AMD Ryzen SKU and Price Information/Speculation.

[simas]

any word on MB pricing?

 

[inf64]

Re-post from benchmarks topic:
Someone at reddit claims the 1700(no X) is 8C/16T 3.1Ghz base and 3.7Ghz max Turbo SKU. Fits well with 65W bracket.

 

[Khato]

When did this -30% expectation change to 0%?

It didn't? At $500 for the top end SKU, it's about 85% the retail price of an i7-6850k.

When has AMD not priced based on performance compared to Intel's offerings?

 

[CatMerc]

Re-post from benchmarks topic:
Someone at reddit claims the 1700(no X) is 8C/16T 3.1Ghz base and 3.7Ghz max Turbo SKU. Fits well with 65W bracket.

I asked him for source, and he claims he has an insider. I don't know the guy, so take it with a large grain of salt.

 

[Doom2pro]

I asked him for source, and he claims he has an insider. I don't know the guy, so take it with a large grain of salt.

 

[Dresdenboy]

It didn't? At $500 for the top end SKU, it's about 85% the retail price of an i7-6850k.

When has AMD not priced based on performance compared to Intel's offerings?

They surely have a pattern. I initially expected 20%. But they don't have to use past factors based on perf. Platform features, power efficiency are somewhat different, too.

 

[LTC8K6]

Re-post from benchmarks topic:
Someone at reddit claims the 1700(no X) is 8C/16T 3.1Ghz base and 3.7Ghz max Turbo SKU. Fits well with 65W bracket.

https://twitter.com/CPCHardware/status/829685483408089090

 

[Atari2600]

https://twitter.com/CPCHardware/status/829685483408089090

Needless to say, that should probably be caveated with "that they know of".

 

[Agent-47]

Is it just me or does the naming scheme makes no sense sense at all. I mean why call two 8c16t cpus 1700 and 1800 over a 200MHz clock difference and then call a 6c12t core 1600? what if they realize a new 6c12t with a 200Mhz bump, are they going to call it 1700-6c??

Also I am not sure a 200MHz clock difference between 1800X and 1700X justify a £90 price difference considering both are unlocked with XFR. maybe £30 at most. I along with most people in this thread would probably go for the 1700X instead.

I call BS.

 

[Doom2pro]

Is it just me or does the naming scheme makes no sense sense at all. I mean why call two 8c16t cpus 1700 and 1800 over a 200MHz clock difference and then call a 6c12t core 1600? what if they realize a new 6c12t with a 200Mhz bump, are they going to call it 1700-6c??

Also I am not sure a 200MHz clock difference between 1800X and 1700X justify a £90 price difference considering both are unlocked with XFR. maybe £30 at most. I along with most people in this thread would probably go for the 1700X instead.

I call BS.

I call marketing wank... Too many sources to be BS at this point... Sounds to me like AMDs idle marketing team got a wild hair up their ass and are getting over excited for Ryzen... You know the gems AMD Marketing is capable of, why is this so surprising?

 

[richierich1212]

I really hate AMD's marketing. Why continue to go with 7/5/3 naming scheme? This is a brand new platform, these idiots need to break free from the old AMD and quit copying Intel in their naming schemes (who copied BMW lol)

 

[Doom2pro]

I really hate AMD's marketing. Why continue to go with 7/5/3 naming scheme? This is a brand new platform, these idiots need to break free from the old AMD and quit copying Intel in their naming schemes.

Sometimes you have no choice, the market leader sets the status quo and everyone has to follow suit... Look at the RX line, you have the XXX system like nVidia, GTX 970, 980, 990, AMD RX 460, 470, 480...

If you have been behind as a company for so long, sometimes you have no choice but to play follow the leader, because that is what consumers know, and anything you introduce that is different from the status quo is going to naturally be considered foreign and inferior.

 

[Agent-47]

I call marketing wank... Too many sources to be BS at this point... Sounds to me like AMDs idle marketing team got a wild hair up their ass and are getting over excited for Ryzen... You know the gems AMD Marketing is capable of, why is this so surprising?

Yes, too many sources indeed. But I will still be surprised if this naming is true.

but the prices still makes no sense. £90 over a 200 MHz bump? a FX 6300 and 6350 has a £20 difference over 400Mhz. i5 7400 and 7500 has a £15 difference over 400 Mhz.

I still call BS, otherwise their marketing has just redefined horses assery.

 

[CatMerc]

If they went with Ryzen#, should have at least called it according to core number. I.E Ryzen 8 1800X.

Makes more sense, and also has the benefit of one upping Intel.

 

[Mopetar]

I was expecting higher prices. Either AMD is going for massive sales to drive the stock price or the performance is a bit below what I expected.

 

[psolord]

Will it be shipped worldwide at the same time or will the EU get it first, like Skylakes iirc?

Also since the cpus will be unlocked, why do I need xfr? I mean if the 1700X gets another 200-300Mhz, isn't it nearly the same if I give the multiplier 2X-3X myself? It's not gonna break the wattage bank, is it? Nor the temps with the D14 I am going to slap on it.

 

[Agent-47]

Will it be shipped worldwide at the same time or will the EU get it first, like Skylakes iirc?

Also since the cpus will be unlocked, why do I need xfr? I mean if the 1700X gets another 200-300Mhz, isn't it nearly the same if I give the multiplier 2X-3X myself? It's not gonna break the wattage bank, is it? Nor the temps with the D14 I am going to slap on it.

The point is it is done automatically by the system and you don't have to hours of stress test to ensure stability. But importantly, it will be most effective to use XFR in single threaded and dual threaded environment. XFR should be able to reach a higher OC on 1c or 2c compared to a 8c manual OC. This is how i think AMD looks to combat intel's superior ST performance in single threaded applications/benchmarks, i.e. by aggressively OCing individual cores.

Having said that i think there will still be room for manual overclock in order to push the CPU to its absolute limits.

 

[Jan Olšan]

I asked him for source, and he claims he has an insider. I don't know the guy, so take it with a large grain of salt.

That base clock is IMO too high. Take for example Trinity.

The 95W A10-5800K had base clock 3,8 GHz with turbo boost 4,1 GHz. The 65W SKU that was based on it (A10-5700) had base clock 3,4 GHz with 4,0 GHz max boost. The base clock penalty in other words was more than 10 %. But it was also a second generation 32nm product (after Llano - or third if we count Zambezi).

So IMHO The base clock of Ryzen 7 1700 will be 3,0 GHz at best, but that is IMHO still not likely, I expect 2,8-2,9 GHz to be safe.

 

[cytg111]

And once more, pricing is the best early indicator of performance. Assuming that AMD is indeed pricing their top end SKU comparably to an i7-6850k, then that's roughly the level of performance that should be expected from it. (Probably a bit better in multi-threaded workloads, which is balanced by being worse in single-threaded.) Which, as usual, doesn't bode well for the consumer markets where usage scenarios beyond 4 cores are few and far between.

Then how are they gonna sell? You can bet your dollars that these prices br based on focus groups, how much stock they have to sell and at what margins.
How are you gonna tap someone with a quad sandy or later for a new system? You want that with double cores sir? Same price, yes sir. I can see that strategy. Market penetration and they've got inventory to sell.

 

[Doom2pro]

We don't really know squat at this point in regard to base or boost... There is lots of conflicting information on base and boost, only SKU and Prices are consistent.

How can anyone say 65W be out of the question if we don't even know base-boost for that SKU or binning parameters?

 

[Agent-47]

So IMHO The base clock of Ryzen 7 1700 will be 3,0 GHz at best, but that is IMHO still not likely, I expect 2,8-2,9 GHz to be safe.

without knowing the characteristics of the chips its really hard to say. we have already seen impressive perf/watt with polaris at lower clocks. it might be possible that ryzen inherits that advantage.

Also, CPC asserted that there is no 65W 8c16t SKU.

 

[JDG1980]

That base clock is IMO too high. Take for example Trinity.

The 95W A10-5800K had base clock 3,8 GHz with turbo boost 4,1 GHz. The 65W SKU that was based on it (A10-5700) had base clock 3,4 GHz with 4,0 GHz max boost. The base clock penalty in other words was more than 10 %. But it was also a second generation 32nm product (after Llano - or third if we count Zambezi).

So IMHO The base clock of Ryzen 7 1700 will be 3,0 GHz at best, but that is IMHO still not likely, I expect 2,8-2,9 GHz to be safe.

I don't think you can extrapolate the frequency/power curve of 14nm FinFET from what was seen on the comparatively ancient 32nm SOI.

 

[Nachtmaer]

I really hate AMD's marketing. Why continue to go with 7/5/3 naming scheme? This is a brand new platform, these idiots need to break free from the old AMD and quit copying Intel in their naming schemes (who copied BMW lol)

Someone else brought this up a while ago too, but it would've made more sense if they went with R(yzen) 8/6/4. That way they continue their one-up game with Intel like they did with their chipset naming and it'd also reflect physical core count of their SKUs.

 

[Dresdenboy]

Is it just me or does the naming scheme makes no sense sense at all. I mean why call two 8c16t cpus 1700 and 1800 over a 200MHz clock difference and then call a 6c12t core 1600? what if they realize a new 6c12t with a 200Mhz bump, are they going to call it 1700-6c??

Also I am not sure a 200MHz clock difference between 1800X and 1700X justify a £90 price difference considering both are unlocked with XFR. maybe £30 at most. I along with most people in this thread would probably go for the 1700X instead.

The 6C name could be OK for this line up. Next will be 2xxx. And there might be not such a long time till then. Also 1650 would be possible.

The price delta is for Zen with 8C+SMT, so it lifts the performance of that, not of a 8C Vishera. And the amount of dies you get for this SKU are likely much less than for the lower SKUs due to binning.

Gesendet von meinem GT-I9300 mit Tapatalk

 

[jihe]

Is it just me or does the naming scheme makes no sense sense at all. I mean why call two 8c16t cpus 1700 and 1800 over a 200MHz clock difference and then call a 6c12t core 1600? what if they realize a new 6c12t with a 200Mhz bump, are they going to call it 1700-6c??

Also I am not sure a 200MHz clock difference between 1800X and 1700X justify a £90 price difference considering both are unlocked with XFR. maybe £30 at most. I along with most people in this thread would probably go for the 1700X instead.

I call BS.

Dude, Intel charge $180 for 200Mhz and some artificially gimped PCI-e lanes.