It seems like even most of the smaller mini itx cases can still fit a 120mm radiator (like in linuses recent video with the silverstone case) and the nano itself isnt much smaller than the Fury X. So for the same money who really buys this over Fury X which is faster, cooler, more headroom, looks better, has backplate etc etc......
I don't get it......needs to be same price as regular Fury to make sense.....
Ya, that's true. I was expecting a price of $449, maybe $499 but $649 is way too high. The other thing is they are putting the best binned Fiji XT chips inside but apparently have a 175W hard power limit? They should have at least allowed the full 225W power limit.
Honestly, they should have given this card a dual personality. Why not add a 6+8-pin connector and force it to run at 175W power limit on BIOS 1 but have BIOS 2 (Uber BIOS) that unlocked full potential for 250-300W TDP? This way someone could easily pay $649 for the best binned Fiji XT GPU and slap their own water-block/AIO CLC but get benefit of the least leaky Fiji XT chips.
As it stands, the Fury X is small enough to fit in most miniITX cases and will be typically operating at 16-17% higher clocks (1050 vs. 900). The only two positives I see in this case is the Nano foreshadows next generations of HBM2 GPUs and how small they will be and it's a glimpse at the Fury X2 as Nano is essentially 1/2 of the Fury X2.
Excitement around the high-end products creates a halo around the entire lineup that pulls in low information consumers. AMD is missing this.
An excellent post!! Pleasure to read.
I'd add that previous management at AMD did try your approach though and it failed, every generation since HD4000. HD4850/5850 had spectacular price/performance, HD6950 unlocked to 6970, R9 290 = 290X, and yet AMD hardly made much $. What they got was 'empty' market share with little profits to show for it. What's better to have 10% market share with profits OR 50% market share with no profits/losses?
But I think while Lisu Su had the right idea overall (Letting products at a loss to gain market share is a horrible strategy long-term), she raised prices too high at a vulnerable time for AMD. She should have coasted this generation with something special like Fury at $399-429, Fury X at $549 and once AMD was stronger, tried raising prices next gen with 16nm HBM2.
Unfortunately this basically means no more historical price/performance from AMD or NV. Now that AMD has abandoned price/performance for the most part, NV will for sure go for a 3rd generation in a row of mid-range Pascal at $500+ prices. For many gamers that means delaying their GPU upgrades every 3-5 years instead of every 2-3 years in the past because we now have to wait much longer to get a great increase in price/performance when upgrading.
Sadly Intel has money, NVIDIA licenses and influence.
Now that MS and Intel is marrying again, expecting to.see the next Xbox (if it survives) made exclusively with Intel and NVIDIA parts, like the original Xbox. Maybe a Core i5 with a Gtx 970 with insane low prices.
I've heard the same wishful thinking for 10 years in a row. There have been quite a few veteran AT posters on here that left the forums or their reputation got destroyed because they kept insisting on the above. Before PS4/XB1 came out, some even insisted that in no way will Sony/MS move to x86 and abandon the awesome sauce (sarcasm) that was the Cell.
While it's hard to predict what new consoles will have as they are still 3-4 years out, Nintendo's next console is rumoured to have AMD parts in it so your theories continue to be wrong.
And it will perform up to 700% better than any AMD solution it has. Give up.
Whatever Intel/AMD have in 2018-2020, we would need to compare to AMD's CPU/GPU parts of that time. Considering how far behind Intel is in graphics and NV's non-existent x86 CPU presence, the most logical solution is actually an AMD APU. Since we don't know what parts AMD will have for sale in 2018-2020, your statement is just a conjecture.
AMD failed hard and deserves a cruel death.
With that one statement you sir have just destroyed your reputation on AT because even if some people prefer NV/Intel, no one sane wants to have no competition in the CPU/GPU space -- only shareholders, short-sellers and blind fanboys.
The nano can fit in a VERY tiny case.
Ya, that's true. In that scenario, the Nano and to an extent the Fury X have no competition. There are plenty of mini ITX cases that will not accept a GTX980 inside. Some people in this thread stating that we've had this level of performance in that form factor in this TDP is also BS. You cannot fit an after-market 980 inside these cases:
Also, if AMD's statements that the Nano = Fury nonX in performance are too be believed, then the 980 isn't even as fast and 970 is even further behind.
If the Nano = Fury 4GB at 98%, that's pretty remarkable for a 175W card as nothing from AMD/NV can touch that. We are talking just 16% behind a reference GTX980Ti.
The price is the biggest obstacle for this card but as far as technological breakthrough, no other card can match it in this form factor in this power usage envelope. Too bad that it appears AMD set a hard 175W power limit. Should have made this card with a 250-300W TDP headroom and 6/8+8-pin connectors.