The TR Ultra-90

[firewolfsm]

It's selling for only $20 and I'm really considering it, how much worse than the Ultra-120/Ultima-90 is it? How much better than stock? Assuming it's running a quiet 120mm fan.

http://www.svc.com/ultra-90-775.html

 

[firewolfsm]

No one?

 

[thilanliyan]

CHeck the Anandtech review. It does fairly well I think.

 

[Quiksilver]

Originally posted by: thilan29
CHeck the Anandtech review. It does fairly well I think.

TR Ultra 90 not TR Ultima 90

 

[Quiksilver]

Soz, Double Post.

 

[VinDSL]

I don't know if I'd trust it...

The base is flat "well finished and attractive", which TR now claims is a BAD thing!

 

[kef7]

For the price I'm tempted to try it in a qpack as a replacement for a zalman 7000.

 

[BonzaiDuck]

It'll be better than your stock cooler, almost as good as the U-120 (original), slightly behind the U-120-Extreme, the Ultima-90 and the Sunbeam Tuniq. Go through the Anandtech comparison reviews of coolers since April. Cut to the chase and go to the comparison tables. I'm sure it's in there somewhere.

If it isn't, OverClockers.com has a review for it, and actual measurement of thermal resistance. I fhte U-120 (original) is around 0.12 C/W, I'd expect the U-90 to be somewhere between 0.12 and 0.13.

You might be able to OC your Q6600 to 3.0 Ghz with the stock multiplier, but I'd aim for keeping the hottest core temperature below 68C at room ambients approaching 80F. If 3.0 Ghz seems uncomfortable, try dropping the multiplier to 8 and boosting the FSB so the CPU runs at between 2.8 and 2.88 Ghz. That will certainly be manageable for an Ultra-90.

For those earlier Ultra tower coolers, the base of the cooler is flat. At least, I can guarantee such for the U-120 (original), so the U-90 would seem a near-certainty.

 

[firewolfsm]

Bonzai, I have it at 2.9GHz on the stock cooler already (loads at 66C), if this thing is even close to as good as the U-120, I should make it to 3.4GHz, thanks for the advice, I'm ordering it now.

I also lapped my Q6600 so this will save me the time of doing the heatsink.

...I wonder why this thing is so cheap?

 

[BonzaiDuck]

Originally posted by: firewolfsm
Bonzai, I have it at 2.9GHz on the stock cooler already (loads at 66C), if this thing is even close to as good as the U-120, I should make it to 3.4GHz, thanks for the advice, I'm ordering it now.

I also lapped my Q6600 so this will save me the time of doing the heatsink.

...I wonder why this thing is so cheap?

_______________________

Ah! You have the G0 stepping. Depending on the room ambient, I'm guessing that 3.2Ghz might show you loads of 65C in the high 70F ambient range -- even between 78 and 80, but this isn't rational calculation on my part -- I'm making an educated guess. With my B3, my load value at 78F is between 64 and 65C. But I'm using the TR-Ultra-120-Extreme. For an E6600 as opposed to a Q6600, moving from the U-120 to U-120-Extreme meant an improvement of 5 degrees Celsius. Citarella's review of the Ultra 90 proves a thermal resistance of 0.11 C/W, but with fan speeds around 4,000 rpm. I suppose you'll just need to find out . . . .

"Why this thing is so cheap?"

Before I started hobnobbing with Anandtech geeks, I'd wait a year to buy last year's technology and reap the benefits of price decline. Because of people like Anandtech geeks, manufacturers expect to capture their research and development costs in the earlier part of the production cycle, so the supply-price is high, just as you would be willing to pay more in demand-price for a new innovation. As the number of potential consumers become actual consumers and the number of "new" consumers is depleted, you have "market saturation." The manufacturer, even for shipping more units to resellers, will lower their asking price. You also have -- through trial, error and practice -- a reduction in production costs.

It's last-year's cooler. Last year's is always cheaper next year . . . .

It could be that the current reseller price reflects actual manufacturing cost, or it could also be that someone, somewhere -- is taking a hit on either production or inventory investment because it's 'surplus stock" that needs to be unloaded.

Go for it.

 

[firewolfsm]

A sanded G0! I didn't have those temps before the lap. Thanks for the advice, you're one of the few members who knows what he's talking about.

As for the price, coolers and fans usually don't see price drops like that so I think someone is taking a hit, like you said...but that's for another thread.

 

[VinDSL]

Gotta say...

What I've noticed, over the years, is cooler technology keeps up with increases in CPU temps - they have to!

So, if you're running last year's CPU, last year's cooler will work just fine. However...

Put another way, last year's cooler is hardly designed to work with the temps that a Q6600 will produce.

That's why they're dumping them for pennies on the wholesale dollar!

Of course, there are some exceptions to that, like a Tower 120. These things were so brutal from the get-go that they still work great!

Then, there's 'planned obsolescence', so called:

Rationale behind the strategy

A new product development strategy that seeks to make existing products obsolete may appear counter intuitive, particularly if coming from a leading marketer of the existing products. Why would a firm deliberately endeavour to reduce the value of its existing product portfolio?

The rationale behind the strategy is to generate long-term sales volume by reducing the time between repeat purchases, (referred to as shortening the replacement cycle). Firms that pursue this strategy believe that the additional sales revenue it creates more than offsets the additional costs of research and development and opportunity costs of existing product line cannibalization.

That's the marketing model TR uses...

 

[BonzaiDuck]

VinDSL -- wish you'd get off the soapbox selling snake-oil.

It's only about two essential issues: thermal wattage at the processor cap, and thermal resistance of the cooler. Once one admits that fan-speed also makes a difference, you can throw in noise dBA as a consideration -- a toxic by-product of enhanced cooling to an asymptotic limit.

For the history of processor thermal properties, AMD went out on the limb and was awarded for being "green" by comparison to Intel -- and I've been an Intel user since the early 1990s (having flirted with AMD processors during that stone-age.)

For Intel, the TDP spec marched upward from the Northwood "C" processor's 80-watt spec. The Prescott measured at 105W in earlier incarnations, and of course they also followed a die-spec downward from 0.13 micron to 95 nm on the Pressy. At the same time, they were adding millions of transistors to the Pressy that the Northwood didn't have, and this would increase thermal wattage again. They announced they'd hit a "thermal barrier" in processor design in conjunction with voltage leaks, but they turned this around again with the 65 nm size reduction, even though the Smithfield and early Presler dual-cores were toaster-ovens for thermal wattage. Then the C2D brought down the TDP to 65W, but by doubling up two Conroe's to build the Kentsfield, the thermal wattage jumps back again to stock values of between 105 and 110W.

So with the 45 nm generation of cores -- we could expect again an ebb in the thermal wattage, and ANY earlier cooler [only including -- but not limited to -- your precious Tuniq] -- could be resurrected to "make nice" for the newer generation of cores.

 

[BonzaiDuck]

Anyway, I don't disagree with the ideas behind your "planned obsolescence" argument, but in this type of competitive market, the innovations still must perform to meet the competition.

You could call it "planned obsolescence," but that wouldn't fit the case with releasing three different coolers at the same time to fit different market-niche preferences in the enthusiast community -- so that TR put forward the Ultra-120 "extreme" revision, the IFX-14 "inferno," and the Ultima 90 all within the course of a few-months time. Each of those coolers meets a different set of preferences by users. The huge IFX-14 only has 4 heatpipes but large fin area; the 120-Extreme has eight pipes and less in the way of fins; the Ultima offers approximately the same cooling advantages with a cooler that fits better in "tight" situations.

Not too different than Toyota producing a top-end sedan, an SUV and a sub-compact.

But TR is no GM or Toyota in absolute size. I'd only guess -- that for now -- they're a "dominant firm" in a very competitive market. By contrast, Sunbeam doesn't specialize in coolers alone, and they subcontract their products to other firms. They've also been less-than-stellar about "truth in advertising" for at least one of those products.