Why doesnt Anandtech review heatsinks anymore?

[Borealis7]

title says it all really...

how about a nice LGA1366/LGA1156 HSF roundup to kick-start the return of AT Cooling reviews? with blazing O/Ced i7s running at hair-raising temps and eye-popping frequencies!

or have they given up their "market-share" to FrostyTech? (the rulers of everything frosty!)

 

[imported_Scoop]

Don't really care, I only read SPCR for HSF reviews. At least they're consistent and don't use different fans on different heatsinks ffs.

 

[aigomorla]

I think it was gary's responsibility, and i think he got tired.

Also the fact that not many new sinks are out, and they probably arent getting any samples.

 

[DaveSimmons]

All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

 

[aigomorla]

its because everytime he came out with one, we would complain on his methodology.

And instead of praises, we gave him frowns.

He then later tried to adapt our wants into his methodology and i think he found out it was quite taxing.

Guys it takes many many hours to do a review.
The testing section alone takes hours, and then the editing and writing involved, its not an easy task.

But i agree with ya, we should have a heat sink section, which is why i always been telling someone, make a good thread with links on air heat sinks, and maintain it, and i'll keep it stickied.

But the last time i offered this spot (operandi) he quickly turned it down because he remembered how taxing it is to keep up with a thread like that.

But if anyone wants the thread job, go ahead. If its good, i'll sticky it, and all you need to do is maintain it so its kept up to date.

 

[HIO]

Originally posted by: DaveSimmons
All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

Yeah I have spent a considerable amount of time there recently and only this morning realized they were my neighbors in Vancouver, Washington right across the river from me in Oregon. Thirty minute drive tops from my home in Lake Oswego.

I just hope Anandtech doesn't get even close to what Tomshardware is doing as of late. That place with all of the ads and stuff is really annoying now and way, way to much. Ughhh! I don't even like going there anymore. I know, I know everyone's got to make a living,.. but puhleeeze.

HIO

 

[The Boston Dangler]

Originally posted by: DaveSimmons
All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

http://www.frostytech.com/

 

[Ayah]

Originally posted by: The Boston Dangler

Originally posted by: DaveSimmons
All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

http://www.frostytech.com/

They don't do LGA1156 or 1336.

 

[aigomorla]

theres a 1336??

I want a 1336 cpu!

 

[Idontcare]

Originally posted by: Ayah

Originally posted by: The Boston Dangler

Originally posted by: DaveSimmons
All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

http://www.frostytech.com/

They don't do LGA1156 or 1336.

Are there any LGA1156 or LGA1366 only HSF?

My understanding was they are all the same HSF but with a different mounting adapter and as such they should technically all handle the same 150W heatload dissipation test and noise test with the same efficacy.

They are labeled LGA775 vs. AM2 reviews but for all intents and purposes they are really socket-agnostic HSF reviews with the buyer making sure they procure the correct socket adapter.

 

[Rubycon]

Originally posted by: aigomorla
theres a 1336??

I want a 1336 cpu!

Intel 1492 is guaranteed to overclock too! :laugh:

On a serious note...

While it's nice to get an idea what temperatures would be like on YOUR cpu, heatsink testing methodology leaves a lot to be desired. Chips - particularly quad cores heavily loaded, overclocked and over volted - produce hundreds of watts of heat!

I would like to see actual capacity numbers of heatsinks along with the effects of multiple mounting positions which in the case with heatpipes get critical as you approach the vapor capacity of the pipe tally. The final (laymen) results could be a performance index. Also it's VERY important to use IDENTICAL fans with each one. (even though said fan will move differing amounts of air due to fin spacing, etc.)

The idea is to get realistic apples to apples results and total capacity of the units. Heatsinks including their own fans (Tuniq/Zalman, etc.) will just be limited to what their supplied fan can muster. Of course the Tuniq Tower series can be fitted with an aftermarket fan...

 

[HIO]

Originally posted by: Idontcare

Originally posted by: Ayah

Originally posted by: The Boston Dangler

Originally posted by: DaveSimmons
All you need is SilentPCReview for HSF reviews. It's been years since AT had good HSF coverage.

http://www.frostytech.com/

They don't do LGA1156 or 1336.

Are there any LGA1156 or LGA1366 only HSF?

My understanding was they are all the same HSF but with a different mounting adapter and as such they should technically all handle the same 150W heatload dissipation test and noise test with the same efficacy.

They are labeled LGA775 vs. AM2 reviews but for all intents and purposes they are really socket-agnostic HSF reviews with the buyer making sure they procure the correct socket adapter.

> Are there any LGA1156 or LGA1366 only HSF?

This one is getting close.

http://www.nexustek.nl/NXS-LOW...cm-high-CPU-Cooler.htm

 

[DaveSimmons]

Originally posted by: Rubycon

I would like to see actual capacity numbers of heatsinks along with the effects of multiple mounting positions which in the case with heatpipes get critical as you approach the vapor capacity of the pipe tally. The final (laymen) results could be a performance index. Also it's VERY important to use IDENTICAL fans with each one. (even though said fan will move differing amounts of air due to fin spacing, etc.)

The idea is to get realistic apples to apples results and total capacity of the units. Heatsinks including their own fans (Tuniq/Zalman, etc.) will just be limited to what their supplied fan can muster. Of course the Tuniq Tower series can be fitted with an aftermarket fan...

SilentPCReview usually installs the exact same Nexus 120mm fan on all of the HSs as well as testing any included fan.

They generally only test the "standard" installation position for the HSF, but they do give a good idea of the heat dissipation ability of the HS for both low and high airflow situations.

My understanding was they are all the same HSF but with a different mounting adapter and as such they should technically all handle the same 150W heatload dissipation test and noise test with the same efficacy.

Right, the major HSFs come with 3+ sets of mounting hardware for different sockets but there isn't much point to testing the heat dissipation with 10 different CPUs, picking one hot one tells you what to expect about the others.

I'm usually more concerned with lower noise that uber-cooling so SPCR suits my needs but it sounds like FrostyTech is good too.

 

[Zap]

Originally posted by: aigomorla
Guys it takes many many hours to do a review.
The testing section alone takes hours, and then the editing and writing involved, its not an easy task.

That's why there are a lot of crappy 4 page reviews out there with first page being manufacturer's marketing info and box shots, second page being pictures of product, third page being "here's a Prime95/OCCT 30 min run" and fourth page being "conclusion: my CPU didn't burn up so 9/10 flybynite.com award plz visit sponsor links k thx bye."

 

[Rubycon]

Originally posted by: DaveSimmons

SilentPCReview usually installs the exact same Nexus 120mm fan on all of the HSs as well as testing any included fan.

They generally only test the "standard" installation position for the HSF, but they do give a good idea of the heat dissipation ability of the HS for both low and high airflow situations.

I'm sure a special rig could be engineered from the ground up for capacity measurement. This prevents IHS, contact patches, cure time, etc. behind actual performance. In the end a performance index could be established indicative of actual thermal transfer capacity. This (of course) would not be a cut and dry decision for purchasing as brand B with a TPI of 6.6 may NOT keep your quad cooler than brand C with a TPI of 5.7. Many times it can be operator error. A well designed mounting system prevents common mistakes. The Prolimatech Megahalems, for example; has a decent mounting system for consistent results from user to user.

 

[Idontcare]

Originally posted by: Rubycon

Originally posted by: DaveSimmons

SilentPCReview usually installs the exact same Nexus 120mm fan on all of the HSs as well as testing any included fan.

They generally only test the "standard" installation position for the HSF, but they do give a good idea of the heat dissipation ability of the HS for both low and high airflow situations.

I'm sure a special rig could be engineered from the ground up for capacity measurement. This prevents IHS, contact patches, cure time, etc. behind actual performance. In the end a performance index could be established indicative of actual thermal transfer capacity. This (of course) would not be a cut and dry decision for purchasing as brand B with a TPI of 6.6 may NOT keep your quad cooler than brand C with a TPI of 5.7. Many times it can be operator error. A well designed mounting system prevents common mistakes. The Prolimatech Megahalems, for example; has a decent mounting system for consistent results from user to user.

Ruby what you are getting at would be handled by using a standardized test platform akin to what frostytech does (calibrated thermal load at fixed 75, 90, 125 and 150W) but stepped up to test the 200-300W regime as well.

Then the data would need to be crunched to represent Thermal Resistance plotted against Thermal Load...such a plot (Thermal Load vs. Thermal Resistance) would capture the diminishing efficacy of the HSF system as thermal loads increased to the point of being near, or exceeding, the vapor capacity of the pipes.

The same graph could also be used to capture the Thermal Load vs. Temperature data by way of using a second y-axis on the same plot.

(and of course they need to standardize on the same fan for all tests, and do their high/low fanspeed for noise and thermal resistance comparisons)

And then last but not least the data needs to be tabulated in a way that lends itself readily to rank-sorting by price/performance (performance presumably won't change with time, but price will).

 

[Rubycon]

Yes bring the wattage way up to 400+ just to see where the run away takes place under optimal positioning then show where the approach is under ideal positioning (i.e. motherboard vertical in case, etc.).

 

[BonzaiDuck]

Originally posted by: Rubycon

Originally posted by: DaveSimmons

SilentPCReview usually installs the exact same Nexus 120mm fan on all of the HSs as well as testing any included fan.

They generally only test the "standard" installation position for the HSF, but they do give a good idea of the heat dissipation ability of the HS for both low and high airflow situations.

I'm sure a special rig could be engineered from the ground up for capacity measurement. This prevents IHS, contact patches, cure time, etc. behind actual performance. In the end a performance index could be established indicative of actual thermal transfer capacity. This (of course) would not be a cut and dry decision for purchasing as brand B with a TPI of 6.6 may NOT keep your quad cooler than brand C with a TPI of 5.7. Many times it can be operator error. A well designed mounting system prevents common mistakes. The Prolimatech Megahalems, for example; has a decent mounting system for consistent results from user to user.

Joe Citarella did that at the Overclockers.com web-site. I haven't checked that site for good part of a year.

He'd published an article in January, 2007 (Jan 17, I think) on nano-diamond thermal paste. He used a $28,000 calorimeter for testing. He'd always posted a disclaimer in his articles: "The author has invested in a small company pursuing cooling solutions for computers." [for the purpose of making money.] Honest enough! Four months later, Innovative Cooling's Diamond thermal paste appeared on the market. But it wasn't on the market when he published in January 07, showing its favorable performance in comparison tests!

Overclocker's -- at one time, had some very thorough heatsink reviews and rankings, and Citarella was pretty much in charge of it.

 

[Ayah]

Should review heatsinks with Ruby's W3520. Watch most heatsinks get a fail rating.

 

[Rubycon]

Originally posted by: BonzaiDuck

Joe Citarella did that at the Overclockers.com web-site. I haven't checked that site for good part of a year.

He'd published an article in January, 2007 (Jan 17, I think) on nano-diamond thermal paste. He used a $28,000 calorimeter for testing. He'd always posted a disclaimer in his articles: "The author has invested in a small company pursuing cooling solutions for computers." [for the purpose of making money.] Honest enough! Four months later, Innovative Cooling's Diamond thermal paste appeared on the market. But it wasn't on the market when he published in January 07, showing its favorable performance in comparison tests!

Overclocker's -- at one time, had some very thorough heatsink reviews and rankings, and Citarella was pretty much in charge of it.

Yes and the actual performance of (IC7) is more dependent on user application than anything else out there. Do it just ever so slightly (wrong) and you have WORSE temps than distant competitors. Many get discouraged by this and write it off quickly.

Originally posted by: Ayah
Should review heatsinks with Ruby's W3520. Watch most heatsinks get a fail rating.

Depends on how much wrist twisting the reviewer is willing to tolerate.

 

[Zap]

Originally posted by: BonzaiDuck
Joe Citarella did that at the Overclockers.com web-site. I haven't checked that site for good part of a year.

IMO the main site isn't too interesting anymore. Sure, they had decent HSF reviews, but their other product reviews (of mediocre PSUs and an endless stream of external HDD enclosures) weren't that great. They also still owe me $25. :| I was still going there to read the editorials by Ed Stroligo (IMO THE ABSOLUTE BEST TECH EDITORIALS HANDS DOWN ALL CAPS BOLDED) but he quit/retired and it hasn't been the same since. More recently they got acquired by some media company.