It's an 80 mm fan. Using that as a bogo ruler, I'd guestimate 90-100mm X 80mm x 50-70 mm. The narrow dimension is the same size as the fan, so that's a gimmie. It's slightly larger on the other axis, since the shroud extends a little beyond the fan. 5-10mm/side should bracket it, probably on the low end. Thickness is the hardest part, but the fan looks like an conventional 25mm thick design; the fins are slightly thinner than the fan, but the height from the heatpipe contact plate isn't visible; but the entire assembly appears shorter than it's wide.
I would've posted last night, but I had to check and make sure. Sometimes my memory plays tricks on me. ;) Anyways, my X2 4400+ came with that wonderful heatsink. Would still be using it today, but none of my operating PCs are using Socket A. Nice to see that the included HSF is that good.
Actually I think that would be socket 939. Which I think (???) had the same/compatible clips as used on AM2 and 3. So you might actually be able to still use it on an AM3+ board?
I think he's trying to say that he thinks he can use a 939 chip on AM sockets which is incorrect for a number of reasons. First and foremost it is physically incompatible. You can't use 939 chips on AM platforms and vice versa. AM2 and AM2+ were a type of 940 pin socket (not necessarily compatible with any other 940 pin socket), and AM3 and AM3+ were 941 pin sockets. You could use AM3 chips in (some?) AM2 boards IIRC, but not AM2 chips in an AM3 board (again going by memory).
To make matters more complex, there were multiple chips that used the X2 "4400+" name. If it was clocked a hair higher with 512KB L2 cache, it was the later-gen AM2 like a 65nm Brisbane or something (DDR2). If it was clocked slightly lower with 1MB cache then I *believe* it's either a Toledo or the other one I can't remember the name of, both of which were 90nm and I believe both of which were socket 939 (and DDR1). I'm too lazy to look up the precise details but that's the gist of it.
I personally still have a machine with a functioning Epox nForce 2 based board and an FX-60. Doesn't see much use but it's not like I can sell it for anything - DDR1 and such is practically obsolete.
Oh and Zen is AM4. All Zen chips, low-end, APUs, and CPUs. About darn time - for their sake it can't release soon enough.
Well I said clips, not chips. All I meant is you might be able to re-use the Opteron 165 heatsink on an AM3+ board. Not the processor itself obviously.
It would be pretty hilarious if AM4 kept the same/compatible mounting as AM3+ in this case, because an Opteron 165 heatsink on a Zen chip somehow would be re-use awesome =D
On a total tangent, from your mention of the 9NPA...
I actually had an Epox 8NPA, which was a really a bizarre but fun idea - It was a socket 754 board that used the nForce4 ultra. So yes single core but SLI. It made sense for that brief period where 754 chips were super cheap, dual-core wasn't quite there for games - but that window was but a few months before 939 chips became cheap (at least single core 939) which obviously made the board pointless. But still credit for a brave idea.
Sadly Epox retreated shortly after that from the retail market (or at least worldwide) ... I think they're still around as SUPoX or something, but I only find hits in asian markets and maybe just OEM?
I just wished for a last bios update that have let me run the Newark 4000+ mobile chip I had... argh that chip, never had a board that would run it right. *sigh* 1Mb cache, stock voltage of 1.35v @ 2.6Ghz. And none of the 4 754 boards I had ever had a bios that would run it right. aarrghh
Sorry, I misread that. The single core wasn't a big deal in 754's heyday, mostly due to cost being substantially less than 939 setups. I had a friend that went 754 and I helped them with the build - another limitation was that it only had a single memory channel. Though again at that price it wasn't a dealbreaker. I went 939 for the upgrade path, started with a less expensive single core and eventually picked up an FX-60 on the cheap along with more (and faster) RAM.
bigboxes: There were models of 4400+ that were socket 939, and there were models that were AM2. There were obviously differences but based on performance they were both marketed as 4400+. That happened all the time back then.
No, xrror was saying that the 4400+ was Socket 939 (which it was) and was saying that the cooler that came with it may have compatibility with an AM3+ mobo.
My 4400+ was definitely a 939 cpu. I was using Abit at that time. I think it was a Toledo core.
I still own that cooler! What a great piece of steampunk, that was. It shipped with my Athlon XP and never failed. The CPU itself was such a solid performer that it was a long time before I upgraded.
In addition to architecture improvements, this is the sort of thing we need to see from AMD: Consideration for their hardware at every stage. I'm pleased to hear about the Wraith. It is effectively targeted at users who don't know or care that they can make their PCs quieter.
I was also thinking I have seen this cooler before Intel totally trounced them. I'm guessing the fan is quieter though. But the heatsink looks like it has been done before.
They should use a range of fans (speed wise) with it since it seems the fan is fixed speed. What's the point in bundling this with a 65W APU when a slower fan would do the job at lower noise. Assuming it's not gonna be for 90W and up only SKUs.
Because the fan noise is already at the point where ambient noise and the rest of the system would drown it out, so why go slower. Having more air passing through also will allow it to hit turbo more.
Uh, you do both realize that the demonstration in the video was likely run off of a straight 12V DC power supply, right? It demonstrates maximum noise from the cooler, which is kind of the point of a video like that. There is no such thing as a fixed RPM computer fan, and even if it lacked PWM control (which I seriously doubt - haven't seen a CPU cooler without that for at least a decade) it would be controllable through voltage.
Maybe before thinking you know better, you could get informed. The fan is constant speed and if you do that , you might as well use the right fan for the TDP/speed for better results.
Let me get this straight. You're saying this fan (which runs at maximum speed in the demonstration) can only run at this speed and that PWM wouldn't change the speed of the fan like it does for all other fans? Or are you trying to say that they should make fans with varying maximum speed? Both seems fairly wrong. The fan will be temperature regulated through PWM. So it will most likely run lower than the demonstrated speed the majority of the time. A lot of people buy Noctua fans running at 1200 RPM and reduce the speed themselves to like 500 RPM.
That's at full speed, under more typical conditions it will be a lot quieter. Keep in mind this is not an enthusiast cooler, but it is a heck of an upgrade over the existing boxed cooler and probably good enough for most average prebuilts or budget models that aren't being overclocked.
A better cooler means a better baseline and that's always welcome. But what about better processors? I don't know, something that can compete with intel maybe?
You should go get a job with AMD so that you can show them how to do the job properly. I bet if you worked there, Zen would've been released sometime last year, right? Because obviously you know better than they do.
It's nice to see AMD working on cooler improvements, but I'm worried about the 125 watt rating as a sign of power consumption and heat output for the next generation of processors. It would indicate that there isn't going to be substantial improvements in that respect and we can probably anticipate 95 watt chips again which, frankly, is too much for a modern CPU. Hopefully, it's just overengineered and Zen will be released at a more reasonable 45-65 watts, but I suspect that AMD won't be able to pull that off and still offer performance advancements to catch up to the per core capabilities of their nearest competitor.
They are probably just designing a 'do-all' cooler that will be able to work with basically everything. I would assume we would see future Zen APU's in the 45-65-95w ranges like current APU's and then 95-125w cpu only/high core count chips on the high end, which is pretty much just like what intel is doing.
What he said. I've got an MSI GTX970 Gaming 4G card that has a cooler that lights up. You can even change it to pulse to the music or some other nonsense. Am I ever going to "need" it? Hell no. I turn all of that off. Now, if I had a window in my case and kept it on top of my desk to show it off then of course I would love that eye candy.
I was surprised by that too. But my system builder randomly put my AMD system a case with a transparent window that I didn't specify, so I'll be happy to add a backlit CPU fan that I didn't specify.
I think 39 decibels is the max setting, even their current fans run quieter than that with the lowest PWM setting (albeit the last two I've seen from them had coil whine from the fan - though that sound did go away after running the fans through their paces for a couple days straight... and I can hear well above the normal human frequency range).
"The vast majority of enthusiasts, who buy AMD FX CPUs, usually obtain their thermal solutions from companies like Corsair, Noctua or Scythe."
So I normally wouldn't nitpick grammar on a tech article but that comma you have before "who" makes the sentence factually wrong because the vast majority of enthusiasts do NOT buy AMD FX CPUs...
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
54 Comments
Back to Article
Pissedoffyouth - Friday, January 8, 2016 - link
What's the dimensions?DanNeely - Friday, January 8, 2016 - link
It's an 80 mm fan. Using that as a bogo ruler, I'd guestimate 90-100mm X 80mm x 50-70 mm. The narrow dimension is the same size as the fan, so that's a gimmie. It's slightly larger on the other axis, since the shroud extends a little beyond the fan. 5-10mm/side should bracket it, probably on the low end. Thickness is the hardest part, but the fan looks like an conventional 25mm thick design; the fins are slightly thinner than the fan, but the height from the heatpipe contact plate isn't visible; but the entire assembly appears shorter than it's wide.boostern - Friday, January 8, 2016 - link
It's very similar to the one bundled with the Opteron back in the Athlon64 days!BigLan - Friday, January 8, 2016 - link
That's what I was thinking, though if I remember right the opteron one had a copper base.xxtypersxx - Friday, January 8, 2016 - link
Oh man, you're right. Opteron 165 @ 2.6ghz was an absolute BEAST of a chip.bigboxes - Saturday, January 9, 2016 - link
I would've posted last night, but I had to check and make sure. Sometimes my memory plays tricks on me. ;) Anyways, my X2 4400+ came with that wonderful heatsink. Would still be using it today, but none of my operating PCs are using Socket A. Nice to see that the included HSF is that good.PIcs of my old HSF:
http://i256.photobucket.com/albums/hh175/bigboxes/...
http://i256.photobucket.com/albums/hh175/bigboxes/...
xrror - Saturday, January 9, 2016 - link
Actually I think that would be socket 939. Which I think (???) had the same/compatible clips as used on AM2 and 3. So you might actually be able to still use it on an AM3+ board?bigboxes - Saturday, January 9, 2016 - link
You are right. 939 it is. Is Zen not going to be a totally new socket?Alexvrb - Saturday, January 9, 2016 - link
I think he's trying to say that he thinks he can use a 939 chip on AM sockets which is incorrect for a number of reasons. First and foremost it is physically incompatible. You can't use 939 chips on AM platforms and vice versa. AM2 and AM2+ were a type of 940 pin socket (not necessarily compatible with any other 940 pin socket), and AM3 and AM3+ were 941 pin sockets. You could use AM3 chips in (some?) AM2 boards IIRC, but not AM2 chips in an AM3 board (again going by memory).To make matters more complex, there were multiple chips that used the X2 "4400+" name. If it was clocked a hair higher with 512KB L2 cache, it was the later-gen AM2 like a 65nm Brisbane or something (DDR2). If it was clocked slightly lower with 1MB cache then I *believe* it's either a Toledo or the other one I can't remember the name of, both of which were 90nm and I believe both of which were socket 939 (and DDR1). I'm too lazy to look up the precise details but that's the gist of it.
I personally still have a machine with a functioning Epox nForce 2 based board and an FX-60. Doesn't see much use but it's not like I can sell it for anything - DDR1 and such is practically obsolete.
Oh and Zen is AM4. All Zen chips, low-end, APUs, and CPUs. About darn time - for their sake it can't release soon enough.
Alexvrb - Saturday, January 9, 2016 - link
Err, nForce 4. FOUR Alex. I think it was a 9NPA+ with the Ultra chipset. nForce 2 was in my machine before that, Socket A.xrror - Sunday, January 10, 2016 - link
Well I said clips, not chips. All I meant is you might be able to re-use the Opteron 165 heatsink on an AM3+ board. Not the processor itself obviously.It would be pretty hilarious if AM4 kept the same/compatible mounting as AM3+ in this case, because an Opteron 165 heatsink on a Zen chip somehow would be re-use awesome =D
On a total tangent, from your mention of the 9NPA...
I actually had an Epox 8NPA, which was a really a bizarre but fun idea - It was a socket 754 board that used the nForce4 ultra. So yes single core but SLI. It made sense for that brief period where 754 chips were super cheap, dual-core wasn't quite there for games - but that window was but a few months before 939 chips became cheap (at least single core 939) which obviously made the board pointless. But still credit for a brave idea.
Sadly Epox retreated shortly after that from the retail market (or at least worldwide) ... I think they're still around as SUPoX or something, but I only find hits in asian markets and maybe just OEM?
I just wished for a last bios update that have let me run the Newark 4000+ mobile chip I had... argh that chip, never had a board that would run it right. *sigh* 1Mb cache, stock voltage of 1.35v @ 2.6Ghz. And none of the 4 754 boards I had ever had a bios that would run it right. aarrghh
Alexvrb - Sunday, January 10, 2016 - link
Sorry, I misread that. The single core wasn't a big deal in 754's heyday, mostly due to cost being substantially less than 939 setups. I had a friend that went 754 and I helped them with the build - another limitation was that it only had a single memory channel. Though again at that price it wasn't a dealbreaker. I went 939 for the upgrade path, started with a less expensive single core and eventually picked up an FX-60 on the cheap along with more (and faster) RAM.bigboxes: There were models of 4400+ that were socket 939, and there were models that were AM2. There were obviously differences but based on performance they were both marketed as 4400+. That happened all the time back then.
bigboxes - Saturday, January 9, 2016 - link
No, xrror was saying that the 4400+ was Socket 939 (which it was) and was saying that the cooler that came with it may have compatibility with an AM3+ mobo.My 4400+ was definitely a 939 cpu. I was using Abit at that time. I think it was a Toledo core.
hastypixels - Tuesday, January 26, 2016 - link
I still own that cooler! What a great piece of steampunk, that was. It shipped with my Athlon XP and never failed. The CPU itself was such a solid performer that it was a long time before I upgraded.In addition to architecture improvements, this is the sort of thing we need to see from AMD: Consideration for their hardware at every stage. I'm pleased to hear about the Wraith. It is effectively targeted at users who don't know or care that they can make their PCs quieter.
eanazag - Monday, January 11, 2016 - link
I was also thinking I have seen this cooler before Intel totally trounced them. I'm guessing the fan is quieter though. But the heatsink looks like it has been done before.jjj - Friday, January 8, 2016 - link
They should use a range of fans (speed wise) with it since it seems the fan is fixed speed. What's the point in bundling this with a 65W APU when a slower fan would do the job at lower noise. Assuming it's not gonna be for 90W and up only SKUs.tipoo - Friday, January 8, 2016 - link
Because the fan noise is already at the point where ambient noise and the rest of the system would drown it out, so why go slower. Having more air passing through also will allow it to hit turbo more.Valantar - Friday, January 8, 2016 - link
Uh, you do both realize that the demonstration in the video was likely run off of a straight 12V DC power supply, right? It demonstrates maximum noise from the cooler, which is kind of the point of a video like that. There is no such thing as a fixed RPM computer fan, and even if it lacked PWM control (which I seriously doubt - haven't seen a CPU cooler without that for at least a decade) it would be controllable through voltage.jjj - Friday, January 8, 2016 - link
Maybe before thinking you know better, you could get informed. The fan is constant speed and if you do that , you might as well use the right fan for the TDP/speed for better results.Trixanity - Friday, January 8, 2016 - link
Let me get this straight. You're saying this fan (which runs at maximum speed in the demonstration) can only run at this speed and that PWM wouldn't change the speed of the fan like it does for all other fans? Or are you trying to say that they should make fans with varying maximum speed? Both seems fairly wrong. The fan will be temperature regulated through PWM. So it will most likely run lower than the demonstrated speed the majority of the time. A lot of people buy Noctua fans running at 1200 RPM and reduce the speed themselves to like 500 RPM.bigboxes - Saturday, January 9, 2016 - link
"A lot of people buy Noctua fans running at 1200 RPM and reduce the speed themselves to like 500 RPM."*raises hand*
TeXWiller - Friday, January 8, 2016 - link
From the Youtube clip you can see the four pin connection, meaning it is a PWM controlled fan.boostern - Friday, January 8, 2016 - link
Because you can control the fan through the motherboard settings.Flunk - Friday, January 8, 2016 - link
39dba isn't "nearly silent", but it is at least much more tolerable then their current, rightfully derided, solution.Alexvrb - Friday, January 8, 2016 - link
That's at full speed, under more typical conditions it will be a lot quieter. Keep in mind this is not an enthusiast cooler, but it is a heck of an upgrade over the existing boxed cooler and probably good enough for most average prebuilts or budget models that aren't being overclocked.bug77 - Friday, January 8, 2016 - link
A better cooler means a better baseline and that's always welcome. But what about better processors? I don't know, something that can compete with intel maybe?Murloc - Friday, January 8, 2016 - link
this is low-hanging fruit, it's good to pick it, but it won't solve their real issue.tipoo - Friday, January 8, 2016 - link
Check out the edge on this onespeely - Friday, January 8, 2016 - link
You should go get a job with AMD so that you can show them how to do the job properly. I bet if you worked there, Zen would've been released sometime last year, right? Because obviously you know better than they do.The_Assimilator - Saturday, January 9, 2016 - link
In his defense, a braindead snail could do a better job than AMD's management.Sttm - Friday, January 8, 2016 - link
Now leak some Zen benchmarks using that fan to OC. C'mon give Intel something to sweat about.[jF] - Friday, January 8, 2016 - link
Yeah, Intel must be really shitting themselves right about now.... :DIanHagen - Friday, January 8, 2016 - link
So that's it! They gave up fighting Intel and decided to go after Leadership.BrokenCrayons - Friday, January 8, 2016 - link
It's nice to see AMD working on cooler improvements, but I'm worried about the 125 watt rating as a sign of power consumption and heat output for the next generation of processors. It would indicate that there isn't going to be substantial improvements in that respect and we can probably anticipate 95 watt chips again which, frankly, is too much for a modern CPU. Hopefully, it's just overengineered and Zen will be released at a more reasonable 45-65 watts, but I suspect that AMD won't be able to pull that off and still offer performance advancements to catch up to the per core capabilities of their nearest competitor.extide - Friday, January 8, 2016 - link
They are probably just designing a 'do-all' cooler that will be able to work with basically everything. I would assume we would see future Zen APU's in the 45-65-95w ranges like current APU's and then 95-125w cpu only/high core count chips on the high end, which is pretty much just like what intel is doing.jimjamjamie - Friday, January 8, 2016 - link
I do hope this cooler isn't just for their high-end CPUsbigboxes - Saturday, January 9, 2016 - link
Yeah, I don't think you are getting this HSF with their bottom tier offerings.looncraz - Saturday, January 9, 2016 - link
125W for an 8-core CPU with ~Haswell IPC would be on-par, if not slightly superior to, Intel's offerings (which are 140W).Clockspeed, as always with Zen, is the wildcard (assuming they reach our much-assumed IPC target).
zodiacfml - Friday, January 8, 2016 - link
This could be overkill if AMD with CPU/APUs based on the 14nm process.timetowiseup - Friday, January 8, 2016 - link
My CPU is cold with solution.Anonymous Blowhard - Friday, January 8, 2016 - link
Now there's a flashback. BRB gtg nef ATOT more.xenol - Friday, January 8, 2016 - link
"backlit illumination"Why is this remotely important in a CPU heatsink?
jimjamjamie - Friday, January 8, 2016 - link
Because we're nerds and it's cool?bigboxes - Saturday, January 9, 2016 - link
What he said. I've got an MSI GTX970 Gaming 4G card that has a cooler that lights up. You can even change it to pulse to the music or some other nonsense. Am I ever going to "need" it? Hell no. I turn all of that off. Now, if I had a window in my case and kept it on top of my desk to show it off then of course I would love that eye candy.The_Assimilator - Saturday, January 9, 2016 - link
AMD has figured out they can't compete on performance, so now they're trying to attract idiots with bling (see: Fury X).stephenbrooks - Sunday, January 10, 2016 - link
I was surprised by that too. But my system builder randomly put my AMD system a case with a transparent window that I didn't specify, so I'll be happy to add a backlit CPU fan that I didn't specify.nagi603 - Friday, January 8, 2016 - link
"Near silent" and 39dBA should not be in the same sentence...xenol - Friday, January 8, 2016 - link
Maybe AMD thinks a typical room has a noise floor of 35dB? :Plooncraz - Saturday, January 9, 2016 - link
I think 39 decibels is the max setting, even their current fans run quieter than that with the lowest PWM setting (albeit the last two I've seen from them had coil whine from the fan - though that sound did go away after running the fans through their paces for a couple days straight... and I can hear well above the normal human frequency range).boozed - Friday, January 8, 2016 - link
What the titting cock is "supernatural style"?AnnonymousCoward - Sunday, January 10, 2016 - link
It means AMD is dead but their ghost lives on in the form of this HSF.Wolfpup - Monday, January 11, 2016 - link
Good to hear they're upgrading it. I'm interested in neither third party coolers, nor overclocking.TemjinGold - Monday, January 11, 2016 - link
"The vast majority of enthusiasts, who buy AMD FX CPUs, usually obtain their thermal solutions from companies like Corsair, Noctua or Scythe."So I normally wouldn't nitpick grammar on a tech article but that comma you have before "who" makes the sentence factually wrong because the vast majority of enthusiasts do NOT buy AMD FX CPUs...
The12pAc - Friday, January 15, 2016 - link
+1. Sad to say, but yah, +1There will always be a special place in my heart for AMD. Notice I didn't say brain.