AMD Desktop Trinity Update: Now With Core i3 And A8-3870K

Page 5 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
[citation][nom]blazorthon[/nom]If something gets damaged, then maybe the CPU northbridge/memory controller can get damaged, but the memory should not be damaged. Even then, you can look up if the northbridge needs a boost or not to correspond with a memory voltage boost. If it does become a problem, then make sure that your memory modules have heat spreaders. If they do, then I don't think that they would have temperature problems unless they have poor quality heat spreaders.[/citation]
Having your integrated memory controller get damaged doesn't sound reassuring. :-( I hope that's a worst case/very unlikely kind of thing.
What are the symptoms of a lack of corresponding memory controller performance? That would require a boost of the BCLK right?
They sure do have simple heat spreaders, though since I've populate all the RAM slots, they're pretty much flush against each other. I have this, just in case you may have any opinion about their heat spreaders: http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=1412141&CatId=4534
Also, do you think it's worth loosening timings for the sake of increasing clockrate while not increasing? Thanks again! Hehehe... Sorry for so many questions. :-D
 
Wow this is really sad. Too many amd fan boys here. It's already PROVEN an LOW-END i3 STOMPS the best amd has to offer at a LOWER price. But yet ppl trying to think the on-board video cards will magically beat dedicated video cards. AH....NEWS FLASH....Even the 6470 amd card STOMPS the ever living crap of the on-board. amd cannot even compete against THEMSELVES. Get a low end i3 and 6470 or better=complete destruction of anything amd has.
I still see ZERO reason for ANYONE to use an amd. It leads ZERO in any benchmark.
 
[citation][nom]computertech82[/nom]Wow this is really sad. Too many amd fan boys here. It's already PROVEN an LOW-END i3 STOMPS the best amd has to offer at a LOWER price. But yet ppl trying to think the on-board video cards will magically beat dedicated video cards. AH....NEWS FLASH....Even the 6470 amd card STOMPS the ever living crap of the on-board. amd cannot even compete against THEMSELVES. Get a low end i3 and 6470 or better=complete destruction of anything amd has.I still see ZERO reason for ANYONE to use an amd. It leads ZERO in any benchmark.[/citation]

AMD wins plenty of benchmarks in many situations, not that this is a very important feat anyway. There is no such thing as the Radeon 6470 and since even the Llano A6s are faster than the only 6400 card, the 6450, by a large margin, I can't imaging how you could be even remotely correct here, especially with how the new A10s can compete with and with 1866MHz memory or better even beat the Radeon 6670 while having CPU performance that easily competes with the i3s. You should learn how to read benchmarks.
 


I've done plenty of messing around with memory overclocking and I've never had a northbridge failure caused by it. It is most definitely a worst-case scenario that is quite unlikely, especially in the short term. If AMD's memory controller is not good enough, then you boost the northbridge frequency. Boosting the BLCK would also boost this frequency because it is based off of it, but that would obviously change much more than the northbridge frequency and you should be sure that you want to do this before doing it.

Timings versus clock rate, well, it depends on many things. Different programs can favor either one or not care about either of them or even favor both. So long as you don't lose timings more than the clock frequency jump can relieve in actual latency, the latency won't increase, but increasing frequency while retaining the same latency can need voltage hikes (although generally only small ones).
 
Thanks again blaz! I'll try my hand at it. Even though I don't have an APU system (nor AMD, sorry, though I know you aren't a fanboy of either Intel or AMD), maybe the questions I asked on this forum still could help people who do. Hehe...

It's interesting how latency and bandwidth can be both or either favored by an application, though with that thought in mind, I am getting some thoughts about it.

[citation][nom]blazorthon[/nom]AMD wins plenty of benchmarks in many situations, not that this is a very important feat anyway. There is no such thing as the Radeon 6470 and since even the Llano A6s are faster than the only 6400 card, the 6450, by a large margin, I can't imaging how you could be even remotely correct here, especially with how the new A10s can compete with and with 1866MHz memory or better even beat the Radeon 6670 while having CPU performance that easily competes with the i3s. You should learn how to read benchmarks.[/citation]
Um... I was hoping to find benchmarks proving that here on TH, but the old Llano release review had none comparing a discrete as I saw. I checked the Gaming Graphics Hierarchy Chart though and saw that the HD6550D, found on the A8 Llano, was in the same tier as the HD6450. Now even though this seems contrary to what you said blaz, there is no mention of what clockrate of memory they used, and if they used the clockrate used in the desktop Llano release review, that's only 1333MT/s as I saw it.

I was scouring through the older desktop Trinity review and it seemed like they didn't have a discrete comparison as well, but the HD6670 was listed in the test setups.

Maybe you can share some benchmarks you may have that prove your point blaz? 🙂
Also, I didn't like the way computertech82 was "talking."
 
[citation][nom]army_ant7[/nom]Thanks again blaz! I'll try my hand at it. Even though I don't have an APU system (nor AMD, sorry, though I know you aren't a fanboy of either Intel or AMD), maybe the questions I asked on this forum still could help people who do. Hehe...It's interesting how latency and bandwidth can be both or either favored by an application, though with that thought in mind, I am getting some thoughts about it.Um... I was hoping to find benchmarks proving that here on TH, but the old Llano release review had none comparing a discrete as I saw. I checked the Gaming Graphics Hierarchy Chart though and saw that the HD6550D, found on the A8 Llano, was in the same tier as the HD6450. Now even though this seems contrary to what you said blaz, there is no mention of what clockrate of memory they used, and if they used the clockrate used in the desktop Llano release review, that's only 1333MT/s as I saw it.I was scouring through the older desktop Trinity review and it seemed like they didn't have a discrete comparison as well, but the HD6670 was listed in the test setups.Maybe you can share some benchmarks you may have that prove your point blaz? 🙂Also, I didn't like the way computertech82 was "talking."[/citation]

With Intel, the northbridge is less important than on an AMD system or an old Intel system because it is already using a full speed cache and the controller is extremely efficient.

With dual-channel DDR3-1600, the Radeon 6670 is ~50% faster than a Llano A8. With the same memory, the A10 is creeping up on the Radeon 6670 and with 1866MHz or 2133MHz memory, the 6670 can be caught and probably overtaken. Even a Llano A6, with dual-channel DDR3-1600, is significantly faster than the Radeon 6450. If Tom's put them in the same tier, then it's probably just with very low speed memory and with old drivers.
 
I just bought a HP Pavilion p7-1240 Desktop from Amazon. It will be here tomorrow. I just needed a replacement for my good old trusty 4800+ system. For $603.98 including overnight shipping it was just what I needed.
 
Personally, I am finishing up my new rig and I am using a Gigabyte GA-990FXA-UD5 ATX AM3+. I'm running the Phenom II x6 1100T Black 3.3Ghz with a GTX 670. I like to look at things practically not just side by brands. Now With what that processor and GPU I should have no problem running any game or even using Photoshop. I prefer to wait a little while and see what happens with their new line up before jumping right in.
 
[citation][nom]tourist[/nom]What i have asked nicely for is a straight up comparison of the top two K chips fom llano and trinity, side by side, same mhz and run benchmarks using 1333,1600,1866, 2133 and 2400. I think it would be a interesting article and much appreciated by there loyal fan base. I know people think llano/tinity is a low tier cpu not worthy of time but there perception is based off of systems using 1333,1600 memory. And really how many reviews does one need of high end graphics cards. Army to get a idea of how memory scales on llano http://www.hardwarecanucks.com/for [...] ew-16.html[/citation]
Thanks for the link! 🙂 I believe they've done comparisons using different memory speeds I believe, but I remember that they weren't that exhaustive as to answer those things you mentioned and what we (all) want to know.

 
Perhaps I've missed something, but how is it that the AMD's all get to use dedicated graphics cards in the gaming performance tests, whereas the Intel chips must use only their onboard graphics? Doesn't seem like a very useful comparison, and then to say that the Intel chips are rubbish for gaming strikes me as very odd... Explanation, anyone?
 
[citation][nom]coolAlias[/nom]Perhaps I've missed something, but how is it that the AMD's all get to use dedicated graphics cards in the gaming performance tests, whereas the Intel chips must use only their onboard graphics? Doesn't seem like a very useful comparison, and then to say that the Intel chips are rubbish for gaming strikes me as very odd... Explanation, anyone?[/citation]
Those are the "discrete-class" integrated graphics which make them "APU's" I believe. Take not that if they have a D after their names, like HD6550D, they're integrated APU graphics.
 
Would be nice it there's HTML5 based benchmarks in the future, since most people spend more time on the web nowadays.
Thou i have no suggestion which benchmark/test.
 
@coolAlias
They are technically both integrated graphics as they are both on 1 chip.
AMD calls them APU cause they have some technology which shares some resources between the graphics component and the CPU component.
 
[citation][nom]computertech82[/nom]Wow this is really sad. Too many amd fan boys here. It's already PROVEN an LOW-END i3 STOMPS the best amd has to offer at a LOWER price. But yet ppl trying to think the on-board video cards will magically beat dedicated video cards. AH....NEWS FLASH....Even the 6470 amd card STOMPS the ever living crap of the on-board. amd cannot even compete against THEMSELVES. Get a low end i3 and 6470 or better=complete destruction of anything amd has.I still see ZERO reason for ANYONE to use an amd. It leads ZERO in any benchmark.[/citation]

It really depends on what you're using your computer for.
If you crunch numbers everyday with excel or do lots of video with Quicksync then get an Intel, but with a discrete graphics please.
These benchmarks are generally useless since "most" people don't zip/unzip their files everyday.

AMD would be better for me cause:
1. Graphics performance (not a heavy gamer but do play some games)
2. Hybrid crossfire (would be great to have for future upgrades) [note that only some cards can do this]
3. Energy savings assuming i don't use an additional graphics card
4. Hardware acceleration on HTML5 (Unknown, would like to see it being tested)
5. OpenCL more applications may start using this which means AMD would run better if you intend to keep that CPU 5 years later (this is just my assumption, since OpenCL adoption is somewhat slow)
 
[citation][nom]tourist[/nom]Your being nice army +1. what i see is they copied and pasted info from the first article over to this one, nothing wrong with that but a notation would have been nice. 1) in test setup they mention a hd 6670. There is no reference to it in this article, that is what i think coolAlias is referring to ?2) the same for the ddr3 2800 memory, nothing !3) Stating that llano and trinity's bclk is 200 mhz ![/citation]
Thanks! That's what these forums are for after all, sharing info. One reason for their existence at least. 🙂
I noticed that HD6670 as well. As for the DDR3 2800MHz RAM, I forgot if it was this article, but I remember it being mentioned that it was just downclocked. I'm not sure though, unless I re-read.
 
@army_ant7 / kaggy / tourist
Yeah, I noticed they were truly integrated graphics cards after I posted... props to AMD for such nice onboard graphics! I suppose if your dedicated gpu fries and you just can't wait to play your games, then you could still play on ultra-low settings.
@army_ant7
Thanks for being nice! Too bad I just wasn't paying attention... doh. That is curious about the HD6670, however...
 
[citation][nom]coolalias[/nom]@army_ant7 / kaggy / touristYeah, I noticed they were truly integrated graphics cards after I posted... props to AMD for such nice onboard graphics! I suppose if your dedicated gpu fries and you just can't wait to play your games, then you could still play on ultra-low settings.@army_ant7Thanks for being nice! Too bad I just wasn't paying attention... doh. That is curious about the HD6670, however...[/citation]

Nitpicking here, but for Llano, Sandy Bridge, Trinity, and other successors of both, it's on-die for being on the CPU die, not on-board for being on the motherboard. On-board would be the IGPs for AMD'a AM3+/AM3 platforms. I don't think that any other modern platforms except maybe for very low end systems still uses on-board integrated graphics.
 
@blazorthon
Ah, yes. I got lazy with my nomenclature there. Henceforth I'll just refer to them all as either "integrated" or "dedicated."
 
T_T This is about AMD's APU right? why does the "Conclusion" needs to be about defending the performance of the Core i3s T_T,. I dont want to say this is biased because I like tomshardware alot,....
 
[citation][nom]Testgamma1[/nom]T_T This is about AMD's APU right? why does the "Conclusion" needs to be about defending the performance of the Core i3s T_T,. I dont want to say this is biased because I like tomshardware alot,....[/citation]
As I remember seeing it, it was just giving credit to whom credit is due. 🙂
 
[citation][nom]Testgamma1[/nom]T_T This is about AMD's APU right? why does the "Conclusion" needs to be about defending the performance of the Core i3s T_T,. I dont want to say this is biased because I like tomshardware alot,....[/citation]

Looked more like constructive criticism of the i3s to me.
 
[citation][nom]csbeer[/nom]I wish these sites would do a thorough review of the dual graphics/asymmetrical crossfire set up of the APU chips. From what I've read there are certain situations where they shine but are highly dependent on the drivers.[/citation]
core i3 chips dont have turbo boost
 
"Ivy Bridge-based Core i3s will help Intel’s case with regard to x86 performance. However, it’s not yet clear if any of them will include HD Graphics 4000. Should the company choose to keep its higher-end implementation a differentiating feature for Core i7 CPUs, you’ll see the more entry-level parts limited to HD Graphics 2500. That’d just be bad news for Intel."

The i3-3225 will have it.

Still, the fact that the HD 4000 can barely keep up with last year's 6550D -- even with the vast raw processing power of a $330 3770K behind it -- doesn't bode well for what will likely be a $140 chip.
 
[citation][nom]KingmanIII[/nom]"Ivy Bridge-based Core i3s will help Intel’s case with regard to x86 performance. However, it’s not yet clear if any of them will include HD Graphics 4000. Should the company choose to keep its higher-end implementation a differentiating feature for Core i7 CPUs, you’ll see the more entry-level parts limited to HD Graphics 2500. That’d just be bad news for Intel."The i3-3225 will have it.Still, the fact that the HD 4000 can barely keep up with last year's 6550D -- even with the vast raw processing power of a $330 3770K behind it -- doesn't bode well for what will likely be a $140 chip.[/citation]

HD 4000 on the i7s seems to be a better contender for the 6530D than the 6550D. Considering that the i3's HD 4000 is weaker than the i7's HD 4000, it would likely be weaker than the 6530D.
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom