• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

AMD 7600 reviews

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Even if the engineers explain to me I'll never understand this "stupid" decision. Why put so much time and resources engineering a feature that they will not be able to make use?
Because you don't know that ahead of time. Both developments (hardware, drivers) go in parallel.
 
Ok something not working correct with NAVI33.

According to AMD the new RDNA3 Compute Unit is 17% faster vs RDNA2 clock to clock.
Also according to AMD the new RT Cores in RDNA3 are 50% faster vs RDNA2 RT Cores.

Since both NAVI 23 and 33 have the same Compute Units (32) and number of RT Cores (32) and it seems both 6650XT and 7600 have almost the same gaming clocks , then RX 7600 should have been 15%+ faster in Raster and up to 50% faster in RT.
According to the reviews nothing of the above was measured.

resize
\

rdna-3-ray-tracing.jpg
 
Last edited:
It has about 15% smaller die than Navi 23, N6 vs. N7 to offer additional fabrication savings, and performance increase between 5% and 10%.

People saying there is no point in introducing this card are not thinking straight.

They are comparing last gen card that have been partially written off and therefore heavily discounted vs. what is sustainable and manufacturable going forward.
Navi 23 has been at these low prices since last November. There's no doubt AMD has been making these knowing they must sell at these heavily discounted prices.

Their margin will be improving a lot (if any sell) since it is +70$ from current market price of RX6600, the die is cheaper to make too. The problem is they are more competing with RX 6600 XT. So it ends up being 20$ more for a slightly smaller die.
 
Ok something not working correct with NAVI33.

According to AMD the new RDNA3 Compute Unit is 17% faster vs RDNA2 clock to clock.
Also according to AMD the new RT Cores in RDNA3 are 50% faster vs RDNA2 RT Cores.

Since both NAVI 23 and 33 have the same Compute Units (32) and number of RT Cores (32) and it seems both 6650XT and 7600 have almost the same gaming clocks , then RX 7600 should have been 15%+ faster in Raster and up to 50% faster in RT.
According to the reviews nothing of the above was measured.

resize
\

rdna-3-ray-tracing.jpg

It's more likely AMD lied and/or that IPC gain in only possible when you increase the ROPS/bandwidth.

And even with that, AMD 17.4 percent does not make sense.

20% more shaders + 17.4% IPC + 13% clock advantage means Navi 31 should be almost 60% faster than a 6950XT. But it is only 35%.

What Navi 33 launch shows if anything is the method AMD used to double tflop count was very cheap in terms of silicon which meant none of this really was going to turn into actual performance in games broadly speaking. You can't magically increase IPC without spending silicon typically, particularly for an architecture as efficient as RDNA2.
 
Ok something not working correct with NAVI33.

According to AMD the new RDNA3 Compute Unit is 17% faster vs RDNA2 clock to clock.
Also according to AMD the new RT Cores in RDNA3 are 50% faster vs RDNA2 RT Cores.

Since both NAVI 23 and 33 have the same Compute Units (32) and number of RT Cores (32) and it seems both 6650XT and 7600 have almost the same gaming clocks , then RX 7600 should have been 15%+ faster in Raster and up to 50% faster in RT.
According to the reviews nothing of the above was measured.

resize
\

rdna-3-ray-tracing.jpg

I was going to suggest maybe it is being let down by the small IC or the general memory bandwidth, but it seems like the 4060ti can do more raw performance with the same general layout (at a very high level).

Perhaps its still memory bandwidth related? Was much said about better compression, IC efficiency, etc. with RDNA3? I have not been able to keep up. The 7600 doesn't have much (any?) more of this than the 6650XT does.
 
Ok something not working correct with NAVI33.

According to AMD the new RDNA3 Compute Unit is 17% faster vs RDNA2 clock to clock.
Also according to AMD the new RT Cores in RDNA3 are 50% faster vs RDNA2 RT Cores.

Since both NAVI 23 and 33 have the same Compute Units (32) and number of RT Cores (32) and it seems both 6650XT and 7600 have almost the same gaming clocks , then RX 7600 should have been 15%+ faster in Raster and up to 50% faster in RT.
According to the reviews nothing of the above was measured.

resize
\

rdna-3-ray-tracing.jpg

RDNA3 has an R520 feel. Let's hope a refresh can give us something more like R580.
 
Looks like Newegg has a 30 day return policy. I asked a customer rep about a price match a while back on a nvme drive and from what I remember they gave me a promo credit that I think had an expiration date in something like 90 days? which I was ok with. It was a bit over a year ago, so the details are a bit fuzzy.
& @DAPUNISHER

Did a chat on Newegg and turns our the return policy is 14 days and yesterday was my 14th day. 🙁 I'm not going to try getting around their policy, as I'm happy with the card and I thought the price was fair. In case anyone is curious about the policy, here is the link.

Price Match Guarantee

 
Navi 23 has been at these low prices since last November. There's no doubt AMD has been making these knowing they must sell at these heavily discounted prices.

Their margin will be improving a lot (if any sell) since it is +70$ from current market price of RX6600, the die is cheaper to make too. The problem is they are more competing with RX 6600 XT. So it ends up being 20$ more for a slightly smaller die.
We don't know when the last Navi 23 die came out of TSMC fab. Probably a long time ago.
The GPU collapse started in Q2 2022. AMD has probably been selling chips from inventory, at prices that reflect write-down of their value. AIBs had their inventories to sell, in market that collapsed.

There is no way that AMD can re-start manufacturing of Navi 23 and have the resulting cards more cost competitive than Navi 33. So Navi 33 is it for the low end for the next year. We may get a little bit of a discount, but that's it.

Like 6700 and 6800 XT that are running out, the same will happen with 6600 XT in not too distant future.
 
What? No discussion of 7600 temps so far?


1684961578190.png
Cooling solution seems to be inadequate. Not worth the money even if it goes on sale. Don't want my RAM chips dying an early heat death. Though I would get the snowy screen well before that death. Not a card I would want to use in summer.
 
& @DAPUNISHER

Did a chat on Newegg and turns our the return policy is 14 days and yesterday was my 14th day. 🙁 I'm not going to try getting around their policy, as I'm happy with the card and I thought the price was fair. In case anyone is curious about the policy, here is the link.

Price Match Guarantee

Your attitude is understandable John.

In your position, I would have told the rep that either they issue the refund for the difference or I leverage the 30 day hassle free return. They pay shipping and there is no restock fee. Then I would have reordered the card. It benefits them to give the refund instead of getting stuck with shipping and what is now an open box card. Rep could ask the supervisor to approve it.

It was enough money to be worth the effort IMO, but again, your position on it is reasonable.
 
I don't know what AMD was planning with the 7600, but it's just sad. They should have priced this at $249. However, you can count on AMD making a bad launch worse with the last minute price change.

Navi 33 is basically equal to Navi 23, in almost everything. Just awful.
 
I was expecting Navi 33 to be an actual improvement over the 6650XT. I have zero interest in ambiguous "AI" accelerators. This is depressing.
 
Last edited:
Maybe they are the secret sauce behind FSR3 which finally brings serious competition for DLSS. We can dream 🙂
I'm starting to suspect that it's just a matrix of bolted on FPGA IP blocks that will do jack unless specifically programed for. Which probably won't happen (at least not for consumer scenarios).
 
They should have made this card exclusive to OEM desktop for the time being. There was no reason to release this to DIY right now given the available alternatives. They would have looked a lot better dropping this to DIY around $200 in the fall when the last gen stuff dries up. I thought this card was going to be 10% better (which would have been less meh at $270) and was still going to end up selling around $200 by the end of the year. Given the reviews it will probably be $200 by July.
 
They should have made this card exclusive to OEM desktop for the time being. There was no reason to release this to DIY right now given the available alternatives. They would have looked a lot better dropping this to DIY around $200 in the fall when the last gen stuff dries up. I thought this card was going to be 10% better (which would have been less meh at $270) and was still going to end up selling around $200 by the end of the year. Given the reviews it will probably be $200 by July.

-Imagine the waves this thing would have made launching at $199. Jesus Christ these things would have come in hot and had a hard time staying on shelves.

People would be calling it the RX580 reborn and stuff.

AMD still has a narrow window to bring in a 28CU model for $229, maybe bring back the LE moniker to avoid the x5xx stigma.
 
They should have made this card exclusive to OEM desktop for the time being. There was no reason to release this to DIY right now given the available alternatives. They would have looked a lot better dropping this to DIY around $200 in the fall when the last gen stuff dries up. I thought this card was going to be 10% better (which would have been less meh at $270) and was still going to end up selling around $200 by the end of the year. Given the reviews it will probably be $200 by July.

It's not like AMD is spamming SKUs. It is a single SKU.

It seems its presence is upsetting you. Is the presence of any other graphics card SKUs upsetting you?
 
It's not like AMD is spamming SKUs. It is a single SKU.

It seems its presence is upsetting you. Is the presence of any other graphics card SKUs upsetting you?
Not upset just confused. Not sure why they bothered releasing this to DIY right now. Seems rather pointless unless it’s replacing the 6600’s current selling price of $200.
 
It's not like AMD is spamming SKUs. It is a single SKU.

It seems its presence is upsetting you. Is the presence of any other graphics card SKUs upsetting you?
I think what is upsetting is that AMD released a single SKU that is not competitive with the disappointing 4060ti.
 
Not a great launch. I think if the 7600 is compared to the 6650 XT, the performance increase over the 6600 is there, but not huge. Would make more sense is to price it lower. Or price it lower and call it the 7500XT or something.

That said, I will say that the good parts of this generation are the 7900XTX, and to a lesser degree the 7900 XT and RTX 4070. The 7900 XTX has some great performance, along with plenty of memory, as does the 7900XT to a degree, we just needed to wait for prices to come down a bit. The 4070 is only 12GB, but it's performance and memory amount is ok for what it is, and is one of the few Geforce cards to offer 8 pin connectors on the current gen, though price still seems a bit high. Which brings us to the 4090. A top tier card, unfortunately with that, it is quite expensive. It would still be a decent halo product, except that it only has the ghastly fire connector, so I wouldn't personally consider one, and I think many others share my view there.
 
Back
Top