AMD Athlon 5350 And AM1 Platform Review: Kabini In A Socket

[Reynod]

Would this be sufficient for a HTPC Don?

 

[mikat]

"It's going after a new pseudo-desktop arena referred to as "PC-like devices". This is the battleground where Android-equipped set-top boxes and media players are taking pieces of the traditional desktop machine's pie."

Agree

"A PC is so much more than just a CPU and motherboard, though. The rest of the components, such as memory, hard disks, and an operating system, already make up much of a budget machine's price tag. So I'm skeptical of this platform's ability to reclaim ground for the PC."

Hmm, now you seem to be contradicting yourself.

Well... streaming boxes for example don't have hard disks. I don't see the memory differentiating them because they both need it. If the box has less, so will the Kabini equivalent. Actually the new Amazon FireTV has a good amount of memory. Operating system? All streaming boxes have something so why does this rule out Kabini? So if a company wants a way more capable set top or streaming box, Kabini could be a way to go. It will cost a little more for more performance and flexibility (more PC-like).

 

[HomeSkillenSlice]

I really need to see this CPU paired with something like an R7 250 or 250x . I bet this thing would kill for really low end budget gaming builds. I read the Hexus review linked below and I saw that at ultra settings the CPU was able to keep up with the 750Ti and play both Bioshock Infinite and GRID 2 at ultra settings with around 30-40 FPS. I really want to see some hard data on its gaming capabilities with Discrete graphics. JAGUAR ARCHITECTURE OVERVIEW MAYAAAAN !!!!

Sure you can easily conclude with a cinebench benchmark that this thing has half the single threaded performance of something like an A10 6790k (check the PCPER review) but the thing those stupid synthetics dont have is the ability to let people know how these things would do with some of the more demanding CPU settings turned down to bring a greater experience. I need some benchmarks! Imagine it : Cheap $35 Sata 3 MB , Quad Core Athlon for future multi-threaded titles , 4GB's of RAM , and a subpar but still more then capable R7 250/250x , all for like $350

http://hexus.net/tech/reviews/cpu/68473-amd-athlon-5350-28nm-kabini/?page=8

 

[dusty13]

this could be useful for a htpc ... but with boards that max out at 4 sata-ports and often only one pci-e port its just no use. 1x blu-ray, 1x system 3x storage (raid5) ... you run out of ports ... same with pcie if you use a dvb-decodercard and a ci-module you need at least two ports, so even in the best case, where you have those ports you have no expansion to solve the sata-port problem through extension-cards ...

would have loved to build a low-power system like this but as is i will fall back to core i5 with 35 watt and 1150 socket. for those you at least get nice boards.

those systems just dont cut it ...

 

[Wisecracker]

mATX AM1 motherboards have 3 X PCIe slots will bunches of USB2/3 connectivity. I'm sorry that is not enough for an Intel elitist who finds the need to leap over Bay Trail for an over-priced i5 system.

But, I keep my media library on 2 8-bay eSATA external enclosures accessible to my home network, or anywhere else for that matter. I guess I just don't 'cut' it with my near-30TB of accessible storage.

 

[ewok93]

These look interesting, might use one of these in a small office/schoolwork build.

 

[cleeve]

Really? I thought they are low power consumption systems, I mean who would buy a 850 watt PSU to power any of these systems seriously. I just felt it was a strange decision to use a 850 w PSU in this review.

This is not a build, this is a test. We aren't suggesting you pair this PSU with this board, we're simply testing it.

The power capacity of the PSU doesn't affect the draw. It doesn't matter what we use to test, as long as its an efficient PSU.

 

[cleeve]

Would this be sufficient for a HTPC Don?

For sure. that's probably one of its best-case scenarios.

 

[cleeve]

Hmm, now you seem to be contradicting yourself.

Well... streaming boxes for example don't have hard disks. I don't see the memory differentiating them because they both need it. If the box has less, so will the Kabini equivalent. Actually the new Amazon FireTV has a good amount of memory. Operating system? All streaming boxes have something so why does this rule out Kabini? So if a company wants a way more capable set top or streaming box, Kabini could be a way to go. It will cost a little more for more performance and flexibility (more PC-like).

I don't agree.

Hard disks... if you're honing in on streaming media box exclusively, we're talking about a very specific use case here. You can run the OS off of a USB, sure, but you're limiting the system in pretty much every other scenario. And don't forget, you could run a streaming OS off of an FM2+ platform, too.

But that's getting off topic a bit. The real point is that the motherboard and CPU are not the entire expense in a system. Price out a cheap machine and you will find that the other components are a significant percentage of cost.

So we're not talking about $60 motherboard & CPU vs $110 motherboard and CPU, which sounds like a huge savings... we're talking $450 system vs $490 system. Know what I mean?

 

[firefoxx04]

What a dumb power supply choice. Not only is it unrealistic for these builds but it makes both setups look like they pull too much power.

To put into perspective, my FM2+ board w/ FM2 A4 runs in the low/mid 20s and rarely goes over 30W while streaming a 1080p movie over wifi from my server.

Idling at 30W really made me sad but then I realized my HTPC is running a very low wattage seasonic SFX psu that is made for these low power CPU/APU builds. No one uses a 850W PSU on these. I know you guys like to overbuild the test setups to avoid bottlnecks but neither of these machines would have been bottlnecked by a nice gold rated sub 400W psu.

 

[cleeve]

What a dumb power supply choice. Not only is it unrealistic for these builds but it makes both setups look like they pull too much power.

That doesn't make sense.
The power supply's maximum capacity doesn't have anything to do with draw.

You're thinking efficiency, and the 800W PSU is 80-plus certified, like pretty most PSUs out there.

On top of that, I don't think there's any implication that this is a suggested build. it's a test bench.

All this in mind, calling it a 'dumb choice' is kind of sensationalist, even for an internet forum.

 

[cablechewer]

I wonder how this would perform for home labs. I am always watching for cheaper hardware platforms that would let me run a number of VMs. In most cases my VMs are largely idle. I use them for simple repros and the interface of the servers I work with. I see the specs of at least 1 board say 32GB of RAM is supported, but I don't see any 16GB DIMs for sale

 

[ewok93]

The board has 4 slots, cablechewer. Four 8gb sticks.

 

[__Miguel_]

The power supply's maximum capacity doesn't have anything to do with draw.

You're thinking efficiency, and the 800W PSU is 80-plus certified, like pretty most PSUs out there.

Actually, he does have a slight point.

Granted, he was a bit out there when it comes to calling the choice stupid, but he did point out a valid flaw on your test setup.

As you well know, 80+ doesn't mean the PSU is 80%+ efficient at every point of its output line. It means that efficiency appears at (usually) 20% and up of its rated max output. On an 800W PSU, that's 160W+ of constant power draw on the PSU side, just to meet the 80% efficiency rating.

Below that, efficiency drops, often abruptly, meaning that, for instance, a 400W PSU with the same 80+ rating would most likely be pulling less power from the outlet, because to get to 80%+ efficiency you would need only to be pulling 80W from the PSU side.

With systems measuring in at sub-40W peak power load, testing them on an 800W PSU means that you're using, at most, 5% of its rated output. At that point, you can bet you'll be pulling more juice from the outlet than you would be if you were using something like a 200W 80+ certified PSU. Or even a 75W PicoPSU + power brick. If even an E7200+G31 motherboard full system (check Tom's past articles) was seen idling at around 30W all that time ago, you can bet that this day and age this kind of setup WILL idle closer to 20W, if not even less.

My point, and I assume firefoxx04's, is this: for low power systems, power draw figures need to be measured with much lower wattage PSUs; otherwise, you'll end up with figures that are at least 5-10W too high.

Great review, in any case. By the way, considering HTPCs, are you planning on doing an article of its HTPC capabilities?

Cheers.

Miguel

 

[cleeve]

My point, and I assume firefoxx04's, is this: for low power systems, power draw figures need to be measured with much lower wattage PSUs; otherwise, you'll end up with figures that are at least 5-10W too high.

In my experience that's higher than the actual difference we'd see, but I'll see if I have a low-power high-efficiency model around and test it for comparison.

By the way, considering HTPCs, are you planning on doing an article of its HTPC capabilities?

I'll probably include it in our next video playback quality analysis, it's been a while since we've looked into that and it's probably time.

 

[Wisecracker]

In a 'real world' power delivery scenario it's just not a big deal ... wasting 25% of power drawn versus 20% of power drawn when working in a 10-30w envelop ain't much. I'm math-challenged this AM but it's less than a watt.

That said, using an inexpensive (and power-efficient) power brick on a $200 low-budget HTPC STB coupled with the graphics power of 2 GCN CUs is a winner for everyone.

 

[Jaroslav Jandek]

I actually expect only 1-4W difference (unless the mini PSU is total crap).

I would really like to have a NUC with the Atom Z3770 - too bad they only use them in tablets (AFAIK). I'd like to see how J1800 compares to Z3770, because from first hand experience, Z3770 is more than powerful enough for my needs (I may end up buying a tablet instead of getting an HTPC - even though the screen would be unused).

 

[Wisecracker]

My understanding ...

At 10w system power the draw from the wall is 12w at 80% efficiency.

At 75% efficiency power at-the-wall would be 12.5w. Efficiency at the low end would have to fall below 70% (which can be the case in some poorly build PSUs) to grow beyond a single watt.

 

[Jaroslav Jandek]

At 75% efficiency power at-the-wall would be 12.5w. Efficiency at the low end would have to fall below 70% (which can be the case in some poorly build PSUs) to grow beyond a single watt.

Actually, even with 80+ PSUs, it can drop to 60% if the power draw is extremely low - like in this case.

Also, you can clearly see the lower power PSUs doing way better than higher power PSUs with lower power consumption:

(www.silentpcreview.com)

 

[Isaiah4110]

My understanding ...

At 10w system power the draw from the wall is 12w at 80% efficiency.

At 75% efficiency power at-the-wall would be 12.5w. Efficiency at the low end would have to fall below 70% (which can be the case in some poorly build PSUs) to grow beyond a single watt.

That's not how I've always heard PSU efficiency described. The way I've heard it described is a PSU is 80% efficient (and current power draw) if the system is currently utilizing 800W of power but the PSU is pulling 1000W at the wall. (System usage / Measurement @ Wall = efficiency)

Therefore I would expect a system using 10W of power at 80% efficiency to be pulling 12.5W at the wall. A system using 10W of power on a PSU with 75% efficiency would then be pulling 13.333W at the wall and that same system with a 50% efficient PSU would actually be pulling 20W at the wall.

Ecova Plug Load Solutions (the company that defines the requirements for 80Plus certification and runs testing) actually has a new level call 80Plus Titanium that, in addition to slightly raising the bar from the Platinum requirements at 20%, 50% and 100% load, requires the PSU to also be 90% efficient at 10% load.

According to testing done by JonnyGuru.com, the specific model of power supply used in this test bench is just over 75% efficient at 10% load. This means when the PC is actually using 85W of power, we can expect to see the PSU pull a little over 110W at the wall. As for what happens when you go lower than 85W with this PSU I can't really calculate, but I would venture to guess the efficiency drops below 70%.


EDIT: Checked the calculations here on wikipedia along with those in the above linked JonnyGuru review. Efficiency is calculated as actual system usage divided by measurement from the wall as I noted above.

 

[Wisecracker]

I don't understand the math. :lol:

The point being ... using our examples the difference in power drawn is still .833w (instead of my fail at .5w). That's like $0.125 per 1,200 hours of use.

I have no issue with beating up on Don when he deserves it but in this case it's really too much. I also have no issue with seeking the utmost efficiency but what sense does it make to purchase a platinum power supply for a $200 HTPC to save $.50 in electricity over a years time ?

 

[RedJaron]

I have no issue with beating up on Don when he deserves it but in this case it's really too much. I also have no issue with seeking the utmost efficiency but what sense does it make to purchase a platinum power supply for a $200 HTPC to save $.50 in electricity over a years time ?

I think the complaint was whether putting such a low load on a large PSU would impact the power usage readings, particularly if it would have exaggerated the margin of error, and reported a significantly higher or lower than actual power consumption for the CPUs.

 

[Isaiah4110]

I have no issue with beating up on Don when he deserves it but in this case it's really too much. I also have no issue with seeking the utmost efficiency but what sense does it make to purchase a platinum power supply for a $200 HTPC to save $.50 in electricity over a years time ?

I think the complaint was whether putting such a low load on a large PSU would impact the power usage readings, particularly if it would have exaggerated the margin of error, and reported a significantly higher or lower than actual power consumption for the CPUs.

That was absolutely my understand as well. I thought it was a great review overall, I was simply pointing out that those who mentioned there may be a flaw in the power draw readings at idle due to general PSU inefficiency at low overall voltage may have a point.

 

[__Miguel_]

I think the complaint was whether putting such a low load on a large PSU would impact the power usage readings, particularly if it would have exaggerated the margin of error, and reported a significantly higher or lower than actual power consumption for the CPUs.

That was absolutely my understand as well. I thought it was a great review overall, I was simply pointing out that those who mentioned there may be a flaw in the power draw readings at idle due to general PSU inefficiency at low overall voltage may have a point.

Personally, that was exactly my point.

And I wasn't trying to bash the reviewer in any way, since the point of having a fixed test setup is to achieve consistency in the results. And, if it really turns out that even at under 5% load, the PSU's efficiency doesn't impact power figures negatively, my gripe will actually be with both Intel and AMD, for not being able to drop idle power draw for the whole system below ~30W since 2008 (!) (check Tom's "G31 And E7200: The Real Low-Power Story" article).

 

[de5_Roy]

my gripe will actually be with both Intel and AMD, for not being able to drop idle power draw for the whole system below ~30W since 2008 (!) (check Tom's "G31 And E7200: The Real Low-Power Story" article).

idle power is dependent more on the platform, bios and o.s. even if the cpus/socs have power saving tech built in. most motherboards disable some of the power saving options for wider range of compatibility. reviewers often test platforms at default settings (leaving optimizations to vendors) which doesn't often show how low/high a customized setups can go.