Should AMD release an updated 8 core FX (improved cores, smaller process)?

[Justinbaileyman]

I voted yes as I would love to see something, anything from AMD to hold us over till ZEN but I fear AMD is dead in the water. I am even starting to have my doubts we will even see ZEN released at all. The only reason I dont switch over to Intel completely is the cost,all there products are way way way overly priced for poor people like me

 

[VirtualLarry]

The only reason I dont switch over to Intel completely is the cost,all there products are way way way overly priced for poor people like me

I hear you there. The only "affordable" Intel CPU worth using is the G3258, and they destroyed that value, by forcing CPU microcode updates that block overclocking on anything but their most expensive chipset-based boards.

AMD likes to create value for their customers (Mantle, FreeSync, etc.), while Intel likes to destroy it (forced product crippling / segregation, charging an effective tax on overclocking, by locking it down, etc.).

 

[ShintaiDK]

I voted yes as I would love to see something, anything from AMD to hold us over till ZEN but I fear AMD is dead in the water. I am even starting to have my doubts we will even see ZEN released at all. The only reason I dont switch over to Intel completely is the cost,all there products are way way way overly priced for poor people like me

You will pay how it performs.

If Zen performs like the hype. Then it will be very expensive. if it doesn't, then it wont.

If you need an upgrade you can just as well get it now.

 

[AtenRa]

There is zero chance of them making enough money off it to cover the cost. So no, they absolutely should not and will not do something like this. Unless somebody comes to them with a large chunk of money and says we'll buy 50000 pieces if you can reduce the TDP by 25% while providing the same performance at whatever specific application they need. AND also assuming that AMD engineers were able to tell the sales guys that this could actually be done.

In order for AMD and Intel to spend time and resources to create a 8-core CPU they need to be able to sell them to the Server market.
A 28nm planar 8-Core SteamRoller/Excavator is not competitive against Intels 22-14nm server chips in 2015.

If AMD had available the 22nm SOI process in 2014, they could make such a chip but not at 28nm planar today.

 

[Justinbaileyman]

Personally I don't think AMD FX cpu's are all that bad they could use some tweaking here and there, but AMD acts like they have ants in there pants and keep bouncing around from architecture to architecture in stead of staying with what they have and improving upon it. Maybe AMD should have just dug in and kept the Phenom II Architecture and added more cores, more cache L3 and L4, and went from there? Same goes for the FX line they should of just stuck with it and improved upon what they have now. They could have easily Increased IPC count and added more instruction sets and tweaked the shit out of it including dropping to TDP and everything would be even stevens when it comes to competition with Intel. Or Am I just in dream land?

 

[NTMBK]

Personally I don't think AMD FX cpu's are all that bad they could use some tweaking here and there, but AMD acts like they have ants in there pants and keep bouncing around from architecture to architecture in stead of staying with what they have and improving upon it. Maybe AMD should have just dug in and kept the Phenom II Architecture and added more cores, more cache L3 and L4, and went from there? Same goes for the FX line they should of just stuck with it and improved upon what they have now. They could have easily Increased IPC count and added more instruction sets and tweaked the shit out of it including dropping to TDP and everything would be even stevens when it comes to competition with Intel. Or Am I just in dream land?

You're in dream land. Your advice is like telling Intel to just knuckle down and stick with Netburst. If the design has fundamental flaws, all the tweaking in the world won't help it.

 

[dark zero]

I hear you there. The only "affordable" Intel CPU worth using is the G3258, and they destroyed that value, by forcing CPU microcode updates that block overclocking on anything but their most expensive chipset-based boards.

AMD likes to create value for their customers (Mantle, FreeSync, etc.), while Intel likes to destroy it (forced product crippling / segregation, charging an effective tax on overclocking, by locking it down, etc.).

Sorry, but this is the monopoly. That’s why I really want a software/gaming crash, far worse than 1983 in order to purge the industry hard from that monopolies and the mediocrity from them and maybe to resurrect some old formats.

Remember than after the gaming crash, games improved dramatically until few years ago.

 

[AtenRa]

Personally I don't think AMD FX cpu's are all that bad they could use some tweaking here and there, but AMD acts like they have ants in there pants and keep bouncing around from architecture to architecture in stead of staying with what they have and improving upon it. Maybe AMD should have just dug in and kept the Phenom II Architecture and added more cores, more cache L3 and L4, and went from there? Same goes for the FX line they should of just stuck with it and improved upon what they have now. They could have easily Increased IPC count and added more instruction sets and tweaked the shit out of it including dropping to TDP and everything would be even stevens when it comes to competition with Intel. Or Am I just in dream land?

AMD has done exactly that, Bulldozer to Piledriver to Steamroller to Excavator.

The only reason you havent seen a 8-core after Piledriver is the lack of availability of a better process than 32nm SOI.

 

[Shehriazad]

Also...didn't they relatively recently enhance the FX-E series? I'm pretty sure that is all we can expect.

Hell...they haven't even released the 870K on FM2+ xD

 

[superstition]

Also...didn't they relatively recently enhance the FX-E series? I'm pretty sure that is all we can expect.

Some think that lower-leakage transistors were used but others aren't so sure and just think the E series reflects a more mature 32nm process and are simply lower-clocked chips that are binned (the 8370E in particular being the most stringently binned).

The tight binning theory, though, isn't necessarily borne out by the tests, though. The "95W" 8320E came in at 86 watts for one site (maybe this one). The "95W" 8370E came in at 117 or so as I recall.

Of course, there are no 88W 4970K chips either. The 88 number is marketing magic, regardless of what its defenders say.

 

[superstition]

Need 8 threads but don't have a lot of money and possibly want to play the newest games with it? FX 8XXX will do just fine....since it'll still be able to deliver 60+ fps on the vast majority of new games. (At least if they're optimized enough to actually use more than 2 threads...which ARE the vast majority of new releases nowadays.

I don't know how much I'd count on developers really utilizing threads.

The Ashes results, for instance, clearly point to a game using 4 or fewer threads. In fact, the results suggest, to me at least, that they're optimizing for 2 cores — given that a 3.5 GHz i3 delivered 40 fps at 1600p on high with a 390X but an 8370 was not particularly competitive. Overclocked to 4.5 the FX might match the i3 at 3.5 GHz but how exciting is that? It's OK, though, for people doing Blender, Handbrake, and other things that benefit from having more cores.

With the shrinking of PC sales devs are probably going to target dual cores like i3s more than they otherwise would going forward. Anything more than 4 cores is probably not going to be very helpful. It's sad that even an RTS DX12 title is performing significantly better on a slow i3 than on an 8 core FX. (And, Intel's 5960X is beaten by a 6700K, so it's not just the FX architecture that's the issue. The issue is that the game apparently doesn't take advantage of 8 cores.)

DX12 looks to be a major boon for AMD's 290/390 customers but not so much for its processors.

 

[superstition]

I voted yes as I would love to see something, anything from AMD to hold us over till ZEN but I fear AMD is dead in the water. I am even starting to have my doubts we will even see ZEN released at all.

The delay is what bugs me. It gives Intel so much time to stay out in front.

 

[DrMrLordX]

At 20nm SOI or 14nm FF I would sure would like an 8-Core Excavator, not at 28nm though.

Actually, the new GV-A1 stepping would make it somewhat-tolerable. Or you could use the same process variant used in Carrizo, depending on what you wanted.

Consider Carrizo. It has superior voltage scaling with clockspeed to GV-A1 Kaveri up to about 2.5 GHz. Beyond that, GV-A1 Kaveri is better, thanks to HDL on Carrizo and all that.

At speeds of 2.5 GHz, Carrizo burns up maybe 42W when strapped to a 500-600 MHz 512-shader GCN iGPU. Eliminate the iGPU and replace it with two more XV modules. The die size should be smaller, and your power usage is probably going to be in the 50-70W range. Bear in mind that's a 4M/8T chip at 2.5 GHz. Add in Carrizo's more-sophisticated-than-Kaveri power management, and you could get some fairly sophisticated turbo settings on that chip. It would be a low-price competitor to Broadwell-D. The performance/watt would be inferior, but it would be better than the empty void that AMD has on the market right now.

Or, consider XV with the same refinements put into the GV-A1 stepping (assuming some/all of those refinements are applicable; that may be an erroneous assumption). XV should exhibit similar/identical voltage scaling as compared to GV-A1 Kaveri in that configuration. They could probably get a 4m/8t XV running around 3.4 GHz in a TDP of maybe 80-100W. I'm leaning closer to 80W but maybe I'm just being optimistic. Again, many would point out that such a part might not be competitive with (insert name of Intel CPU here), but it's better than the gaping void in the AMD lineup out there right now. The only problem I see with this particular scenario is that the die size might be too great.

What you would not get is something like a 4m/8t XV chip hitting 5 GHz with power consumption lower than that of PD at the same clockspeed. You probably wouldn't see clocks over 4.5 GHz with any degree of reliability. It would be a non-starter for most enthusiasts. AMD would have to position the product elsewhere.

That being said, AMD has too few resources and too little time to bring such a product to market. Hindsight is 20/20. Had AMD known how Carrizo would turn out, or had they known how GV-A1 stepping would affect Steamroller, maybe they could have started work sometime last summer on a stopgap XV solution with simplified power planes vs. Carrizo (read: no iGPU) that would work on AM3+. I do not think AM4 is ready, and with 2016 being so close to the Zen launch, I don't think they want to invest in a 4m XV chip that would launch with AM4.

AMD would be better served in getting an EX core APU into AM4 (or whatever zen's platform will be called) with a proper die shrink down to 14/16FF along with whatever new graphics AMD has planned for that node. 2017 is a long ways off.

I agree with that, though again, there is a question of what it is they can do with their remaining resources and talent. The Zen team seems to be sucking up all the oxygen these days. It would make sense to port XV to GF's 14nm as a test-case.

I don't even think you will see socket Carrizo either. AMD's only real new product in 2016 might be Arctic Islands.

If we see anything from AMD, it will probably be 28nm Bristol Ridge launching AM4 in Q3 2016. Zen follows up in Q4 2016 with limited supply and/or Q1 2017 in greater quantity . . . assuming AMD lasts that long.

 

[cbn]

Consider Carrizo. It has superior voltage scaling with clockspeed to GV-A1 Kaveri up to about 2.5 GHz. Beyond that, GV-A1 Kaveri is better, thanks to HDL on Carrizo and all that.

Didn't the Stilt's testing show 6% lower power consumption for Carrizo at 2.6 Ghz and 9% higher power consumption at 3.4 Ghz (compared to Kaveri).

So maybe the actual point where the voltage scaling with clockspeed are equal is closer to 3Ghz?

P.S. Carrizo also has 5% to 10% better IPC (and although this is unrelated to voltage scaling with clockspeed argument, it does factor into overall performance per watt).

 

[cbn]

as for 6 cores or more fm2+, I think AMD would not want to use resources for that because they want the same die to be used for laptops, and over 4 cores there is no market for laptops? they are better using the space for their IGPs I guess

Six cores in a laptop would be sweet:

http://www.anandtech.com/show/9518/the-mobile-cpu-corecount-debate

When I started out this piece the goals I set out to reach was to either confirm or debunk on how useful homogeneous 8-core designs would be in the real world. The fact that Chrome and to a lesser extent Samsung's stock browser were able to consistently load up to 6-8 concurrent processes while loading a page suddenly gives a lot of credence to these 8-core designs that we would have otherwise not thought of being able to fully use their designed CPU configurations.
[...]
What we see in the use-case analysis is that the amount of use-cases where an application is visibly limited due to single-threaded performance seems be very limited. In fact, a large amount of the analyzed scenarios our test-device with Cortex A57 cores would rarely need to ramp up to their full frequency beyond short bursts (Thermal throttling was not a factor in any of the tests). On the other hand, scenarios were we'd find 3-4 high load threads seem not to be that particularly hard to find, and actually appear to be an a pretty common occurence. For mobile, the choice seems to be obvious due to the power curve implications. In scenarios where we're not talking about having loads so small that it becomes not worthwhile to spend the energy to bring a secondary core out of its idle state, one could generalize that if one is able to spread the load over multiple CPUs, it will always preferable and more efficient to do so.
[...]
In the end what we should take away from this analysis is that Android devices can make much better use of multi-threading than initially expected. There's very solid evidence that not only are 4.4 big.LITTLE designs validated, but we also find practical benefits of using 8-core "little" designs over similar single-cluster 4-core SoCs. For the foreseeable future it seems that vendors who rely on ARM's CPU designs will be well served with a continued use of 4.4 b.L designs.

Maybe 3 to 5 watts per Excavator module (for a 15W hexcore Excavator SKU):

 

[DrMrLordX]

Didn't the Stilt's testing show 6% lower power consumption for Carrizo at 2.6 Ghz and 9% higher power consumption at 3.4 Ghz (compared to Kaveri).

So maybe the actual point where the voltage scaling with clockspeed are equal is closer to 3Ghz?

P.S. Carrizo also has 5% to 10% better IPC (and although this is unrelated to voltage scaling with clockspeed argument, it does factor into overall performance per watt).

The Stilt flat-out said that the crossing point is 2.6 GHz. If he misinterpreted his own data, then so be it.

In the limited testing I've seen, Carrizo has had up to 21% better IPC than Kaveri. 5-11% seems to be the more common range. Therefore, I think that using the 20/20 of hindsight should tell us that AMD should have launched low-power, low-clockspeed 4m XV parts in the 2-2.5 GHz range instead of the Seattle vaporware. I'm not sure that 4m XV at higher clockspeed using standard 28nm with the refinements of the GV-A1 stepping would have worked out for them.

It's too late now for either chip.

 

[SlowSpyder]

You're in dream land. Your advice is like telling Intel to just knuckle down and stick with Netburst. If the design has fundamental flaws, all the tweaking in the world won't help it.

The original plans, with four Excavator modules on an improved manufacturing process, with L3 (especially if they improved it's performance) might not be bad at all. But, I think when they fell behind schedule, AM3+ aged. How many AM3+ users are out there compared to users with Intel sockets over the last several years? As much as I'd like to see it, I'm not sure there'd be money to be made, but I don't think it'd be a 'bad' part by a long shot.

 

[mrmt]

The original plans, with four Excavator modules on an improved manufacturing process, with L3 (especially if they improved it's performance) might not be bad at all. But, I think when they fell behind schedule, AM3+ aged. How many AM3+ users are out there compared to users with Intel sockets over the last several years? As much as I'd like to see it, I'm not sure there'd be money to be made, but I don't think it'd be a 'bad' part by a long shot.

Since there wouldn't be money to be made this would be a bad part by definition.

 

[cbn]

Didn't the Stilt's testing show 6% lower power consumption for Carrizo at 2.6 Ghz and 9% higher power consumption at 3.4 Ghz (compared to Kaveri).

So maybe the actual point where the voltage scaling with clockspeed are equal is closer to 3Ghz?

P.S. Carrizo also has 5% to 10% better IPC (and although this is unrelated to voltage scaling with clockspeed argument, it does factor into overall performance per watt).

The Stilt flat-out said that the crossing point is 2.6 GHz.

You are right.

According to the following post the 6% lower power consumption corresponded to 2.1 GHz (not 2.6 Ghz), with 9% higher power consumption at 3.4 Ghz:

http://www.overclock.net/t/1560230/jagatreview-hands-on-amd-fx-8800p-carrizo/520#post_24343181

 

[Puffnstuff]

I'd like to see amd release an extremely competitive cpu that is strong enough to force intel to innovate even further.

 

[dark zero]

Sorry but that Cpu won't even come from AMD who are deader than Disco.

It.will come from Apple

 

[coffeemonster]

with 2016 being so close to the Zen launch, I don't think they want to invest in a 4m XV chip that would launch with AM4.

Since they hyped Carrizo and released a mullins refresh called Carizzo-L, they could Release the 4m XV chip and call it Zen-L or something as a stopgap until the real Zen release

 

[DrMrLordX]

Zen-L huh. That or call it Excavator+.

 

[SlowSpyder]

Since there wouldn't be money to be made this would be a bad part by definition.

I meant from an overall performance, satisfaction of use stand point. Obviously it's a bad part from a business perspective, and doesn't exist for a reason.

 

[Killrose]

Maybe Via will make a CPU based on the AM3+ Socket.

Wait a minute.., I have an AM3+ system that's not very funny Maybe the new Zen could still drop into the AM3+ socket like an AM2 CPU will and still use DDR3. Oh come on, why not?