This is ostensibly more of a workstation/dev thing. The integrated GPU is more or less like a very power efficient laptop 4070/4080 with unlimited VRAM, depending on which APU you pick, and the CPU is very fast, with desktop Ryzen CCDs but double the memory bandwidth of what even an 9800 X3D has. In that sense, it’s a steal compared to Nvidia DIGITs or an Apple M4 Max, and Mini PC makers alternatives haven’t really solidified yet.
I think Framework knows they can’t compete with a $900 Walmart laptop and the crazy bulk pricing/corner cutting they do, nor can they price/engineer things (with the same bulk discounts) at the higher end like a ROG Z13/G14.
So… this kinda makes sense to me. They filled a gap where OEMs are enshittifying things, which feels very framework to me.
The integrated GPU is more or less like a very power efficient laptop 4070/4080 with unlimited VRAM
I’d love to see some benchmarks comparing the two. I don’t believe that this integrated GPU is able to hold a candle to a laptop’s 4070, but I’d love to be proven wrong. It also looks, per the words of the article, the CPU with performance comparable to a “mid-range” laptop GPU (think xx60s) will cost $2000. Literally paying OVER TWICE the fucking price as a laptop with a 4070. AND LAPTOPS COME WITH SCREENS! This is a complete scam.
As far as CPU speed goes, couldn’t people simply choose whatever CPU they want when building their computer without paying the “framework tax”? Are you trying to claim there’s something “special” about this CPU like the PS3’s cell processor?
Are you trying to claim there’s something “special” about this CPU like the PS3’s cell processor?
There is. It’s a 16-core desktop CPU kinda like Ryzen, but hooked up to an IO die with very fast (8533MHz) quad channel memory. It’s more like a small, power efficient threadripper CPU, if you want to look at it that way, and that is a $2K+ platform.
But that IO die has a 40 CU GPU (compared to 12 CUs for the previous highest end IGP, or 8 for the Steam Deck), so yeah, somewhere in the ballpark of a 4070, or (to be more accurate than my previous guess) the 4060 at the lower end of SKUs.
But the real appeal isn’t “magic”: its like an Apple M Pro or Mac, a decent GPU hooked up to a huge pool of reasonably fast VRAM without having to pay freaking $4000+ for a Quadro or A100. Workstation people are going crazy over this thing.
I’d buy it at the drop of a hat for workstation stuff. I also have some friends who want one as a home server, since it’s so powerful but power efficient, and more modular/repairable than some Chinese mini PC.
For $2000, you and your friends could be getting better hardware.
I’m sorry, this is a product for suckers. Anyone who buys this has more money than sense. It’s just a fact of life.
It’s okay if you don’t realize this now, but hopefully one day you will. It’s alright to be wrong and taken for a ride. We’ve all been there, including me. It’s part of why I’m so keen on identifying bullshit like this now; I’ve seen it before and will continue to see it again.
I’m sorry, this is a product for suckers. Anyone who buys this has more money than sense. It’s just a fact of life.
There’s literally no better hardware. If I want 48Gb+ vram for $2000, it’s literally a bank of ancient, power sucking Tesla P40s or this… there’s nothing else because Nvidia/AMD price gouge everything else. Heck, the used 24GB 3090 I bought has skyrocketed in price. I’d still be paying l$1500+ for a bare minimum 3090 system now.
I’d love to be wrong, but the GPU market is totally fucked.
GPU aside, it’s the same for those that want a really fast transcoder box or whatever.
Honestly I don’t even care if it’s half as fast as a 4070, that’s fast enough for me.
In workstation workloads, some stuff just will not run unless you have a ton of VRAM, and running slower is fine. Or in other cases, you get a gigantic speedup from the virtue of simply having tons of VRAM. That’s the value, not pure core speed compared to some 8GB GPU.
This is ostensibly more of a workstation/dev thing. The integrated GPU is more or less like a very power efficient laptop 4070/4080 with unlimited VRAM, depending on which APU you pick, and the CPU is very fast, with desktop Ryzen CCDs but double the memory bandwidth of what even an 9800 X3D has. In that sense, it’s a steal compared to Nvidia DIGITs or an Apple M4 Max, and Mini PC makers alternatives haven’t really solidified yet.
I think Framework knows they can’t compete with a $900 Walmart laptop and the crazy bulk pricing/corner cutting they do, nor can they price/engineer things (with the same bulk discounts) at the higher end like a ROG Z13/G14.
So… this kinda makes sense to me. They filled a gap where OEMs are enshittifying things, which feels very framework to me.
I’d love to see some benchmarks comparing the two. I don’t believe that this integrated GPU is able to hold a candle to a laptop’s 4070, but I’d love to be proven wrong. It also looks, per the words of the article, the CPU with performance comparable to a “mid-range” laptop GPU (think xx60s) will cost $2000. Literally paying OVER TWICE the fucking price as a laptop with a 4070. AND LAPTOPS COME WITH SCREENS! This is a complete scam.
As far as CPU speed goes, couldn’t people simply choose whatever CPU they want when building their computer without paying the “framework tax”? Are you trying to claim there’s something “special” about this CPU like the PS3’s cell processor?
There is. It’s a 16-core desktop CPU kinda like Ryzen, but hooked up to an IO die with very fast (8533MHz) quad channel memory. It’s more like a small, power efficient threadripper CPU, if you want to look at it that way, and that is a $2K+ platform.
But that IO die has a 40 CU GPU (compared to 12 CUs for the previous highest end IGP, or 8 for the Steam Deck), so yeah, somewhere in the ballpark of a 4070, or (to be more accurate than my previous guess) the 4060 at the lower end of SKUs.
But the real appeal isn’t “magic”: its like an Apple M Pro or Mac, a decent GPU hooked up to a huge pool of reasonably fast VRAM without having to pay freaking $4000+ for a Quadro or A100. Workstation people are going crazy over this thing.
I’d buy it at the drop of a hat for workstation stuff. I also have some friends who want one as a home server, since it’s so powerful but power efficient, and more modular/repairable than some Chinese mini PC.
For $2000, you and your friends could be getting better hardware.
I’m sorry, this is a product for suckers. Anyone who buys this has more money than sense. It’s just a fact of life.
It’s okay if you don’t realize this now, but hopefully one day you will. It’s alright to be wrong and taken for a ride. We’ve all been there, including me. It’s part of why I’m so keen on identifying bullshit like this now; I’ve seen it before and will continue to see it again.
There’s literally no better hardware. If I want 48Gb+ vram for $2000, it’s literally a bank of ancient, power sucking Tesla P40s or this… there’s nothing else because Nvidia/AMD price gouge everything else. Heck, the used 24GB 3090 I bought has skyrocketed in price. I’d still be paying l$1500+ for a bare minimum 3090 system now.
I’d love to be wrong, but the GPU market is totally fucked.
GPU aside, it’s the same for those that want a really fast transcoder box or whatever.
You’re deluding yourself into thinking this iGPU is more powerful than it actually is.
Have you seen benchmarks comparing it to some of the competition?
Honestly I don’t even care if it’s half as fast as a 4070, that’s fast enough for me.
In workstation workloads, some stuff just will not run unless you have a ton of VRAM, and running slower is fine. Or in other cases, you get a gigantic speedup from the virtue of simply having tons of VRAM. That’s the value, not pure core speed compared to some 8GB GPU.
But I am not deluding myself, the core performance is in the ballpark of a laptop 7700S: https://www.notebookcheck.net/AMD-Radeon-RX-8060S-Benchmarks-and-Specs.942049.0.html
It’s also (on paper) slower than my desktop 3090, but in practice, an order of magnitude faster for stuff that struggles to fit on the 3090.
You’re the one who’s trying to argue its power, and then when presented with a better option you say “it’s fast enough for me.”
Have you seen benchmarks comparing the performance of this iGPU to dGPUs?
Please share.
See the notebookcheck page above.