Best value "Entry Level" gaming PC.

[cbn]

Here the 5350 is 49 and the Celerons are 38-40, nices dedicated MB s are 30 and 65 repectively, do the maths.

And what are the prices for uATX AM1 and uATX LGA 1150?

At Newegg, the cheapest LGA 1150 is the Biostar H81MHV3 at $39.99 with $5.39 shipping ---> http://www.newegg.com/Product/Produc...82E16813138407

And the cheapest AM1 BIOSTAR AM1ML is $30.99 with $2.99 shipping ---> http://www.newegg.com/Product/Produc...82E16813138401

 

[VirtualLarry]

Fair, there s inherently some redundancy between such threads, on another note for those who have only 2 SATA 3 ports like in your board it s not a problem, use one for the main drive while you can plug several items on the other SATA port, AMDs SATA controler can manage several devices in a single port , bandwith per port is at roughly 4gbit/s.

You are talking about SATA port multiplier support, no? Is that supported in Windows? What port-multiplier controller board do you suggest, and how much does it cost, compared to just a PCI-E SATA controller card?

 

[Abwx]

And what are the prices for uATX AM1 and uATX LGA 1150?

At Newegg, the cheapest LGA 1150 is the Biostar H81MHV3 at $39.99 with $5.39 shipping ---> http://www.newegg.com/Product/Produc...82E16813138407

And the cheapest AM1 BIOSTAR AM1ML is $30.99 with $2.99 shipping ---> http://www.newegg.com/Product/Produc...82E16813138401

There s no benefit price wise, in Europe, to buy a uATX over a mini ITX when it comes to AM1, your MSI mITX can be found at 27-28€ while the Asrock AM1B mITX is at 30€, uATX are at the same prices.

H81 mITX are 65€ at the lowest, uATX are as low as 39-40€ but these are inferior to the 30€ Asrock AM1B mITX connectivity wise, H81 has to 2 SATA 3gb/s ,two SATA 6 gb/s and two USB3 ports, to compare with said Asrock that has 4 SATA 3 + 4 USB3.

From all thoses numbers and from technical considerations we can deduct that there s no point using uATX MBs and hence casings for plateform with as low power comsumption as an AM1, indeed there s more mITX offerings than uATX , it say it all.

You are talking about SATA port multiplier support, no? Is that supported in Windows? What port-multiplier controller board do you suggest, and how much does it cost, compared to just a PCI-E SATA controller card?

A sata port support several devices, depending of the controler, this is implemented in Kabinis since it s also used in low power servers, cost should be a cable.

 

[cbn]

Regarding testing for the future, here is part of my plan:

1. Revisit the Dirt 3 built in benchmark (results so far posted here--> http://forums.anandtech.com/showpost.php?p=36730608&postcount=133) and answer the question I posted at the end of this post--> http://forums.anandtech.com/showpost.php?p=36731634&postcount=138

2. Look for more games with built-in benchmarks.

3. Test with more processors. So far I have Pentium G3258, Athlon 5350, A6-5400K (still need a FM2+ mobo). I would also like to add Celeron G1820.

4. Multiplayer testing with Team Fortress 2. I figure since we are discussing the most entry level iGPU processors capable of gaming, having a popular game that is free on Steam should be included. (Data will be difficult to reproduce here, but I should still be able to give an opinion)

 

[cbn]

I may also add some single channel testing to the processor list (using 1 x 4GB DDR3 1600 DIM rather than 2 x 2GB DDR3 1600 DIMMs).

It may prove to be useful information for extreme budget iGPU gamers. (re: 1 x 4GB DDR3 1600 is usually cheaper than 2 x 2 GB DDR3 1600, and 1 x 4GB DDR3 1600 is the type of RAM a person will most likely encounter in value bundles)

 

[Enigmoid]

That s not possible, take a laptop with a 35W CPU and run just CB 11.5,
do you realize what is 50W.?.

The case has a 90W PSU, why would it need to be upgraded.?..

Many thin and light laptops support way more power than a case that size. Its just a case of designing a cooling system that can handle that heat.

The Antec ISK 110 (picture you linked) can easily support a SB i3 (65W).

Anandtech even did an review.

http://www.anandtech.com/show/6192/antec-isk-110-vesa-case-review-just-about-as-small-as-it-gets/4

Ah, and Celeron dont even support 4K and a few other desirables features.

4K and AM1/Celeron really don't go together.

 

[frozentundra123456]

4K and AM1/Celeron really don't go together.

This. I was just sort of sitting here laughing to myself about promoting 4k support on such anemic processors. It is like putting racing tires on a prius. I also doubt anyone using these processors is going to be handicapped by the other so called "missing" features of the celeron. What they are going to be handicapped by is the lousy single thread performance, especially on the athlon.

 

[Abwx]

Many thin and light laptops support way more power than a case that size. Its just a case of designing a cooling system that can handle that heat.

The Antec ISK 110 (picture you linked) can easily support a SB i3 (65W).

Anandtech even did an review.

http://www.anandtech.com/show/6192/antec-isk-110-vesa-case-review-just-about-as-small-as-it-gets/4

That s not adapted, the fan will have to pull a lot of air making it noisy, the HDD and SSD are on the other side, the case coverage is mainly plastic giving it a poor thermal resistance, the Kabinis fits well because they are not actual 25W APUs, this case is good up to 50W, the power supply is 90W, this is not enough with 65W CPUs.

4K and AM1/Celeron really don't go together.

Certainly that Celeron are not adapted but AM1 has support whatever the sayings i read here and there and wich amount to pure denial.

For Frozentundra who do not understand we are talking of video content, not of games...

 

[cbn]

The Antec ISK 110 (picture you linked) can easily support a SB i3 (65W).

Anandtech even did an review.

http://www.anandtech.com/show/6192/antec-isk-110-vesa-case-review-just-about-as-small-as-it-gets/4

Here is another review with an Intel and AMD comparison at 65 watts:

http://www.anandtech.com/show/5696/puget-systems-echo-intel-and-amd-showdown-at-65-watts

 

[VirtualLarry]

A sata port support several devices, depending of the controler, this is implemented in Kabinis since it s also used in low power servers, cost should be a cable.

I think that you are confused. SATA ports and devices are 1:1, unless the controller port supports "port multiplier", which requires an expansion card (a "sata expander"), which takes the controller's port, and splits it out into multiple ports, much like a USB hub.

You can't do it with just a 'Y' cable. (Much like USB ports).

 

[Abwx]

Here is another review with an Intel and AMD comparison at 65 watts:

http://www.anandtech.com/show/5696/puget-systems-echo-intel-and-amd-showdown-at-65-watts

That's great for cooling, but does absolutely nothing to muffle noise.

The result is that both systems never see temperatures above the low 60C on the cores, and noise is essentially a non-issue unless you're looking for a "silent" system.

Notice that an AM1 system at full throughput would still be roughly below 30W in this case, i got 29.6W at the main with Prime 95.

82W with a 90W PSU.??.

Wasnt Puget the corporate that was boasting about being a reliable brand.?.

More seriously, the point is about low budget systems.

I think that you are confused. SATA ports and devices are 1:1, unless the controller port supports "port multiplier", which requires an expansion card (a "sata expander"), which takes the controller's port, and splits it out into multiple ports, much like a USB hub.

You can't do it with just a 'Y' cable. (Much like USB ports).

Some chipsets support multiple connections on a single port, no need of a card.

A Serial ATA port multiplier is a device that allows multiple SATA devices to be connected to a single SATA host port. Many common controllers do not support this feature,[1] as it is not a requirement for a SATA controller

http://en.wikipedia.org/wiki/Port_multiplier

https://ata.wiki.kernel.org/index.php/SATA_hardware_features

 

[frozentundra123456]

That s not adapted, the fan will have to pull a lot of air making it noisy, the HDD and SSD are on the other side, the case coverage is mainly plastic giving it a poor thermal resistance, the Kabinis fits well because they are not actual 25W APUs, this case is good up to 50W, the power supply is 90W, this is not enough with 65W CPUs.



Certainly that Celeron are not adapted but AM1 has support whatever the sayings i read here and there and wich amount to pure denial.

For Frozentundra who do not understand we are talking of video content, not of games...

I understand perfectly.

Edit: although you might want to revisit the title of the thread, seems like I saw the word "gaming" in there.

 

[Enigmoid]

Notice that an AM1 system at full throughput would still be roughly below 30W in this case, i got 29.6W at the main with Prime 95.

82W with a 90W PSU.??.

Wasnt Puget the corporate that was boasting about being a reliable brand.?.

More seriously, the point is about low budget systems.

The point is that a case and PSU that small can easily power much more power hungry chips. There is no advantage for an AM1 platform using the antec 110 for example.

 

[Abwx]

The point is that a case and PSU that small can easily power much more power hungry chips. There is no advantage for an AM1 platform using the antec 110 for example.

The advantage is total silence, even in silent mode Prime 95 doesnt manage to get the fan spin faster, the fan will pull very little amount of air from oustside tha case, with a bigger set up like the one on AT review the fan is spinning hard, the air is pulled with enough torque that any dust that passe close is pulled inside the case, it wont be long before the case is dust filled.

Besides i dont expect the 82W set up to be reliable, it wont be long before some peaks damage the DC converters, Puget are crazy to use a 90W only circuit to feed such PCs, 30% margin wouldnt even be totaly safe, let alone 9%, if one has the bad idea to run Linpack, Prime or even Luxmark i predict big amounts of smoke.

It s not like a pico PSU where the 12V is taken directly from the DC adaptator, in this case there s a 19V DC adaptator and the 12V has to be produced by the cased converter.

 

[Enigmoid]

The advantage is total silence, even in silent mode Prime 95 doesnt manage to get the fan spin faster, the fan will pull very little amount of air from oustside tha case, with a bigger set up like the one on AT review the fan is spinning hard, the air is pulled with enough torque that any dust that passe close is pulled inside the case, it wont be long before the case is dust filled.

Besides i dont expect the 82W set up to be reliable, it wont be long before some peaks damage the DC converters, Puget are crazy to use a 90W only circuit to feed such PCs, 30% margin wouldnt even be totaly safe, let alone 9%, if one has the bad idea to run Linpack, Prime or even Luxmark i predict big amounts of smoke.

It s not like a pico PSU where the 12V is taken directly from the DC adaptator, in this case there s a 19V DC adaptator and the 12V has to be produced by the cased converter.

AT doesn't give hard numbers but

Thankfully, the circulating cool air does wonders to keep either heatsink's fan from having to work particularly hard. Neither system is actually audible while running unless you put your ear up to the open side; any kind of ambient noise in the room will muffle them. The result is that both systems never see temperatures above the low 60C on the cores, and noise is essentially a non-issue unless you're looking for a "silent" system.

Noise is essentially a non-issue. Temps are fine.

The i7-2600S is way more power hungry than a pentium/celeron. It will throttle on stock to stay in the 65W tdp window. Without any turbo on a pentium/celeron there will not be any short term power peaks. Lets be realistic, with 3.2 ghz quad + HT + GT2 (i7-4700m) laptop cpus at 47 W there is no way a 2 core 3.2 GT1 chip is going to be bumping against a 54W tdp.

The 90W power adapter is plenty and is roughly equivalent to a 47W quad on a 15.6" laptop on a 90W adapter which can be seen a lot. Hell, apple manages to cram a 47W CPU + 750m + retina screen and run it on 85W (note that the system does throttle) brick.

 

[Abwx]

AT doesn't give hard numbers but
Noise is essentially a non-issue. Temps are fine.

Are you sure thyatnoise doesnt matter.?.

Lets be realistic, with 3.2 ghz quad + HT + GT2 (i7-4700m) laptop cpus at 47 W there is no way a 2 core 3.2 GT1 chip is going to be bumping against a 54W tdp.

The 90W power adapter is plenty and is roughly equivalent to a 47W quad on a 15.6" laptop on a 90W adapter which can be seen a lot. Hell, apple manages to cram a 47W CPU + 750m + retina screen and run it on 85W (note that the system does throttle) brick.

To be realistic would be to check Anand s review of the 4700MQ with a minimalist plateform and about 86W power draw, i wouldnt advise anybody to build such a system, neverless you are stretching the thing far from low power/low cost plateforms.

 

[cbn]

Are you sure thyatnoise doesnt matter.?.

Neither system is actually audible while running unless you put your ear up to the open side

And AM1, even at 1200 rpm, can't do any better than that:

http://www.silentpcreview.com/article1390-page4.html

Editor's Note (April 23): Reader reaction in the forum to the minimal comments made here by Lawrence Lee regarding the noise of the fan on the ASUS heatsink prompted me to revisit the anechoic chamber with the test system. I can confirm the original comment that the fan produces very little noise. It is low enough that at 1m distance in the chamber, the fan is inaudible; ie, below the ambient level, which is 10~11 dBA — extremely, unnaturally quiet. This is at the default/idle speed of ~1,200 RPM. When left under 100% load with Prime95 for about 45 minutes, the CPU temperature rose to around 46°C (26°C above ambient), and the fan sped up to nearly 1,400 RPM. The noise level barely increased, however. At one meter distance it was still more or less inaudible and very beinign, with a soft muted quality; the measured SPL was <13 dBA.

1 meter vs. putting ear up to the side of an open case is quite a difference in distance as well.

 

[MiddleOfTheRoad]

AMD can design or have intentions for whatever goal they want to, but it doesn't mean they will meet that goal.

I'm sorry but fact remains Athlon 5350 is still a $59 chip that must compete with a $42 Celeron G1820 (in some instances) due to its cost.

Now if you told me AMD has designed a $5 to $15 SOC for third world countries, then I wouldn't be comparing to the Celeron G1820.

You do realize you are completely looney tunes with these comparisons. So now I'll flip the manufacturers to prove how insane it is comparing completely different platforms with completely different missions.... This time FM2 versus Bay Trail Atoms.

I can build an A6-5400k CPU with Motherboard for the same price as this J1900 Intel Atom Bay Trail motherboard:

http://www.newegg.com/Product/Produc...-495-_-Product

But, oh man..... The Intel J1900 is completely worthless (by your logic) because the A6-5400 Radeon IGP can get double to quadruple the frame rates playing Tomb Raider and Battlefield 4 versus the sorry Intel HD Graphics.

But of course, I would never compare them. One is Intel's low power platform while the other is AMD's mainstream offering, Apples To Oranges. But you would -- because they are the same price. The Atoms and Athlon 5350 were not intended for serious gaming..... A little common sense could go a long way.

To now paraphrase using your flawed logic:
"I'm sorry but fact remains the Intel J1900 is still an $82 chip (tray price) that must compete with a $59 A6-5400K (in some instances) due to its cost."

http://ark.intel.com/products/78867/Intel-Celeron-Processor-J1900-2M-Cache-up-to-2_42-GHz

You're freaking Avatar has the Intel logo in it -- so let's get real. You are just looking for a pathetic excuse to trash talk the Athlon 5350, but you don't have a leg to stand on.... The 5350 mops the floor with the Intel J1900 in anything 3D and for many productivity tasks (Powerpoint, InDesign, 7-Zip). That is the CPU the 5350 was designed to compete against (and the AMD consistently outperforms it for $20 less).

http://www.techspot.com/review/806-amd-kabini-vs-intel-bay-trail-d/

 

[cbn]

I can build an A6-5400k CPU with Motherboard for the same price as this J1900 Intel Atom Bay Trail motherboard:

http://www.newegg.com/Product/Produc...-495-_-Product

But, oh man..... The Intel J1900 is completely worthless (by your logic) because the A6-5400 Radeon IGP can get double to quadruple the frame rates playing Tomb Raider and Battlefield 4 versus the sorry Intel HD Graphics.

That is a really bad comparison for at least three reasons:

1. You picked one of the most expensive J1900 motherboards @ $98.99 plus $5.39 shipping when there a much cheaper choices for J1900 Mini-ITX at $69.99 plus $4.99 shipping ---> http://www.newegg.com/Product/Produc...82E16813157497

2. J1900 is passive and the 65 watt Trinity/Richlands are known to be quite noisey according to some newegg reviews.

3. A6-5400K and FM2 or FM2+ Mini-ITX is classically much more expensive than the typical J1900 Mini-ITX. Even on Micro ATX, the A6-5400K will classically be a good deal more expensive.

So, in a nutshell, you are attempting to compare a somewhat noisey and more expensive platform to one that is cheaper and passively cooled.

At least with our G1820 to Athlon 5350 comparisons, both platforms are much closer in cost (actually for uATX I believe the G1820 is cheaper most of the time) and have fans on them.

Do you see the difference?

Your comparison: Two different price points and different methods of cooling (fan vs. passive)

My comparison: Same price point and same method of cooling the processor.

 

[Abwx]

And AM1, even at 1200 rpm, can't do any better than that:

http://www.silentpcreview.com/article1390-page4.html

1 meter vs. putting ear up to the side of an open case is quite a difference in distance as well.

When i mounted the thing the AM1 was very silent, inaudible actualy, it became, relatively, quite noisy once i added a 2.5" HDD to side with the SSD...

While we re discussing casing it s worth noting that the Antec ISK 110 is plastic covered , dont be fooled by the inner metallic structure, it has very poor thermal dissipation capability compared to a regular uATX capable box, Kabini fits perfectly the usage because it s not an actual 25W chip.

The power dissipated inside the box when running Prime 95 was 27W and this result in 20°C increase in inner temp, that is, about 0.8°C/W, if a system is to dissipate 65W inside such a case the inner temp would be 52°C above ambiant, at 25°C ambiant, to get a convenient value, this yield 77°C inside the case and it is in the range measured by Anand for normal use, now try running Linpack or Prime assuming the DC converters give you enough time to end a few runs, likely that even X264 encoding will be dreadfull.

Edit : The 12°C delta is for 15-18W, all numbers corrected in the post, the Puget system fan must really be pushing hard...

 

[MiddleOfTheRoad]

That is a really bad comparison.

So, in a nutshell, you are attempting to compare a somewhat noisey and more expensive platform to one that is cheaper and passively cooled.

My comparison: Same price point and same method of cooling the processor.

Comparing AM1 to 1150 is just as Apples To Oranges as comparing FM2 to Bay Trail. Have a nice day.
Your main point was that if it similarly priced, we should compare them. But please continue, I'm enjoying the backpedalling.

BTW, the G1820 pulls more electricity under load than the 5350 -- but you seem to love glossing over that issue, too.
Remember, we've got to keep spreading the bull**** that AMD chips are always more power hungry, too.

 

[Enigmoid]

Are you sure thyatnoise doesnt matter.?.

To be realistic would be to check Anand s review of the 4700MQ with a minimalist plateform and about 86W power draw, i wouldnt advise anybody to build such a system, neverless you are stretching the thing far from low power/low cost plateforms.

The review even said that the noise is minimal.

I'm still looking for the 4700mq review. Can you find it for me? The one where he gives the 86W power?

For the iris pro testing platform this was the fan they used (techreport).

The platform is idling at 37W with average power under load of 92W. The system is not optimized for mobile use at 10W higher than the 4770k. Furthermore this is power before the power brick.

When i mounted the thing the AM1 was very silent, inaudible actualy, it became, relatively, quite noisy once i added a 2.5" HDD to side with the SSD...

While we re discussing casing it s worth noting that the Antec ISK 110 is plastic covered , dont be fooled by the inner metallic structure, it has very poor thermal dissipation capability compared to a regular uATX capable box, Kabini fits perfectly the usage because it s not an actual 25W chip.

The power dissipated inside the box when running Prime 95 was 27W and this result in 12°C increase in inner temp, that is, about 0.5°C, if a system is to dissipate 65W inside such a case the inner temp would be 32.5°C above ambiant, at 27.5°C ambiant, to get a convenient value, this yield 60°C inside the case and it is in the range measured by Anand for normal use, now try running Linpack or Prime assuming the DC converters give you enough time to end a few runs, likely that even X264 encoding will be dreadfull.

Yet reviews have shown that it is perfectly capable of supporting more power hungry processors.

Comparing AM1 to 1150 is just as Apples To Oranges as comparing FM2 to Bay Trail. Have a nice day.
Your main point was that if competes on price, we should compare them. But please continue, I'm enjoying the backpedalling.

BTW, it is spelled Noisy.

PS, the G1820 pulls double the electricity under load as the 5350 -- but you seem to love glossing over that issue, too.
Remember, we've got to keep spreading the bull**** that AMD chips are always more power hungry, too.

Your comparison was terribly flawed as cbn pointed out.

Twice is untrue.

 

[MiddleOfTheRoad]

Your comparison was terribly flawed as cbn pointed out.

Twice is untrue.

It was intentionally flawed -- to mimick CBN's equally flawed comparison of AM1 to 1150.

The Sempron 2650 / 3850 / Athlon 5150 / 5350 ARE SYSTEMS ON A CHIP - SOC. Socket 1150 IS NOT.

It really doesn't get simpler than that. Just because AMD socketed a SOC so people had an upgrade path -- doesn't change the fact that it has nothing in common with 1150 architecture. That is the definition of Apples to Oranges.

 

[Abwx]

Yet reviews have shown that it is perfectly capable of supporting more power hungry processors.

Yes, with consumer grade components required to run within military specs, wich most are capable of but i wouldnt try such random builds, and reviews dont tell how this will age, likely that it will be short lived sytems.

Btw : You realise that scores of people tested AM1 plateforms and that none did find 44W unless he had a phony set up, so why are you using graphs that are just not credibles.?.

 

[MiddleOfTheRoad]

Yes, with consumer grade components required to run within military specs, wich most are capable of but i wouldnt try such random builds, and reviews dont tell how this will age, likely that it will be short lived sytems.

Btw : You realise that scores of people tested AM1 plateforms and that none did find 44W unless he had a phony set up, so why are you using graphs that are just not credibles.?.

Agreed -- those numbers seem wildly inconsistent compared to the majority of reviews out there for the 5350:
I have never seen anyone exceed 35 watts, much less 40. My Kill-A-Watt showed 26 watts maximum under
full load on my personal Sempron 3850.

http://hexus.net/tech/reviews/cpu/68473-amd-athlon-5350-28nm-kabini/?page=9

http://www.legitreviews.com/amd-athlon-5350-apu-am1-platform-review_139224/6

http://www.kitguru.net/components/c...b-apu-review-w-sapphire-r7-240-low-profile/9/