R&D costs. Meanwhile, let's suppose that the pessimistic rumors are true and all the cards except Fiji are straight rebrands. No port to GloFo process, just better binning, higher clocks, and more RAM. This leaves AMD's entire lineup in an uncompetitive position; without bringing anything new to the table, they'll continue to have to compete on price with the legion of ex-mining cards out there, not to mention the cheap R9 200 series clearance sales. AMD continues to hemorrhage market share for 18 months or more, waiting for the FinFET+ process to arrive... assuming they survive that long. The only real hope at that point is a buyout offer from Samsung.
Facebook
Twitter