Yesterday the embargo lifted on the Nvidia RTX 4060 Ti and today it hits store shelves, but AMD is hoping that before you consider that you take a look at their new RX 7600. Today the 7600 embargo lifts and we can dive into what it is all about and check out its performance, then it hits store shelves tomorrow. The RX 7600 is targeting the mainstream segment as well which means it is designed for 1080p performance and is priced to match the RTX 4060 which Nvidia announced for a July launch. Mainstream users can finally get excited, the next generation of cards is finally here, and with 64.52% of Steam users on the latest survey still gaming at 1080p that is still by far the largest market. So I’m excited to see how the new RX 7600 performs!

Product Name: AMD Radeon RX 7600

Review Sample Provided by: AMD

Written by: Wes Compton

Amazon Affiliate Link: HERE

 

All about the RX 7600

I spoke about it in yesterday's Nvidia RTX 4060 Ti review and I mentioned it in the opening here but it is easy to forget that a majority of gamers are playing at 1080p. With so many new cards coming out and being so capable at 1440p and 4k you would think that is what most people are using. So with the 65% of gamers who are playing at 1080p a lot of those higher-end cards might get them higher frame rates but in most cases, it would be overkill. Obviously, if you are planning a monitor upgrade soon it would still make sense to focus on a card capable of that. But if 1080p is still in your future, why not focus on cards designed specifically for that? The RX 7600 that we are looking at today falls into that category but if you were curious you can usually get a good idea of what a card's target is just by looking at the VRAM. AMD has their lineup broken up this way and Nvidia is similar as well. You can see AMDs current lineup in the picture below from AMDs presentation, interestingly we can see that right now they still have the RX 6650 XT and the RX 6600 as well in their 1080p focus, and because they skipped from their new RX 7900 XT, and RX 7900 XTX down to the RX 7600 everything in the lineup in between that is still from the 6000 Series.

When comparing the new RX 7600 to the previous generation card, the RX 6600 we can see that the transistor count has increased from 11.1 billion to 13.3 billion but with the smaller 6nm build process the die size has also gotten smaller going from 237 mm² to 204 mm². In that smaller space, AMD has increased the compute unit count from 28 up to 32 and the new architecture also adds in the new AI accelerators as well. The stream processor count has improved from 1792 up to 2048. The RX 7600 also has higher game and boost clock speeds with the new boost clock at up to 2655 MHz. All of this together has made huge improvements in the single precision performance which on the RX 6600 was up to 8.93 TFLOPS and the RX 7600 is now up to 21.75 TFLOPS. The 8GB GDDR6 hasn’t changed at all, this is still a 1080p focused card and the RX 7600 still has a 128-bit memory bus but AMD did increase the memory speed up to 18 Gbps. With the higher clock speeds, even the power efficiency that the smaller build process wasn’t enough, the RX 7600 has a higher total board power of 165 watts compared to the RX 6600’s 132 watts. The smaller die size did help with the pricing as well as not being in the middle of a crypto (and covid) fueled card shortage with the price dropping from $329 of the RX 6600’s launch price down to $269 for the RX 7600.

Specifications	Radeon RX 7600	Radeon RX 6600
Architecture	RDNA 3	RDNA 2
Manufacturing Process	6nm	7nm
Transistor Count	13.3 billion	11.1 billion
Die Size	204 mm²	237 mm²
Compute Units	32	28
Ray Accelerators	32	28
AI Accelerators	64	-
Stream Processors	2048	1792
Game GPU Clock	2250 MHz	2044 MHz
Boost GPU Clock	Up to 2655 MHz	Up to 2491 MHz
Peak Single Precision Perf.	Up to 21.75 TFLOPS	Up to 8.93 TFLOPS
Peak Half Precision Perf.	Up to 43.5 TFLOPS	Up to 17.86 TFLOPS
Peak Texture Fill-Rate	Up to 339.8 GT/s	Up to 279 GT/s
ROPs	64	64
Peak Pixel Fill-Rate	Up to 169.9 GP/s	Up to 159.4 GP/s
AMD Infinity Cache	32 MB (2nd Gen)	32 MB (1st Gen)
Memory	8GB GDDR6	8GB GDDR6
Memory Speed	18 Gbps	14 Gbps
Effective Memory Bandwidth w/AMD Infinity Cache	Up to 476.9 GB/s	Up to 412.9 GB/s
Memory Bus Interface	128-bit	128-bit
PCIe Interface	PCIe 4.0 x8	PCIe 4.0 x8
Total Board Power	165W	132W
Launch MSRP	$269	$329

 

AMDs presentation did also highlight that they are supporting AV1 which is starting to pick up and will be very important. AV1 allows for better video quality at lower bitrates which has a few important uses. For streamers or if you are streaming your game on Discord to your friends this allows for better quality without clogging up your upload connection for those of us with cable internet. It’s not only streamers who will benefit though, AV1 support means less data transferred for companies like YouTube as well.

Before getting into testing I did also run GPUz to double-check that our clock speeds match up with the specifications. Our default clock speed lines up but the boost clock in GPUz is showing 2655 MHz which does line up with AMDs specification. GPUz also has our Bios revision in case that information is needed in the future and shows the driver I tested with which is the 23.140.01.16 Beta driver that AMD provided to the press ahead of the launch.