Sapphire Radeon RX 7600 Pulse OC

[W1zzard]

any chance of trimming some games down.. i think some of them are getting old now

Intel Arc A750 Limited Edition 8 GB Graphics Card Discounted to $199

Newegg has reduced the asking price for Intel's Arc A750 Limited Edition 8 GB graphics card - this reference model is now available at an all-time low of $199. It is too early to tell whether this a permanent discount or just a seasonal sale - but the online retailer could be reacting (in...

www.techpowerup.com

 

[Chrispy_]

This Sapphire model is now one of the cheapest in the UK, and honestly it's looking better value than any of the 6600-series stock. So yes, the 7600 launch was a huge disappointment but at the same time, it's actually quite easy to recommend this card if someone cannot stretch their budget to a 12GB model like the 6700XT or 4070.

Problem of this cards against NVIDIA it's it 6 nm that makes 165W TDP vs 160W on RTX 4060 Ti that is 25% faster, so future RTX 4060 with 115W will be better than this RX 7600 on all elements: RT, DLSS, perf/w, etc..

There's a pretty good chance that the $270 7600 is faster than the $300 4060. Looking at the specs of the 4060 it's not going to be much faster than the 3060 12GB; Single-digit gains even.
For sure Nvidia will have the efficiency advantage, but I do honestly question how useful RT and DLSS are at this tier. RT performance is going to be too bad for any serious RT games even on Nvidia cards, while DLSS, FSR, and XeSS all look like garbage if you drop the native render resolution below about 900p. Like frame generation, it's a nice-enough-feature to have if your card can cope at 1080p without it, but DLSS is terrible if the base resolution is only 540p or 720p, just as frame-generation is feels horrible unless you can manage at least 60-70fps without it.

 

[doc7000]

While it wasn't the greatest of launch I Feel like it beat the 6600 launch by a good margin seeing the 6600 had a launch MSRP of $330, you are getting some 20-30% performance up lift at a $60 discount, yes the old cards are discounted and you can get a 6600 for like $200 which comparing the non discounted price to a discounted price generally favors the discounted card.

It feels like some people want to be more disappointed in this card then it really deserves, sure it may hit $250 soon though from the sounds of it the 7600 is actually selling.

 

[Roroleblaireau]

6600 had a launch MSRP of $330, you are getting some 20-30% performance up lift at a $60 discount

That's pretty honest on AMD's part as you also get 14% less silicon for this 18% "discount".
Let's not forget that these high prices were justified by silicon pricing in the first place

 

[Chrispy_]

While it wasn't the greatest of launch I Feel like it beat the 6600 launch by a good margin seeing the 6600 had a launch MSRP of $330, you are getting some 20-30% performance up lift at a $60 discount, yes the old cards are discounted and you can get a 6600 for like $200 which comparing the non discounted price to a discounted price generally favors the discounted card.

It feels like some people want to be more disappointed in this card then it really deserves, sure it may hit $250 soon though from the sounds of it the 7600 is actually selling.

The 6000-series MSRPs were based around street pricing during the ETH mining days. It isn't comparable to today's MSRPs, and it's telling that all RX6000-series cards have been significantly below their MSRPs since the ETH merge ended mining - whilst Nvidia's MSRPs didn't factor in the ETH mining supply/demand issues and as a result most of them stayed above MSRP for the entirety of 2022 and even into 2023 for some models.

ETH mining officially ended in Sept 2022, but even the dumbest miners stopped buying cards by early summer. It has been comfortably under $300 for most of its product lifecycle and selling at under $250 for almost an entire year at this point. Comparing that to something on the Nvidia side like the 3060Ti (which had an almost unrealistically-low MSRP during the ETH scalpocalypse) which still retails at it's MSRP 2.5 years later and a full year after the miners stopped buying them.

MSRPs for GPUs have stopped being meaningful for years now. Not only were they wrong back during the mining boom, they are wrong now - with all 40-series cards below the 4090 sat on shelves unsold, even with a significant discount. Some models like the 4060Ti and RX 7600 actually being discounted within mere days of launch.

 

[Kitsone]

Is that blocking a 6+2pin power header with the PCB? Teardown photo looks like it'll block the +2 plug's bump.

Bump is above the 2 pins so it looks like that PCB will block for those that(right side in picture) have that +2 plug design. I realize now I have the type on the left so I'm clear.

EDIT:
Scratch that. The PCB would need to OVERHANG the plug like the AMD "sample" version's backplate:

 

[bug]

Is that blocking a 6+2pin power header with the PCB? Teardown photo looks like it'll block the +2 plug's bump.
View attachment 298839 View attachment 298841
Bump is above the 2 pins so it looks like that PCB will block.

Not a very good picture. It looks like the one on the right would be blocked, but the one on the left might just fit.

 

[Hyderz]

lets hope there is a revision for this gpu to make the connector a better fit

 

[Kitsone]

lets hope there is a revision for this gpu to make the connector a better fit

It's all good. I edited my post. The PCB would have to overhang and it doesn't. Crisis averted

 

[sLowEnd]

lets hope there is a revision for this gpu to make the connector a better fit

AMD Confirms: RX 7600 Reference Cards in Retail will Not Have Power Connector Flaw

In the course of our testing of the reference-design AMD Radeon RX 7600 graphics card, we noticed a flaw in the physical design that could impede certain kinds of 6+2 pin PCIe power cables, causing improper power connector contact, posing a potential fire hazard, theoretically. The flaw centers...

www.techpowerup.com

 

[Aerpoweron]

@W1zzard About the cleaning the fans in the review. In the review you said:
"Sapphire Radeon RX 7600 Pulse OC exudes simplicity, so taking it apart is pretty straightforward. The cooler shroud comes off more easily than the rest of the heatsink, so you can clean its fans without having to take the whole cooler apart (and disturb the TIM or thermal pads)."

The great thing about the Sapphire cards are the easy to replace fans. One screw and you can pop out the fan. Which makes it even easier for cleaning. In that case you need to remove 2 screws (one for each fan) instead of 4 for the shroud

I also recently got 2 fans from Sapphire for the RX 480 Nitro+ for 17€ as replacements. Same easy to replace fan technology

 

[dafiek]

I mixed up something with the data, let me investigate and fix.

Edit: fixed now, thanks for reporting

25 dBA normalized isn't possible for many cards because the lowest fan speed you can set is 30% on NVIDIA, also 25 dbA isn't happening on higher-end cards. I rather have something standard, maybe 30 dBA, but some cards will overheat, which is not the goal of the test

I was comparing theese 2 cards in terms of temps/noise:
- Sapphire RX 6600 XT test
- Sapphire RX 7600 test

Here are results:

| model                            | temperature [°C] | noise [dBA] | fans [RPM] |
|----------------------------------|------------------|-------------|------------|
| RX 6600 XT                       | 70               | 27.9        | 1739       |
| RX 7600 (AMD 23.4.2 WHQL driver) | 73               | 31.4        | 1579       |
| RX 7600 (AMD Press Driver)       | 65               | 33.5        | 2145       |

From what I see, the cooling of these cards is the same or very similar, but as you can see, the RX 7600 is much louder, even with lower RPM. I was thinking what is the reason. Maybe different test methodology (2 years gap)? Then I realized that die sizes are different:
- Sapphire RX 6600 XT die size: 237 mm²
- Sapphire RX 7600 die size: 204 mm²

Does the fact that the die size is smaller affect the temperature and noise of the graphics card that much? This surprises me very much, because the RX 7600 is made using a newer technological process, I expected it to perform better.

 

[W1zzard]

but as you can see, the RX 7600 is much louder, even with lower RPM

The size of the fan, fan blade design, motor design, plenty of reasons. RPM is not fan noise, unless using the same fan config.

Maybe different test methodology

Same methodology, I just checked

Does the fact that the die size is smaller affect the temperature and noise of the graphics card that much?

It doesnt. The plate on the bottom of the cooler spreads heat out evenly and averages the effect of heat output spikes.

 

[dafiek]

The size of the fan, fan blade design, motor design, plenty of reasons. RPM is not fan noise, unless using the same fan config.


Same methodology, I just checked


It doesnt. The plate on the bottom of the cooler spreads heat out evenly and averages the effect of heat output spikes.

The fans look identical. Even if they are different, it seems that in 7600 they are much worse, even unacceptable, since at 160 RPM less less they generate 3.5 dBA more noise than in 6600 XT. I was considering buying a RX 6600XT or 7600, and was initially leaning towards the 7600, but after this review I'm very hesitant - 3.5 dBA is a huge difference. I once read, but I don't know how true it is, that a 3 dBA difference means a 2x increase in noise

 

[Chrispy_]

I mean, I have the (vanilla) RX 7600 sat at home on an open test bench right now, having previously had a Pulse RX 6600, Pulse RX 6700 (10GB), and Pulse 6700XT all on the same test bench, and realistically the noise differences between them all were very minor. The dual-fan Pulses aren't the quietest cards on the market but they're adequately cooled and the sound profile they make seems to be very broad-spectrum so even when you can hear them it's an unobtrusive wind/air-movement sound rather than any kind of whine/whistle/howl that some fan/shroud/heatsink combos tend to make.