• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Piledriver stacking up

Caporegime
Joined
17 Mar 2012
Posts
50,858
Location
ARC-L1, Stanton System
Lets see if we can't inject some positive AMD news in here.

The first (indications) of Bulldozers replacement (Piledriver) are out. so how does it stack up?

First Per-Core Performance Bulldozer

iTunes.

All @ 3.3Ghz

1# i7 2600K = 1.09

2# x6 1100T = 1.35

3# FX-8150 = 2.03

Piledriver core vs Bulldozer core @ 3.8Ghz

3.8Ghz is stock for this Piledriver core (without Turbo core)
As you can see the performance is up 15% clock for clock and core for core.
The FX-8150 runs 3.6Ghz stock. so take off about 8% to BD's 3.3Ghz iTunes time and you get about 1:54. (114sec + 8% = 123sec)
Trinity stock is 1:31, thats 26% faster overall than Bulldozer (91sec + 26% = 114sec)

The new chart now looks like this

(all CPU's at stock)

1# i7 2600K = 1:08

2# Piledriver = 1:31

3# x6 1100T = 1:35

4# FX-8150 = 1:54

As you can see the FX-8150 is about 40% behind the 2600K and Piledriver now replaces the x6 1100T in second place (all be it by about 2%)
Its a vast improvement on Bulldozer, keep in mind that Trinity has no L3 and that the FX-8350 is likly to be clocked at 4Ghz stock, the indications are a real gain on Thuban even if its not a big one.

Multithreaded Productivity

FX-8150 = 1:50

Trinity

Piledriver Core = 3:0

Its difficult to predict Core scaling, But Trinity has 50% less cores than the FX-8150 and again no L3 cache, if one was to double it up the FX-8350 would score 1:30, be that as it may it actually beats the 2600K. i would say its probably a dead match., perhaps slightly better, certainly it would continue to out class the Thuban and 2500K.

That's just a couple, but there is plenty more there, if you take your time and think you can see the Piledriver core is a vast improvement over Bulldozer, its even an Ivy Bridge vs Sandy Bridge level of improvement over Thuban (tho not in all areas), a step in the right direction...
 
Last edited:
I can't wait to see it out and then we can see what it'll really do. I know we're going off that 1 Trinity test from Tom's hardware but if that replicates into the full Piledriver CPU then it will be interesting to see how it fares against Sandy and Ivy. It'll have the multi-threaded performance but if it can pull back some ground in single core software it'll be great for the competitiveness of the CPU market in general.
 
I can't wait to see it out and then we can see what it'll really do. I know we're going off that 1 Trinity test from Tom's hardware but if that replicates into the full Piledriver CPU then it will be interesting to see how it fares against Sandy and Ivy. It'll have the multi-threaded performance but if it can pull back some ground in single core software it'll be great for the competitiveness of the CPU market in general.

Unfortunately they are the only ones which have a Desktop Trinity CPU, so far.

But at least we can compare using the same reviewers making the same benches.

There is also a larger selection of Mobile Trinity reviews, they also suggest the same thing this one review does, the Piledriver Core absolutely hammers Bulldozer. :)
 
Lets hope AMD are actually going in the right direction with their next CPU (it looks like it), they definitely need to pull something decent out of the hat otherwise they will only leave more underwhelmed enthusiasts! Doesn't have to beat Intel, but at least beating the old Phenom II is something.
 
When is the piledriver meant to be released guys?

In Trinity, next few months, in Vishera, meant to be again later this year, wouldn't be surprised to see it slip to Q1 2013 though, although that would be disappointing, really hoping AMD can bring out a heavy hitter with Vishera.
I'm hoping for at least a high clocking Phenom II X8 performance wise.
Depending on Haswell, Vishera may be worth a purchase for myself.
 
The problem AMD have is that all Intel need to do is lower 3930K to ~£250 and AMD will be playing catch up again, that's a 50% performance improvement for Intel right there.
 
The problem AMD have is that all Intel need to do is lower 3930K to ~£250 and AMD will be playing catch up again, that's a 50% performance improvement for Intel right there.

Yeah, though i would think they would only do that if they loose significant sales to AMD, which probably isn't going to happen, AMD might improve their sales with this. but i doubt it will scratch let alone dent Intel's vast sales.
 
The problem AMD have is that all Intel need to do is lower 3930K to ~£250 and AMD will be playing catch up again, that's a 50% performance improvement for Intel right there.

If AMD do bring out a CPU to compete, and Intel do drop the hex, while that's not what AMD want, it's best for consumer.

Have to give praise to AMD for making hex/octo mainstream.
 
is there any likelihood of a 10 core vishera chip? I thought I saw an old roadmap with that on it but that was last year sometime.
 
oh well 10 core would have been awesome, but I shouldn't rule out the 8 core, if it clocks well and betters phenom ii's single threaded performance it will likely be the last best cpu for the am3+ platform so could displace my x4, it would probably be fun to overclock. Would guess whatever comes after will need a new chipset to get the best out of it and possibly ddr4.
 
I think they need to concentrate on getting the performance up on one core before they start adding yet more cores lol....
 
one thing they could do right would be not to do away with the solder and use cheap tim on the heat spreader lol.
 
one thing they could do right would be not to do away with the solder and use cheap tim on the heat spreader lol.

Again yeah.... its not very often you see Intel making mistakes these days, but that is one blunder and one would think AMD are very very aware of it.

Unless they are catastrophically dumb, they will continue to solder their DIE cap to the heat spreader.
 
Last edited:
Back
Top Bottom