Jump to content
Main menu
Main menu
move to sidebar
hide
Navigation
Main page
Recent changes
Random page
freem
Search
Search
Appearance
Create account
Log in
Personal tools
Create account
Log in
Pages for logged out editors
learn more
Contributions
Talk
Editing
Openai/6907fc4f-3f34-8000-8487-2b4781f86078
(section)
Add languages
Page
Discussion
English
Read
Edit
Edit source
View history
Tools
Tools
move to sidebar
hide
Actions
Read
Edit
Edit source
View history
General
What links here
Related changes
Special pages
Page information
Appearance
move to sidebar
hide
Warning:
You are not logged in. Your IP address will be publicly visible if you make any edits. If you
log in
or
create an account
, your edits will be attributed to your username, along with other benefits.
Anti-spam check. Do
not
fill this in!
===== AI・LLM用途で見ると ===== # ピークAIスループット - RTX 5090の3,352 AI TOPSはR9700のINT4 1,531 TOPSのざっくり2倍強です。これはBlackwell世代TensorコアがFP4や低精度を前提にしているからで、Stable Diffusion系やTensorRT-LLMの最新ビルドを素で走らせると5090が有利になります。PC Watch<ref>{{cite web|title=PC Watch|url=https://pc.watch.impress.co.jp/docs/topic/feature/1656874.html|publisher=PC Watch|access-date=2025-11-13}}</ref> - R9700もINT4 1,531 TOPSでこのクラスとしてはかなり強く、300Wでここまで出るのは効率が良いです。sapphiretech.com<ref>{{cite web|title=sapphiretech.com|url=https://www.sapphiretech.com/ja-jp/commercial/radeon-ai-pro-r9700|publisher=sapphiretech.com|access-date=2025-11-13}}</ref> # ソフトウェア栄養価 - 5090はCUDA 13世代でTensorRT-LLMや各種OSSがどんどん追随しているので「新しいモデルを持ってきたらとりあえず動く」側。WSL2でも事例が出ています。Reddit<ref>{{cite web|title=Reddit|url=https://docs.nvidia.com/deeplearning/tensorrt/latest/getting-started/support-matrix.html|publisher=docs.nvidia.com|access-date=2025-11-13}}</ref> - R9700はROCm 7.0.xで公式サポートが入り、vLLMのテストでもRTX 6000 Adaと良い勝負をしていて、2枚構成だと追い抜くケースもありますが、Linux前提がまだ強いです。Windowsは2025/9時点でPyTorch on AMDが「公開プレビュー」で、実用はできますがNVIDIAほどの“そのまま感”はありません。gpuopen.com<ref>{{cite web|title=gpuopen.com|url=https://www.phoronix.com/review/amd-radeon-ai-pro-r9700|publisher=phoronix.com|access-date=2025-11-13}}</ref> # メモリの質 - 両方32GBなので「量」では並びますが、帯域は5090が約1.8TB/sでR9700の約640GB/sの3倍近いです。RLHFやMoEでアクティベーションをたくさん回すと効いてきます。NVIDIA<ref>{{cite web|title=NVIDIA|url=https://www.nvidia.com/en-us/geforce/graphics-cards/50-series/rtx-5090/|publisher=nvidia.com|access-date=2025-11-13}}</ref> # 複数枚運用 - R9700は300W・ブロワ・2スロットなので、ワークステーションに4本指すような使い方をかなり現実的にできます(AMD自身もマルチGPU推し)。asrock.com<ref>{{cite web|title=asrock.com|url=https://www.asrock.com/Graphics-Card/AMD/Radeon%20AI%20PRO%20R9700%20Creator%2032GB/index.jp.asp|publisher=asrock.com|access-date=2025-11-13}}</ref> - RTX 5090を4本は電源も冷却もかなり大掛かりになります。1〜2枚で最高性能を引き出すカードと考えた方が自然です。The Verge<ref>{{cite web|title=The Verge|url=https://www.theverge.com/2025/1/6/24337396/nvidia-rtx-5080-5090-5070-ti-5070-price-release-date|publisher=The Verge|access-date=2025-11-13}}</ref>
Summary:
Please note that all contributions to freem are considered to be released under the Creative Commons Attribution-ShareAlike 4.0 (see
Freem:Copyrights
for details). If you do not want your writing to be edited mercilessly and redistributed at will, then do not submit it here.
You are also promising us that you wrote this yourself, or copied it from a public domain or similar free resource.
Do not submit copyrighted work without permission!
Cancel
Editing help
(opens in new window)