2060 KO is actually a 2080 Chip?

[TheoneandonlyMrK]

So you're saying that they should always announce that there might be a new better version of what you are buying coming out at some point in the future whenever you buy something. Yeah, the world doesn't work that way and not doing so is in no way dishonest, not even borderline.

No not really , just that this info out there can inform the minority it matters to, no drama required, your hyping my opinion of what i might think?,I didn't buy one, yet.

It's ok to dissagree btw , people can be dramatic about opinions, Even borderline ones ,remember they're personal not even I expect to change yours and my opinions are not me just an expression of my thoughts.

 

[efikkan]

Not so possibly, I watched the same video and while games perfom as expected, close to a prior 2060 the KO performs /50% better in pro apps, no reason known.

I don't agree with Steve at GamersNexsus' theory about differences(?) in fp32 units is causing the performance difference between TU104-150 vs TU106, that would have caused a performance delta across any workload, including gaming, which is very fp32 intensive. The chips in a GPU family have mostly the same configuration except for the scalable part of the design, and while there can sometimes be minor differences in cache or register files etc. that makes a minor differences, it would then benefit most workloads. My theory is the largest difference between dies (which most people keep forgetting about); the scheduling/management engine. We've seen this kind of behavior with cut down dies before, like GTX 970 vs. GTX 980, where the GTX 970 gained more performance per Gflop/s. TU104-150 is a very significantly "cut down" die, so the management resources vs. resources to manage should be plentiful, which is quite likely a contributing factor to only certain compute workloads scaling better. (While gaming usually stresses the management engine differently)

2. Cache. TU104 has more cache. How much of it is left enabled on 2060 with TU104 is a good question. Is anyone able to check?

What matters here is cache per computational resource. I don't know how much of the L2 cache is enabled, but if all of it were, I would imagine a performance difference in most/all workloads.

 

[Darmok N Jalad]

I get what you are saying. All of the cards you mentioned have the same chip and for me the main difference is in the cooler. There is also the level of factory OC that I see in different models of the same GPU, not the chip that resides on the PCB.

This has still been done before. The HD 7870 XT was a 7870 based on a cut down version of the 7970.

 

[TheoneandonlyMrK]

This has still been done before. The HD 7870 XT was a 7870 based on a cut down version of the 7970.

The 7870 was not a cutdown 7970 ,the 7870 and Xt were the same cut down silicon 7970 7950 were different.

 

[Darmok N Jalad]

The 7870 was not a cutdown 7970 ,the 7870 and Xt were the same cut down silicon 7970 7950 were different.

No, the 7870 XT was based on the larger Tahiti chip, while the original 7870 was based on the Pictarin chip. The 7870 XT launched later, at a price premium, and performed better than the Pictarin-based 7870. I may be getting the “XT” name mixed up, but there were still 2 7870 models based on different chips.

PowerColor PCS+ Radeon HD 7870 (Tahiti LE) 2GB Myst Video Card Review

We check out the new Tahiti LE based HD 7870 from PowerColor and give you the low down.

www.tweaktown.com

 

[TheoneandonlyMrK]

No, the 7870 XT was based on the larger Tahiti chip, while the original 7870 was based on the Pictarin chip. The 7870 XT launched later, at a price premium, and performed better than the Pictarin-based 7870. I may be getting the “XT” name mixed up, but there were still 2 7870 models based on different chips.

PowerColor PCS+ Radeon HD 7870 (Tahiti LE) 2GB Myst Video Card Review

We check out the new Tahiti LE based HD 7870 from PowerColor and give you the low down.

www.tweaktown.com

Just checked tpu database sorry your right.

But then that had a different name and wasn't this.

 

[Darmok N Jalad]

Just checked tpu database sorry your right.

But then that had a different name and wasn't this.

Actually, the 7870 example was probably worse—the terms I’ve been using are technical terms, not marketing terms. The Tahiti LE card in the review I posted was called the “PowerColor PCS+ HD 7870 2GB Myst Edition,” which doesn’t differentiate from a Pictarin card in any obvious sense. It’s the usual case of “buyer beware,” as these naming shenanigans have been going on for a long time. Both AMD and Nvidia have been guilty of it. I don’t know why they don’t just make a new model, but maybe that is harder for partners to grasp. I don’t know.

 

[TheoneandonlyMrK]

Actually, the 7870 example was probably worse—the terms I’ve been using are technical terms, not marketing terms. The Tahiti LE card in the review I posted was called the “PowerColor PCS+ HD 7870 2GB Myst Edition,” which doesn’t differentiate from a Pictarin card in any obvious sense. It’s the usual case of “buyer beware,” as these naming shenanigans have been going on for a long time. Both AMD and Nvidia have been guilty of it. I don’t know why they don’t just make a new model, but maybe that is harder for partners to grasp. I don’t know.

Naming at the same spec we decided was fine just before anyway your quoting a guy who said so what, like some un abashed defender, no need or do you think there is.