AMD Bristol/Stoney Ridge Thread

Page 57 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DrMrLordX

Lifer
Apr 27, 2000
21,582
10,785
136
Is the GDDR latency really that bad? I was under the impression that while it was higher in cycle count, because the frequency was so much higher that the latency in NS was kind of a wash.

I looked into it and at least GDDR5 doesn't seem that bad. Apparently it's more often the controller than the RAM itself that is the underlying cause of higher latencies in GDDR deployments. The GDDR5 in the PS4 apparently had CL of around 15.
 
  • Like
Reactions: NTMBK

burninatortech4

Senior member
Jan 29, 2014
661
369
136
Technically Carrizo not Bristol Ridge, but seems relevant to this thread: https://www.anandtech.com/show/14368/biostar-a10n-8800e-motherboard-review

I can't seem to find that board for less than $99 US. It doesn't make any sense at that price with the A300/W existing and 2200GE and 200GE going for less than $70 and $45, respectively, on eBay. Excavator is really long in the tooth at this point.

If the price dropped closer to $50 then …. maybe I could see using it for a grandma style office machine.
 

NTMBK

Lifer
Nov 14, 2011
10,208
4,940
136
I can't seem to find that board for less than $99 US. It doesn't make any sense at that price with the A300/W existing and 2200GE and 200GE going for less than $70 and $45, respectively, on eBay. Excavator is really long in the tooth at this point.

If the price dropped closer to $50 then …. maybe I could see using it for a grandma style office machine.

Yeah, I just don't get how it makes sense at that price. The CPU is too weak compared to closely priced Zen systems. The big GPU is nice, but it's so hamstrung by the low TDP, poor CPU, and DDR3 that it can't shine in games.
 

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
Yeah, I just don't get how it makes sense at that price. The CPU is too weak compared to closely priced Zen systems. The big GPU is nice, but it's so hamstrung by the low TDP, poor CPU, and DDR3 that it can't shine in games.

The benchmarks where the 8800 lose badly are those who are not that significant for this kind of product, dunno if Blender is to be used even with the Athlon set up, in Integer tests it not that bad, also for a system that is always on it has 7W lower idle comsumption, not sure that a 1200W PSU is adequate for such test but with a lower power one this thing can be brought at 10W or so if not lower.
 

burninatortech4

Senior member
Jan 29, 2014
661
369
136
The A10N-8800E is selling for $69 to $77 currently in markets Biostar actually aims at.It is DDR4.

It is DDR4 but it maxes out at 2133mhz. Bristol Ridge's memory controller can't do past 2400mhz and I'm not convinced that board will even give you a 2400mhz option.
 

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
The A10N-8800E is selling for $69 to $77 currently in markets Biostar actually aims at.It is DDR4.
So approx 75usd vs 115usd for a zen solution.
40 usd difference.
This product is relevant in households with an income of aprox 15 usd a day as the technology wonder in the hut together with the moped.
That's actually an income group that is the most prevalent on this earth.
So lets say you save up a dollar a day for your new rig. Is it worth waiting one and a half month to get zen.
Yes. Because it will last for 2 more years.

Lets see. Imo 14 and 12nm zen will flood the market. The 40 usd difference will shrink to 20 in half year. 10 usd in a year and 5 in two.

At that time stoney can rest in peace with Atom.
 
  • Like
Reactions: VirtualLarry

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
Carrizo is 2x Stoney, by the GPU and by the CPU, and well better than any Atom, besides that s a ITX form factor, dunno how much is the equivalent AM4 ITX + Athlon 200GE, but certainly not 115$.

Looking again at the benches they give an innaccurate idea as there s several that are mainly single threaded, hence giving a big advantage to the Athlon, but total throughput is not that bad when all threads are more or less loaded, this set up will last as much as an Athlon 200GE :


91396.png


91397.png

91398.png


91401.png

91399.png

91402.png


https://www.anandtech.com/show/14368/biostar-a10n-8800e-motherboard-review/6
 
  • Like
Reactions: Tlh97 and amd6502

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
Carrizo is 2x Stoney, by the GPU and by the CPU, and well better than any Atom, besides that s a ITX form factor, dunno how much is the equivalent AM4 ITX + Athlon 200GE, but certainly not 115$.

Looking again at the benches they give an innaccurate idea as there s several that are mainly single threaded, hence giving a big advantage to the Athlon, but total throughput is not that bad when all threads are more or less loaded, this set up will last as much as an Athlon 200GE :


91396.png


91397.png

91398.png


91401.png

91399.png

91402.png


https://www.anandtech.com/show/14368/biostar-a10n-8800e-motherboard-review/6

None of these test is representative of what such a machine would do. User experience in real world is what matters and stoney and carizo is just bad here. Normal browsing media and office use is just nearly half speed. Every second. Every click. While zen is building up brand stoney will erode it.
 
  • Like
Reactions: NTMBK

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
There s no browser real time like test in the review, Mozilla Kraken is using several tasks, so what are those browsing related tests that show half speed..?

And one more time, what about the price that you said is 115$ for an AM4 ITX + A200GE.?

In Germany at the cheapest a A10N-8800E is 80€ while an AM4 ITX + A200GE will cost you roughly 145€ if taken at the same retailer...
 

amd6502

Senior member
Apr 21, 2017
971
360
136
Considering that the A10 is running at 15W these results compare very nicely; carizzo is doing splendid there. I mean for someone who wants something like a passive cooled silent desktop this is a really nice option; or for someone set on a super small SFF ITX form factor. This makes for a fairly decent bare basics programming/email/www station.

It would be nice though if they allowed some upward cTDP. Personally the passive cool option isn't very important for me and I'd like higher base frequencies; would prefer to run somewhere around 30W or so. Here is a lower binned chip at 35W http://www.cpu-world.com/CPUs/Bulldozer/AMD-A10-Series PRO A10-9730B.html (runs at 2.8-3.5ghz).


Carrizo is 2x Stoney, by the GPU and by the CPU, and well better than any Atom, besides that s a ITX form factor, dunno how much is the equivalent AM4 ITX + Athlon 200GE, but certainly not 115$.

Looking again at the benches they give an innaccurate idea as there s several that are mainly single threaded, hence giving a big advantage to the Athlon, but total throughput is not that bad when all threads are more or less loaded, this set up will last as much as an Athlon 200GE :


91396.png


91397.png

91398.png


91401.png

91399.png

91402.png


https://www.anandtech.com/show/14368/biostar-a10n-8800e-motherboard-review/6
 
Last edited:

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
Heh, wonder from whom they stole that process.
It isn't as simple. Its legal and it's worse.
What I have seen is technology transfers as part of an access to the Chinese market. So its state controlled and approved by the ones doing the transfer. Top management. They dont steal it. They get it. You just dont steal such technology. Its transferred legally.
So in this sense the communist state controlled system is more effective than a free market based system where eg short term bonuses for sale makes long term loss for innovation and loss for all.
 

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
Yeah, I know, which is part of the reason for the trade war etc. etc. Just wondering who got pushed into divulging that tech: GF, Samsung, or . . .?

They poached from Samsung the famed guy who was previously poached at TSMC....
 

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
Considering that the A10 is running at 15W these results compare very nicely; carizzo is doing splendid there. I mean for someone who wants something like a passive cooled silent desktop this is a really nice option; or for someone set on a super small SFF ITX form factor. This makes for a fairly decent bare basics programming/email/www station.

It would be nice though if they allowed some upward cTDP. Personally the passive cool option isn't very important for me and I'd like higher base frequencies; would prefer to run somewhere around 30W or so. Here is a lower binned chip at 35W http://www.cpu-world.com/CPUs/Bulldozer/AMD-A10-Series PRO A10-9730B.html (runs at 2.8-3.5ghz).


Actually the 8800FX has some kind of turbo boost extending the TDP at 25W for short durations, but seems that the turbo is tied to frequency and temperature at the same time, so only high IPC code can push the power envelloppe, this can be seen with power under Prime 95, and also POVRay score where the 8800 match an Athlon 200GE while lagging in about everything else.

The 4C 9000 series are Bristol Ridge, although they should make their way to Biostar BGAs MB sooner or later, they benefit from a better 28nm process and have better perf/Watt, as you point it it would be a valuable improvement for those compressed costs SFFs, 35W A10-9800E Bristol Ridge is about on par with an Athlon 200GE, albeit with a little more actual TDP, around 30-35W and 20-25W respective power comsumptions.
 
  • Like
Reactions: amd6502

NostaSeronx

Diamond Member
Sep 18, 2011
3,683
1,218
136
The 4C 9000 series are Bristol Ridge, although they should make their way to Biostar BGAs MB sooner or later
I think Biostar would go for Raven2. There is the Atari VCS potentially and other demand for it reducing its cost to a bulker like Biostar. Making it more ideal for Biostar, if you look at A10N-8800E motherboard, it is pretty cheap.

http://www.biostar-usa.com/app/en-us/mb/socket.php?S_ID=19
Only Carrizo and Carrizo-L only have one model, while the other ones have more than one model.

Top-end with Ryzen 3 3200u and lower-top end with Athlon 300u.

Bristol wants two rails at 35A, Raven/Raven2 only needs one rail at 45A. With that FP5 has reduced pin count and reduced overall size, etc. More expensive SoC with less expensive motherboard, should work out in replacing Carrizo FP4. Rather, than moving to Bristol and ruining the cost dynamic Biostar built. I'm pretty sure Raven2(15W) has been pointed out for the workloads Biostar is selling to of it beating Bristol(15W).

////
New 22FDX boosters incoming.
22FDX(2019+) expects ~1.10x increase in performance from pre-boost 22FDX.
22FDX+(2020?) expects ~1.16x to ~1.20x increase in performance from pre-boost 22FDX+. <- FinKiller®
The boosters should be more effective on BOX15 12FDX, but the numbers are only for BOX20 22FDX and BOX15 22FDX+.
 
Last edited:

dark zero

Platinum Member
Jun 2, 2015
2,655
138
106
Sadly Bulldozer has nothing else to do... not even 25% more performance by node shrink can save it.
 

NostaSeronx

Diamond Member
Sep 18, 2011
3,683
1,218
136
Sadly Bulldozer has nothing else to do...
Micro-op queue, 2 BTB branch prediction(branch calculated per core queue), CX/AX; complex execution(iMUL/iDIV/CRC/Branch+ALUs) and address execution(AGUs+ALUs), more flexible FPU, L0 caches, etc. However, most of that stuff is more ideal on 12FDX which is to 14nm FinFET as 22FDX is to 28nm.
not even 25% more performance by node shrink can save it.
The 22FDX design wouldn't be a 25% more performance node. It would be closer to a 50% or higher frequency node. The 22FDX node re-uses the 20FD/14FD node transistor, which is faster than 20BLK strained(20SHP to GlobalFoundries). What isn't gained in density is gained in absolute and nominal performance which directly translates to frequency.

1V is the safest voltage to expect a 50% increase and ~0.75V with a ~75% increase at minimum;
A9-9420e => 2.7 GHz @ 1V || +50% = 4.05 GHz // 1.8 +75% = 3.15 GHz
A6-9220e => 2.4 GHz @ 1V || +50% = 3.6 GHz // 1.6 +75% = 2.8 GHz
A4-9120e => 2.2 GHz @ 1V || +50% = 3.3 GHz // 1.5 +75% = 2.63 GHz
^This boost is achievable through a minimally redesigned port of 28HPA/9T/114CPP/90Mx to 22FDX/8T/104CPP/80Mx. The lower variability of 22FDX over 28BLK especially strained BLK(28SHP/28A/28HPA) can allow higher frequency at lower power do to a better product sign-off.

The biggest gain 22FDX brings over is a more scalable LVT and a much faster RVT transistor. Which is important for the CPU and GPU. I think the UVD/VCE would also benefit from the opposite with RVT being less leaky at same speed.

However, it only works if AMD gives us a Family 15h + Graphic Core Next + UVD/VCE SoC on 22FDX in the frame of Stoney. If AMD decides to do post-Family 15h(still CMT) + Single-CU WGP RDNA(still 3 CU) + VCN, somewhat like Kaveri(Stoney) to Carrizo(ex: Ernst). Then, some of it has to be reworked to the potential differences in Vt selection, ratio-ed channel lengths, different work units, etc.

However, when has AMD ever not hoisted up a GloFo node themselves. We obviously should be looking at a post-Family 15h w/ (single-cu ver)RDNA+VCN2 /etc. on 12FDX. With an absurd die size of less than 100 mm squared; between 50mm2 and 100mm2. For a price-sensitive, cost-effective, power-optimized, tiny high-performance APU SoC for the pocketbook market, or whatever. (GPD/Chuwi/etc but with AMD as the product focus point and below $399, instead of these $799 intel monstrosities.)

====
Edit, Sep 17;
With the http://www.rock-chips.com/a/en/products/RK18_Series/2019/0529/989.html

Which is on GlobalFoundries' 22FDX, at any month can the AMD Stoney revision feasibly launch.
Which no one told me popped up around the same time period as Synpatics;

Oof~
china.png

Edit: Octo. 17
2020 is looking like another Stoney Refresh on 28nm.
 
Last edited:
  • Like
Reactions: amd6502

amd6502

Senior member
Apr 21, 2017
971
360
136
Now there is this proposal to mix big and small physical cores and share threads between them in this way.

This might mean a return to a Piledriver like core, or maybe slightly more likely, a Jaguar like core with tiny transistor count.
 

NTMBK

Lifer
Nov 14, 2011
10,208
4,940
136
Now there is this proposal to mix big and small physical cores and share threads between them in this way.

This might mean a return to a Piledriver like core, or maybe slightly more likely, a Jaguar like core with tiny transistor count.

Nah, the patent refers to "low power", so that rules out Bulldozer.