AMD Trinity A10 5700 review

[raghu78]

http://hexus.net/tech/reviews/cpu/47257-amd-a10-5700/

This CPU is a very good for a low power HTPC and entry level gaming rig. Hope AMD designs Kaveri with stacked DRAM and releases in H2 2013. Now that should be really nice. :biggrin:

 

[Insert_Nickname]

Thanks for the link, was looking forward to a review of the 65W trinitys. Looks like my main media-centre is in for an upgrade in the near future...😀:thumbsup:

 

[SPBHM]

that's pretty decent I think.

 

[scajjr2]

I just upgraded our HTPC to an A10-5700/Asus F2A85-M-CSM. Fast, quiet, uses less power than the A8-3850 it replaced.

Sam

 

[frozentundra123456]

that's pretty decent I think.

Decent perhaps, for a htpc, but not a very complete review. Only two games tested, neither of which is graphically demanding. And it makes no sense to test it with a gtx 680. Why would you buy a chip whose only superior feature is the igp and then disable it to use a discrete card. Even the a10 is very limited in gaming, and this will be even more so by the lower tdp.

 

[nismotigerwvu]

Decent perhaps, for a htpc, but not a very complete review. Only two games tested, neither of which is graphically demanding. And it makes no sense to test it with a gtx 680. Why would you buy a chip whose only superior feature is the igp and then disable it to use a discrete card. Even the a10 is very limited in gaming, and this will be even more so by the lower tdp.

Looks like Batman was demanding enough to bring the I3 iGPU to its knees.

 

[Silverforce11]

Looks like Batman was demanding enough to bring the I3 iGPU to its knees.

And the I7..

intel iGPU is a waste of die space.

 

[AtenRa]

Just food for thought,

Trinity at 32nm = 246mm2 1.3B transistors.
For simplicity, dual Module occupies half of the die.

Now just imagine what would happen if they would have a 22nm process next year.

1: They could have a 22nm Trinity at ~130mm2 with 30-50% less power
That would put them in the same die size as Core i3, same CPU performance but with 100% more iGPU performance and close the same power consumption.

2: They could have a Quad Module (8 cores SR) plus 384 GCN Shaders at ~190mm2
That would be a FX8350 + 15% more performance from SteamRoller cores + Trinitys iGPU but with GCN shaders and lower consumption (65watts ??)

3: They could have a Quad Module (8 cores SR) plus 512 GCN shaders at ~250mm2 and same TDP as Trinity.
That would have SteamRoller performance + 50% more iGPU performance than Trinity.

4: Dual Module (SteamRoller) + 512 GCN Shaders at ~160mm2.
The iGPU would get more space than the CPU cores. 15% more CPU performance + 50% more iGPU performance at 65W TDP.

I believe number 4 is the 28nm Kaveri at 100W TDP and close to ~220mm2 die size.

Just for reference, Intel 4C Ivy at 22nm = 160mm2 (GT2)

 

[frozentundra123456]

Looks like Batman was demanding enough to bring the I3 iGPU to its knees.

True, without a discrete card, and also true for the i7. But what is the relevance to my post? I never said the any intel igpu was sufficient for gaming. I notice that you were not able to refute what I said, but just switched the topic to criticizing intel without a response which was relevant to what I said.

 

[Centauri]

Does anybody know what the stock vcore on the 5700 is?

 

[Greenlepricon]

Just wondering what happens if you can get away with undervolting/underclocking the 5800k. Would you get similar results?

 

[Nemesis 1]

And the I7..

intel iGPU is a waste of die space.

You should be the head of sales at intel . Straighten them miss guided fools out .

 

[SPBHM]

Just for reference, Intel 4C Ivy at 22nm = 160mm2 (GT2)

well, ivy have 8MB of l3, I think it easily uses as much space as 2 cores.

 

[SithSolo1]

Tigerdirect(Free Shipping) and Frys have this thing in stock for $130.
NCIX.us has it for $138

Sadly thats as much or more than a 5800K

Also, seems like Superbizz has the Asrock FM2 itx board instock for $104 shipped: http://www.superbiiz.com/detail.php?name=MB-F2A75IX
Only US retailer I found with the board in stock.

 

[Yuriman]

Just food for thought,

Trinity at 32nm = 246mm2 1.3B transistors.
For simplicity, dual Module occupies half of the die.

Now just imagine what would happen if they would have a 22nm process next year.

1: They could have a 22nm Trinity at ~130mm2 with 30-50% less power
That would put them in the same die size as Core i3, same CPU performance but with 100% more iGPU performance and close the same power consumption.

2: They could have a Quad Module (8 cores SR) plus 384 GCN Shaders at ~190mm2
That would be a FX8350 + 15% more performance from SteamRoller cores + Trinitys iGPU but with GCN shaders and lower consumption (65watts ??)

3: They could have a Quad Module (8 cores SR) plus 512 GCN shaders at ~250mm2 and same TDP as Trinity.
That would have SteamRoller performance + 50% more iGPU performance than Trinity.

4: Dual Module (SteamRoller) + 512 GCN Shaders at ~160mm2.
The iGPU would get more space than the CPU cores. 15% more CPU performance + 50% more iGPU performance at 65W TDP.

I believe number 4 is the 28nm Kaveri at 100W TDP and close to ~220mm2 die size.

Just for reference, Intel 4C Ivy at 22nm = 160mm2 (GT2)

Your math is a little off.

When Intel moved from 32nm to 22nm, they went from 149mm² (2c Sandy) to 94mm² (2c Ivy) and from 216mm² (4c Sandy) to 160mm² (4c Ivy).

If AMD scales down similarly, Trinity would be 155-180mm², not 130mm², compared to the i3 which is 94mm².

Vishera would be 200-230mm² WITHOUT an iGPU, not a chance it'll get near 180mm² with 384 GCN cores, competing with 160mm² i7's with HD4k graphics (or i5's if you're comparing a cutdown version of the chip).

Interestingly, Trinity even at 22nm would be bigger than an Ivy Bridge i7, because it's already bigger than a Sandy Bridge i7.

I think Intel did very well considering their transistor and die size budget.

 

[nismotigerwvu]

Not all nodes (22nm or otherwise) are created equal. Intel's processes are top notch and I believe (but might need to go check some sources on this) the transistor density of even Intel's 45nm node was better than GF's 32nm. Regardless, transistors on a 22nm process are not half the size of ones a 45nm one.

 

[AtenRa]

Your math is a little off.

When Intel moved from 32nm to 22nm, they went from 149mm² (2c Sandy) to 94mm² (2c Ivy) and from 216mm² (4c Sandy) to 160mm² (4c Ivy).

That's because you forgetting the bigger HD4000(GT2) that IvyBridge has over the HD3000 in SandyBridge. If Ivy had the same HD3000 that Sandy has it would be almost half the size.

Vishera would be 200-230mm² WITHOUT an iGPU, not a chance it'll get near 180mm² with 384 GCN cores, competing with 160mm² i7's with HD4k graphics (or i5's if you're comparing a cutdown version of the chip).

Again, you forgetting that Vishera has 8MB L3, Trinity doesn't. A 22nm Trinity would be close to 130mm2.

 

[AtenRa]

Not all nodes (22nm or otherwise) are created equal. Intel's processes are top notch and I believe (but might need to go check some sources on this) the transistor density of even Intel's 45nm node was better than GF's 32nm. Regardless, transistors on a 22nm process are not half the size of ones a 45nm one.

We almost have 50% reduction(double the transistor count) with every node every almost two years. That makes Moors law keep going 😉

45nm is 2x times the size of 22nm, 22nm is ~50% (half the size of 32nm).

 

[nismotigerwvu]

We almost have 50% reduction(double the transistor count) with every node every almost two years. That makes Moors law keep going 😉

45nm is 2x times the size of 22nm, 22nm is ~50% (half the size of 32nm).

Very true, but the point I want to make is that a node name is just that, a name and that not all 32nm (or 45, or 22...ect ect) are created equal.

 

[lamedude]

Does anybody know what the stock vcore on the 5700 is?

AMD says 0.825-1.475V.

 

[Centauri]

Not its range - its default.

 

[Yuriman]

Not its range - its default.

I have a 1055T and although stock vcore is 1.35v, turbo (before I disabled it) takes it up to 1.475v.

Suggests voltage will be similar.