AMD Chases Crucial $279 Price Point with Radeon R9 280

[alwayssts]

If TSMC doesn’t start verbalizing they will be providing excellent commercial viability on 20Nm by mid-summer, I might say Nvidia could launching another Maxwell on 28Nm to blow this $200-300 segment open. Nvidia wants some of the Litecoin money, and as Hash/watt is so great with Maxwell they’re at a tipping point. Wait around for worthwhileness of 20Nm, which really only brings real significance in the high-end enthusiast parts. Especially when neither (Nvidia/AMD) want a repeat of TSMC ramp-up like the 28Nm again. If Nvidia could deliver a Maxwell that bests a GTX760 on a ≤ 220 mm² die, 130W, and hash that provides what this R9 280 provides, I’d say Nvidia has to do it.

Except all indications seem to point to late q3 to q4, regardless of what tsmc says. Be that a year from the launch of Hawaii, six months or so from this onslaught of refresh parts, the fact amd publicly stated they were taping out chips this quarter, that nvidia said the rest of maxwell was coming later this year, and that the public cost per transistor maps (that has been floating around for years) all pretty much point to exactly that. To be literal to their respective histories on the ramp/price-per-xtor it would suggest late q3 for amd and late q4 for nvidia, but both may end up around the same time.

On top of that, what you're asking for doesn't make sense. Maxwell is clearly based around 6smm and 2mb cache blocks. 750ti is essentially half of what a 8800gt was in a lower market...1 less smm released earlier on a larger process to compensate for the power increase (20% in power from logic + whatever clockspeed increases 20nm gives with such more logic in the same power envelope; probably around 1200mhz to 750ti's ~1150 for a complete difference of around 20-25%) while simultaneously working as class-leading <75w part. What you're asking for is essentially a part that is less than 2x gm107, which really doesn't make sense on 28nm short of something really weird like a low-clocked 9smm part with 24 ROPs/3mb cache on a 192-bit bus, which I doubt we see.

On the flipside, what do you expect of a higher-up part on 28nm? If you're expecting die savings compared to gk104, I think you will be sadly mistaken. The arch again is catered towards the higher logic use of 20nm (ex: the extra cache) while using less power for similar to slightly higher core clocks (rather than scaling to higher clocks with more voltage and less logic like most processes before it) and conceivably smaller/slower mem controllers to keep power down, which will become more-so apparent as they move to larger designs. In short, a 12smm part (the only one that makes sense versus gk104) would be faster per clock than gk104, probably larger (depending on cache + more units vs mem controller size offset), but proportionally use less power because of design efficiency/slower memory speed/bus. While I suppose it's possible they could keep such a design under 225w on 28nm, and it would be good for us, it would probably be bad for nvidia's bottom line (a wholly new design for what amounts to a small improvement).

That all said, when it comes to this 280...the price is indeed a shame compared to old product prices, but also reflective of the situation: the yields are good and the cost savings are already passed on to 280x (at least at msrp). The price difference is pretty much directly proportional to real avg shader perf per clock and memory speed differences....meaning at least these will probably clock decently and similar to 280x (as opposed to 7950 which was power/clock limited) so it should be able to wiggle a decent spot between 760 and 770 if priced accordingly.

 

[vega22]

Pretty much this. How many enthusiasts give a damn about AMD chasing some obscure price point that they see as some kind of sweet spot? Most if not all are either only looking at high end hardware (Your 780, 780ti, 290 and 290X), or they already have 7970's, 7950's and 680's, which are essentially what all this crap is. Everyone else will buy the super low end cards for HTPC systems, or the best price/performance cards available within their budget range, which for some reason seems to be the 760 atm (this is just an observation, I see a lot of people recommending these as mid range cards for those that can't quite reach the high end area).

This is nothing that hasn't happened before, and most enthusiasts won't ever buy them anyway. There is no doubt better cards at better price points.

but it is the sweet spot, the low high end/high mid range gaming cards are the top sellers year after year for both companies. they are seen as the best value by many including the large oem, the average joe public on the street and to really make matters worse miners are going to love them. for a long time the 7950 has been the darling of the bit coin world. it is the reason why my card still sells for over £200 when they only cost £220 new 18 months ago.

if the r9 280 sells at anything like its £170 rrp then its ability to run most games at 1440p with ease alone and monstering them all with 2 in xfire for £350 will make it a hit with gamers. miners will love the kw:hash rate as they already do, only down side as they will drive up the price...i only hope the draw of maxwell will make them wait on any upgrades.

 

[RCoon]

£170 rrp

plus VAT. American RRP doesn't account for UK Tax. This will just be a 7950 or close to it, being sold at the same price it released at (I bought my two 7950's for £210 each)

 

[HumanSmoke]

8800gt
8800gts
9600gt
9800gt
9800gtx
9800gtx+
gtx250
1 core, 3 generations of cards.
nothing new.

Indeed, although the 9600GT doesn't belong in that list since it featured the G94a GPU rather than the G92. You're probably thinking of the 9600GSO which was G92 powered (192-bit, 384 or 768MB) in one guise, although a rarer version used the G94 (128 or 256-bit, 512MB, 1GB). The 8800GS also belongs on your list - the 9600GSO G92 is a rebrand of the same card.
Also, it is the GTS 250 not GTX 250, and if you're including the die shrink (G92b) then that should also include the G92a/b versions of the GTS 240.

at least this has only been rebranded once.

Not quite
HD 7950
HD 7950 Boost
HD 8950 (OEM)
R9 280

Wait around for worthwhileness of 20Nm, which really only brings real significance in the high-end enthusiast parts.

Indications coming out of Taiwan are just the opposite. 20nm looks optimized for SoC/low power GPUs and hasn't shown any tangible performance benefits over 28nm when scaled to a larger die - at least when you take into account the increased wafer cost. It is my understanding that both Nvidia and AMD will stand pat with 28nm for the performance/enthusiast segment until the 16nm FF (FinFET) derivative of the 20nm process arrives. Note how quickly the ramp of 16nm FF (16nm FEOL+20nm BEOL) is due to follow 20nm.

 

[iKhan]

Thanks, but I think I'll wait for mid-range and high end Maxwell cards.

 

[xvi]

 

[Casecutter]

Except all indications seem to point to late q3 to q4, regardless of what tsmc says... should be able to wiggle a decent spot between 760 and 770 if priced accordingly.

Thanks for this always good to read more differing information.
I just think that AMD and Nvidia are not going for another leap-of-faith with risk production. And, it’s just too early to read anything from the TSMC tea leaves at this point. Not exactly understanding that another 28mm Maxwell couldn’t be a good move, but it would hold Nvidia for a year and appears if miners became fond of it could make some good ROI.

 

[Divide Overflow]

Downgraded gap filler for old, re-badged tech.
*Yawn*

 

[arbiter]

8800gt
8800gts
9600gt
9800gt
9800gtx
9800gtx+
gtx250

1 core, 3 generations of cards.

nothing new.

at least this has only been rebranded once.

stick around long enough and you will see it happen again too.

you need to check your facts, 8800GTS was g80 90nm gpu, 9600gt was g94a/g94b 65/55nm part. and well gtx250 was considered mid/low range part which was g94b which is kinda normal.

Downgraded gap filler for old, re-badged tech.
*Yawn*

Yea Nvidia has put out new gpu, if and when they do larger gpu to replace high end with maxwell and they can keep power down to like 200 watts or lower, would be huge.

 

[Casecutter]

If this was a normal gaming market AMD would be pricing below MSRP and they wouldn’t need to be considering moving what little binned (geldings) like this and the R7 265 to feed the demand. None of these SKU’s should be here right now, but the market is burning through everything and AIB are clamoring to get anything.

 

[FX-GMC]

Indeed, although the 9600GT doesn't belong in that list since it featured the G94a GPU rather than the G92. You're probably thinking of the 9600GSO which was G92 powered (192-bit, 384 or 768MB) in one guise, although a rarer version used the G94 (128 or 256-bit, 512MB, 1GB). The 8800GS also belongs on your list - the 9600GSO G92 is a rebrand of the same card.
Also, it is the GTS 250 not GTX 250, and if you're including the die shrink (G92b) then that should also include the G92a/b versions of the GTS 240.

Not quite
HD 7950
HD 7950 Boost
HD 8950 (OEM)
R9 280

Indications coming out of Taiwan are just the opposite. 20nm looks optimized for SoC/low power GPUs and hasn't shown any tangible performance benefits over 28nm when scaled to a larger die - at least when you take into account the increased wafer cost. It is my understanding that both Nvidia and AMD will stand pat with 28nm for the performance/enthusiast segment until the 16nm FF (FinFET) derivative of the 20nm process arrives. Note how quickly the ramp of 16nm FF (16nm FEOL+20nm BEOL) is due to follow 20nm.

Figures he would've been wrong all along.

On point, "AMD chases crucial $279 price point with Radeon R9 280, retailers break their legs."

 

[HumanSmoke]

you need to check your facts, 8800GTS was g80 90nm gpu,

That was the original 8800GTS 640MB/320MB, a relatively expensive and short lived G80 salvage part. The G92 8800GTS should be much more well known, since it essentially made the 8800GTX and Ultra obsolete overnight.

and well gtx250 was considered mid/low range part which was g94b which is kinda normal.

There is no such card. The GTS 250 was essentially a rebranded 9800 GTX+ (as was the OEM GTS 150 and GTS 240)

 

[LeonVolcove]

Another REBRANED with nice sticker

 

[KainXS]

8800GT-> 9800GT(early ones)
9800GTX+ -> GTS250

8800GTS -> 9800GTX(not really, 9800GTX had an enhanced PCB and newer core)
9600GT was never rebranded(not including oem) and was not a G92 either
9800GT(die strunk) -> GTS240(who cares about oem-_-)

I don't know why this comes up they both do it, let it go

 

[vega22]

kain that is the point i was trying to make dude, they both do it when they have a lead in performance.

nice 1 smoke, it was the gso which i was thinking of. one of the best sleeper cards they did at the time as it overclocked like stink and had plenty of mem bandwidth for high res gaming. i was trying to keep away from oem cards like the hd8000 and gts300s which also had another sticker on the same hardware too. mainly as they never hit the mass markets but they are also a great indicator of how long the market had stagnated then too.

the g92b started with the 98gtx+ giving them all the ability to run faster clocks but not by much. a good overclocking g92 would hit the same speeds as a b.

 

[Ja.KooLit]

If you live in the US, expect to pay at least $50 over MSRP.

korea prices are like $100 over MSRP. All the time that is always the trend and it will stay like that. no matter green or red cards.