Observations with an FX-8350

[(sic)Klown12]

Interesting. So I wonder where in the BIOS I need to enable the throttling. I certainly don't want to be needlessly damaging my chip.

I suppose I better ask ASUS.

Hey, another noob question here, is it common to not be able to run ram at T1 command rate with AMD rigs?

I've got this 4x4GB GSkill ram, the exact same sticks I ran at T1 rate on my Intel Asus mobo, and I can run the ram at the same speed, latency and voltage on the AMD mobo but I cannot enable T1 without the mobo locking up on reboot.

Are AMD platforms not compatible with T1 command rate, or are they more sensitive to the timings and so I will need to bump up the voltage or some such?

Since you're using 4 sticks, have you tried giving a little boost the IMC voltage? In order to use the stated timings of my 4 sticks of DDR2 on my X6 i had to push the voltage up one notch.

 

[thilanliyan]

Do you know of any windows utilities that can track (or change) the NB clockspeed in real-time?

I believe the TurboV tab in Asus AISuite can do it. The setting is "CPU bus. frequency".

 

[Idontcare]

Since you're using 4 sticks, have you tried giving a little boost the IMC voltage? In order to use the stated timings of my 4 sticks of DDR2 on my X6 i had to push the voltage up one notch.

I didn't try that yet, but I will. Wanted to make sure I wasn't chasing a pipe-dream first.

I believe the TurboV tab in Asus AISuite can do it. The setting is "CPU bus. frequency".

Thanks, I will go look for it right now.

 

[Arkaign]

It seems that the 8320/8350 are vastly improved over BD, and a fair option for anyone with an existing mobo that is compatible, but they remain of questionable value compared to Intel options. I believe I'd personally rather have a 2500K or 3570K, to be honest.

What should a fair market price be for these, considering the uneven performance and relatively large power draw / cooling requirement? Maybe 8320 @ $109, 8350 @ $139? I bet AMD would move plenty at those prices.

 

[Torn Mind]

It seems that the 8320/8350 are vastly improved over BD, and a fair option for anyone with an existing mobo that is compatible, but they remain of questionable value compared to Intel options. I believe I'd personally rather have a 2500K or 3570K, to be honest.

What should a fair market price be for these, considering the uneven performance and relatively large power draw / cooling requirement? Maybe 8320 @ $109, 8350 @ $139? I bet AMD would move plenty at those prices.

The FX-8320 would definitely sell like hot cakes at that price since, performance-wise, it is plenty competitive with the i3s. But AMD's would probably be selling those things at a loss and thus it wouldn't be "fair" in that sense.

 

[sequoia464]

Interesting. So I wonder where in the BIOS I need to enable the throttling. I certainly don't want to be needlessly damaging my chip.

I suppose I better ask ASUS.

Hey, another noob question here, is it common to not be able to run ram at T1 command rate with AMD rigs?

I've got this 4x4GB GSkill ram, the exact same sticks I ran at T1 rate on my Intel Asus mobo, and I can run the ram at the same speed, latency and voltage on the AMD mobo but I cannot enable T1 without the mobo locking up on reboot.

Are AMD platforms not compatible with T1 command rate, or are they more sensitive to the timings and so I will need to bump up the voltage or some such?

My experience is this -

(1) Kingston HyperX blue - big mistake - 1.65 volts for 1600 MHz would not run T1

(2) replaced that with some high priced GSkill low voltage 1.35 ECO ram that is supposed to run @ T1 - it wouldn't for me - didn't like anything over 1600 either.

(3) bought some low priced Samsung 1600 memory - Runs great at 1866 1.35 volts, decent latencies and T1

This on an Asus 1st generation Sabertooth

 

[0___________0]

It seems that the 8320/8350 are vastly improved over BD, and a fair option for anyone with an existing mobo that is compatible, but they remain of questionable value compared to Intel options. I believe I'd personally rather have a 2500K or 3570K, to be honest.

What should a fair market price be for these, considering the uneven performance and relatively large power draw / cooling requirement? Maybe 8320 @ $109, 8350 @ $139? I bet AMD would move plenty at those prices.

The 4XXX chips are horribly overpriced, they have i3 pricing, $110 at NE. Should be around $60, going up against the Pentiums and Celerons. 8350 is too much as well, it's in i5 territory. Put it in between the i3 and i5, around $150 and it looks much better.

 

[The Alias]

The 4XXX chips are horribly overpriced, they have i3 pricing, $110 at NE. Should be around $60, going up against the Pentiums and Celerons. 8350 is too much as well, it's in i5 territory. Put it in between the i3 and i5, around $150 and it looks much better.

you're kidding right ?

 

[Zodiark1593]

The 4XXX chips are horribly overpriced, they have i3 pricing, $110 at NE. Should be around $60, going up against the Pentiums and Celerons. 8350 is too much as well, it's in i5 territory. Put it in between the i3 and i5, around $150 and it looks much better.

Considering the already slim profits AMD is making from each chip, I doubt a further price drop will happen soon unless stores want to clear inventory.

There are repercussions to manufacturing a physically larger chip, and manufacturing costs is one of them.

 

[BD231]

you're kidding right ?

He's right, and its not our fault AMD is making chips that arent profitable at their appropriate value.

 

[Arkaign]

Considering the already slim profits AMD is making from each chip, I doubt a further price drop will happen soon unless stores want to clear inventory.

There are repercussions to manufacturing a physically larger chip, and manufacturing costs is one of them.

Yep, and not owning your own fabs (they had to sell them due to the idiotic ATI purchase) makes costs much higher than if you don't have to factor in a 3rd party's profits into your prices.

 

[AtenRa]

@IDC

VID with Turbo On = 1.4125v
VID with Turbo Off = 1.3625v

 

[SlowSpyder]

Interesting. So I wonder where in the BIOS I need to enable the throttling. I certainly don't want to be needlessly damaging my chip.

I suppose I better ask ASUS.

Hey, another noob question here, is it common to not be able to run ram at T1 command rate with AMD rigs?

I've got this 4x4GB GSkill ram, the exact same sticks I ran at T1 rate on my Intel Asus mobo, and I can run the ram at the same speed, latency and voltage on the AMD mobo but I cannot enable T1 without the mobo locking up on reboot.

Are AMD platforms not compatible with T1 command rate, or are they more sensitive to the timings and so I will need to bump up the voltage or some such?

I'm not sure if this applies to BD/PD, but if I recall correctly the Phenom II's had trouble running four sticks at 1T (two sticks of ram, 1T no problem if the memory was rated for such). It could be done, but you usually had to mess with the voltages and not every brand/type of ram would work with 1T.

Now that I think of it, I'm not sure if I ever got my system stable at 1T. My DDR2 is already overclocked, so I'm not sure if I'd have any luck. Maybe I'll mess around with that tonight and see if there is any performance difference.

 

[Idontcare]

@IDC

VID with Turbo On = 1.4125v
VID with Turbo Off = 1.3625v

Thanks for the data :thumbsup:

Do you happen to know anything official regarding TJmax for the FX-8350? Know anyone that would know?

If JFAMD was still at AMD then I would ask him...but he's moved on to greener pastures. In fact every single employee of AMD that I use to know have since moved on from AMD.

Now that I think of it, I'm not sure if I ever got my system stable at 1T. My DDR2 is already overclocked, so I'm not sure if I'd have any luck. Maybe I'll mess around with that tonight and see if there is any performance difference.

On the Intel rig which this ram came from, where it was running T1 at the same voltage and timings, going T1 over T2 improved the bandwidth quite a bit (5% or so IIRC).

That said, I never bothered to quantify the impact on performance with real-world apps.

 

[Idontcare]

Thread scaling results with Maxwell Render's Benchwell:

With 8 threads the FX-8350 at stock scores 314.34 points, which is 6.7x higher than the single-threaded score of 46.90 points.

The thread scaling is nearly identical to that observed with CineBench which has 8 thread performance coming in at 6.6x the single threaded performance.

This table shows us the "CMT tax", the performance loss that comes from the sharing of resources within the same module when the module has 2 threads versus 1 thread.

Comparing C0/C1 (module 1 is fully loaded) to M1/M2 (modules 1 and 2 are loaded with just 1 thread each), we see the C0/C1 case delivers 0.84x the performance of the M1/M2 case.

We also see that commensurate with the reduction in performance when fully loading a module (the C0/C1 case), power consumption is lower but the peak operating temperature increases (power density within the module has increased).

If there are any other benchmarks which folks would like to see comparable data generated then please post with a request and include a link to the benchmark. I'll be happy to oblige.

 

[Puppies04]

If JFAMD was still at AMD then I would ask him...but he's moved on to greener pastures. In fact every single employee of AMD that I use to know have since moved on from AMD.

It is a shame, you could have asked him about IPC as well.

 

[SlowSpyder]

IDC, do you still have your 2700K running? I know your Ivy died (RIP buddy!). I know this wouldn't be scientific with different components, but I would be curious to see how eight threads scales on Intel's Hypterthreading. I know the overall score will likely be higher on a 2700K, but I'm more interested in the scaling, just for comparrison and for the sake of curiosity.

 

[guskline]

I'm not sure if this applies to BD/PD, but if I recall correctly the Phenom II's had trouble running four sticks at 1T (two sticks of ram, 1T no problem if the memory was rated for such). It could be done, but you usually had to mess with the voltages and not every brand/type of ram would work with 1T.

Now that I think of it, I'm not sure if I ever got my system stable at 1T. My DDR2 is already overclocked, so I'm not sure if I'd have any luck. Maybe I'll mess around with that tonight and see if there is any performance difference.

Slowspyder has a good point. AMD's 4 stick memory in my machine required a bump in the Dimm voltage to 1.53 from 1.5 to assure enough juice for stability. I also agree that 1T timing is preferred but 2T might be needed. Finally, despited the claim that PileDriver supports 1866 memory I believe the fine print reads that it supports 2 sticks at that rated speed not 4 sticks.

 

[guskline]

IDC: the "8 core" PileDrive really is closer to 6.7 true cores due to the sharing. If only the software would be tailored for this sharing.

My Intel 2500ks don't have hyperthreading but what is added to the 2600/2700/3770 series to allow hyperthreading?

 

[grimpr]

IDC, Give this nice "simulator" a try.

http://www.dimkovic.com/node/1

 

[grimpr]

My Intel 2500ks don't have hyperthreading but what is added to the 2600/2700/3770 series to allow hyperthreading?

Nothing, its just blown fuses on the Core i5.

 

[ShintaiDK]

My Intel 2500ks don't have hyperthreading but what is added to the 2600/2700/3770 series to allow hyperthreading?

Nothing. The feature just aint fused off.

 

[beginner99]

The FX-8320 would definitely sell like hot cakes at that price since, performance-wise, it is plenty competitive with the i3s. But AMD's would probably be selling those things at a loss and thus it wouldn't be "fair" in that sense.

Maybe but they might would lose less than pay GF not to make chips while gaining market share.

 

[SammichPG]

I'd be curious to see how that cpu behaves with 4 modules disabled, I guess that it would clock higher, consume less power and be generally slower while running highly threaded applications.

 

[NTMBK]

I'd be curious to see how that cpu behaves with 4 modules disabled, I guess that it would clock higher, consume less power and be generally slower while running highly threaded applications.

It would clock infinitely higher, have 0W power draw and be infinitely slower.

(It only has 4 modules. )